Go Back   EQEmulator Home > EQEmulator Forums > Development > Development::Development

Development::Development Forum for development topics and for those interested in EQEMu development. (Not a support forum)

Reply
 
Thread Tools Display Modes
  #1  
Old 05-26-2024, 05:08 AM
Torven
Sarnak
 
Join Date: Aug 2014
Posts: 61
Default A Treatise on EverQuest's Experience Calculations

This treatise covers everything I understand about EverQuest's experience calculations. My research focus is older eras of EQ (up though Planes of Power) but much of it is still relevant to Darkpaw servers as of this writing. When I say "Darkpaw" I mean EverQuest in 2024 or recent to 2024. When I say "Sony" I generally mean EQ circa 1999-2004. Read carefully so you do not mix up my explanations of modern EQ with old EQ as there have been many changes over the years and I will sometimes compare the two.

In EverQuest a player's progress is represented by a single experience value that is the total lifetime sum of every experience gain minus every experience loss. (can't be negative) This is stored in a very large integer number. From this single value the player's level is calculated formulaically. This is a bit unusual as the level is subordinate to the experience instead of vice versa. It makes some sense for them to have done this in a game where delevleing is possible, and it makes changing levels after an experience gain or loss easy to calculate. The downside is that it's inflexible to modify: you cannot easily change the amount of experience required for a specific level for example. Thus EQ required some kludgy workarounds which I will explain in the relevant sections.


The Basic Math

There are two basic functions EQ uses for experience: (brackets mean doesn't always apply)

Exp required to level = player_level^3 * class_mod * race_mod * hell_mod

Exp gained from kills = mob_level^2 * ZEM [* group_bonus] [* group_split] [* class_mod] [* HBM] [* MLM]

ZEM stands for "Zone Experience Modifier". (or multiplier if you like) All zones have a ZEM value. A ZEM of 75 is the baseline and is what most outdoor zones had.

Level exp per kill in old EQ was capped at one eighth of the level's exp, which is 12.5%. On Darkpaw's servers this cap is 7.5% at very low levels, which is a rare case of modern EQ being less forgiving, although there is a weird scaling exp bonus that is added after the exp is capped starting at level 8 and increases to 10% after a dozen or so levels. AA exp never had a cap.

class_mod is a number from 9 to 14 depending on the player's class:

Warrior: 9
Rogue: 9.05
Cleric, Druid, Shaman: 10
Wizard, Enchanter, Magician, Necro: 11
Monk: 12
Ranger, Paladin, Shadowknight: 14

The class_mod was applied to exp gains to cancel out the class penalties after January 2001 for all classes except rogues and warriors. This does not apply to AA exp. (this mod would be divided by 10 first or the entire result would be divided by 10 such that a Ranger for example would get 40% more exp if using this double digit notation)

race_mod is a value from 95 to 120 depending on the player's race:

Halfling: 95
Barbarian: 105
Ogre: 115
Troll, Iksar: 120
Everybody else: 100

Note that I will be using the words "modifier" and "multiplier" interchangeably. These terms are all player invented and what Sony called them is unknown. Also Sony's numbers may have had the decimals in different places, e.g. class_mod may have been three digits and race_mod two or whatnot, and sometimes ZEMs are represented in decimal format, but the end result would be identical as long as the decimals were moved complementary since they're all multiplications.

The group_split multiplier is a number between 0 and 1 when sharing exp in a group and works more or less like one might expect (e.g. it's 0.2 in a five-man group of the same levels) although there is more to it. See the groups section for details. In Luclin two new multipliers were added which are called the HBM and MLM for short in this document and are discussed in their sections. Not shown above is a possible server bonus multiplier (Al'Kabor had 20%, so 1.2) and a possible individual NPC multiplier which certain Planes of Power NPCs had.

The exp gain formula on Darkpaw servers has changed significantly and I do not understand how it works now. The above formula should be considered accurate only for older eras of EQ. Exp from kills on Darkpaw servers is significantly higher than in old EQ and I don't know why. A race modifier to cancel racial penalties and bonuses is likely applied to exp gains in modern EQ however.


Hell Levels and hell_mod

hell_mod is a number Sony used to make the higher level game take longer to advance. (require more kills to level) For levels 1 through 29 this value is 1.0. For levels 30-34 this value is 1.1; For levels 35-39, 1.2; For levels 40-44, 1.3; For levels 45-50, 1.4. Kunark levels increase this every level and often by more than 0.1. This number is where so-called "hell levels" come from.

Without the hell_mod, the increase in exp required per level from the previous level quickly trends toward zero percent. For example, level 4 requires nearly twice as much exp to level as level 3 does, but level 29 requires only about 7.4% more exp than level 28 does. Exp gain also trends toward being flat, so without a hell_mod the amount of kills needed to level would otherwise end up about 25 white con kills for all levels with a ZEM of 75.

Hell levels (a term players came up with) were levels that required twice as much exp to complete as a normal level. The hell levels were 30, 35, 40, 45. These arise because the hell_mods are not increased gradually and instead are constant for each tier of levels. Since the exp required for a level is the total lifetime sum of exp, multiplying that number by 1.1 after previously multiplying by 1.0 is essentially multiplying all the previous 29 levels combined by an extra 10% instead of just the singular level 30. If Sony had increased it gradually each level then the hell level effect would be far less noticeable.

Since Kunark only had 10 levels, and because of the flattening exp curve, and to prevent previous levels requiring more exp then the current level, Sony had to crank up the hell_mod every level. Levels 54 and 59 feel like hell levels (they're less than double previous level exp however) because at those levels the hell_mod increases more than in previous levels. (from +0.1 to +0.2 at level 54 and from +0.2 to +0.3 at level 59) It's the rate of increase of the hell_mod which creates the hell level effect. Level 60 is a fast level because the hell_mod doesn't increase at all there when the previous level it had increased by 0.3, so the exp required to go from 59 to 60 is much less than from 58 to 59.

The hell_mods for Kunark levels are: 51 1.5, 52 1.6, 53 1.7, 54 1.9, 55 2.1, 56 2.3, 57 2.5, 58 2.7, 59 3.0, 60 3.0.

Some old exp calculators list level 60 as 3.1 but that is incorrect. We can know the hell mods for 1-60 because they were in the client.

The hell_mods for PoP levels are: 61 3.225, 62 3.45, 63 3.675, 64 3.9, 65 4.125. An increase of 0.225 per level.

I know PoP's hell_mods because of two reasons: one, I estimated them from an Al'Kabor log of my bard being powerleveled and I was able to add up the kills and estimate the experience gains and how much gain was required to level. Also because Darkpaw's servers now show player experience within 1/1000th of a percent; that plus knowing how much numeric experience is lost on death allows me to compute a very close estimate. (see death section for details) The Al'Kabor data and Darkpaw data matched. At first I thought it might be +0.2 and my AK estimates were off, but they were off by about the same amount in the same direction, and the Darkpaw data is very conclusive. My AK estimates ended up being very close using +0.225.

Hell levels were effectively removed in the Luclin era with bonus exp multipliers. I'll go over that in a later section. It's important to understand that the hell_mods have never changed.


Simple Exp Example

A human cleric needs:

1,000 lifetime experience to reach level 2
8,000 lifetime experience to reach level 3
27,000 lifetime experience to reach level 4
64,000 lifetime experience to reach level 5

level 2 requires 1,000 additional experience to complete from 0% in level 1
level 3 requires 7,000 additional experience to complete from 0% in level 2
level 4 requires 19,000 additional experience to complete from 0% in level 3
level 5 requires 37,000 additional experience to complete from 0% in level 4

As a consequence of this system, changing a player's experience value may cause the player's level to change by more than one. If you set a player's total experience value on the server to 40,950 administratively then they will end up level 4.5, no matter what their starting level was, when they kill an experience granting mob or they zone. (a little higher if killing a mob as the kill exp will be added)

Incidentally a cap on experience didn't exist before Kunark. You could fill up your experience bar at level 50 and then continue to add more exp after, then you could die and not visibly lose exp in your bar until it fell below the "full exp" value for level 50. When Kunark launched they removed this excess exp so players didn't log in at a level above 50 on day one of Kunark.

See this link for a classic experience calculator: http://web.archive.org/web/200206122...aps/eqexp.html


Race and Class Penalties and Bonuses

Experience required per level is not the same for every player and changes depending on race and class. (modern EQ however cancels the differences out with multipliers) Prior to Velious, exp gained per kill was the same for everybody and the penalties/bonuses were represented in the exp needed for leveling. Class and race penalties do not apply to AA exp. (at least when exp is set to 100% AA exp. see the AA section for details)

Most races require the same exp. Barbarians, Ogres, Iksars and Trolls require more experience to level. Halflings require slightly less experience per level. The Halfling bonus was supposed to be for Humans but Sony made an error and accidentally applied it to Halflings. (I don't recall seeing a dev quote of this but that's the rumor) Because of the rigid experience system, they couldn't easily fix this without changing every Halfling and Human's levels when they logged in after the patch, so they left it that way. (technically they could have fixed this the same kludgy way they did with class modifiers but they didn't bother)

Classes started out requiring different amounts of experience to level when the game launched in 1999. Hybrids required significantly more (~40%) and casters required slightly more. Rogues and warriors required slightly less. (~10%) Priest/healing classes were the baseline.

The class penalties were effectively removed from the game in January 2001, which was shortly after Velious launched. To accomplish this they simply granted more experience per kill to the classes that required more exp to level by multiplying gains by class_mod, canceling out the penalty. Had they instead set the class_mods to 1.0 it would have granted players extra levels at login which they didn't want. They also applied the class_mod for experience loss on death, otherwise the extra exp would have significantly reduced death penalties for these classes. This probably inspired their hell level smoothing later on as it's a similar solution.

Race penalties were removed in the September 19 2006 patch. The class bonuses for Warriors, Rogues and Halflings were left in much longer, until 2019. It's worth mentioning that race stats do make a large difference in leveling speed at low levels, so it's not like Sony added the penalties for no reason.

This producer's letter describes the reasoning behind the January 2001 changes: https://web.archive.org/web/20010124...s/letter.shtml
Reply With Quote
  #2  
Old 05-26-2024, 05:10 AM
Torven
Sarnak
 
Join Date: Aug 2014
Posts: 61
Default

Light Blues and NPC Consider Color Effect on Exp Gains

The following applies to EverQuest prior to Serpent's Spine.

For the most part, consider colors have no bearing on the experience gained from a kill. I.e. killing a level 10 mob while at level 7 (when it's red) does NOT grant more experience points (I stress points, not exp gauge progress) than killing the same level 10 mob while at level 11. The experience gauge goes up more at level 7 from that kill because the experience required for that level is significantly less. Raw exp values used to be sent over the wire and if reds had awarded more than it would have been noticed by the ShowEQ guys.

Here is the results of an experiment somebody did in the late classic era (pre-Kunark) of a solo druid keeping track of how many kills he needed to level. The results did not indicate any sort of bonus exp for killing reds: https://groups.google.com/g/alt.game...m/0KuHXk_NRAoJ

There are two exceptions to con colors not mattering. One is light blue exp. (plus greens of course grant zero) The second is the high level kill bonus multiplier (MLM) that was put in just prior to PoP's launch; that will be covered in a later section.

Light blue cons were put into the game on November 7 2001. (very late Velious) This is the patch note:

Quote:
Originally Posted by Patch Note
- We have changed the way the /consider command works. We have added a
‘light blue’ area between green and blue. NPCs that used to /consider
green but gave experience will now /consider light blue, as well as
NPCs of slightly lower level than those greens. You will always get
experience for something that is light blue. You will not receive
experience for greens. At the same time, by including lower level NPCs
in the light blue /consider, we have increased the range of lower level
NPCs that will grant experience. This will be most noticeable for
characters of higher level.
Prior to that, at higher levels green cons that were nearly blue could also grant experience. (it was probably no more than two levels) They were the light blues before light blues.

I don't know exactly how the light blue experience reduction worked. I have found two explanations for it. It's very possible both are true and Sony changed light blue exp mid-Luclin. It seems to be the case however that the light blue reduction was not equal for all light blues, and higher level light blues awarded more experience than lower level ones.


Explanation 1: Light blues granted 50% or 25% normal exp, with the lower level ones granting 25%.

"if you kill a high light blue, exp is 50% of that, if you kill a low light blue, its 25%." (01-12-2002)
This post is by Casey Webster, a ShowEQ developer at the time, therefore has high credibility.
https://www.showeq.net/forums/showth...ill-experience

"Green cons that give exp are divided into 2 groups, high and low depending which of the 2 green/light blue levels is closer to you. high gives 50% exp, and low gives 25%. So basically, if its light blue by 1 level, you only get half exp, if by 3 levels 25%" (Mar 12, 2002)
Casey Webster Usenet post: https://groups.google.com/g/alt.game...m/m9N5Q__WctAJ

"Level * level * ZEM * Class <- Blue or higher, lightblue are .5 or .25 times this" (03-14-2002)
https://www.showeq.net/forums/showth...?675-EXP-Chart

I found several older comments which stated that high level greens gave half or one quarter exp like light blues did, but didn't state where they got this information. ShowEQ was starting to become well-known in 2000 however. Here are two, but I've found more:

"If the creature was green to a player, that player may get no experience or 1/2 or 1/4 of their share if the creature was barely green." (April 2000)
https://groups.google.com/g/alt.game...m/A71gx4EkTfAJ

"When the rat turns high green it will give 1/2 the experience, in this example it will give 50 exp points. If you level again and it still gives experience it will give 1/4 the experience, in this example 25 exp points." (March 2001)
https://groups.google.com/g/alt.game...m/EeuCSNfUc1AJ


Explanation 2: Light blues granted 80%, then 60%, then 20% exp.

One of the ShowEQ threads that was discussing the hell level removal multipliers (HBM) and the high level bonus kill multipliers (MLM) had data from users who were getting precise experience values from kills in September 2002. The results showed the two highest level light blues granted 80% then 60% experience with the next two light blues granting 20%. Since Sony was making large changes to exp code at the time, including a very short-lived nerf to low dark blues, it wouldn't be surprising at all if they also changed the light blue returns.
https://www.showeq.net/forums/showth...-6-patch/page4


I've found several posts which stated that green con exp had some quirks. (to reiterate: green con exp was a thing up until the end of Velious before LBs)

I've seen a few comments suggesting that green exp did not work in groups, or with some members getting the exp and some not. Here's one:

"solo, both of us get exp from these guys since they're high green to both of us. But I've been finding that when grouped, we don't get exp for high greens. Anyway, we were doing this at lunch today and the Paladin would get exp for all of the greens, and the warrior would not."
https://groups.google.com/g/alt.game...m/TEUdnSfNOdEJ

The pet mitigation might have worked differently on it:

"when dealing with green mobs of any kind, if the pet outdamages the mage, the mage gets no exp. If the mage outdamages the pet, though, he gets the normal green exp, be it 50% or 25%."
https://web.archive.org/web/20010302...ld/showeq.html


The dark blue exp range shrank after level 60 when Planes of Power launched, but this didn't last. At level 60 dark blue was 15 levels wide and this was the case from Kunark until well after PoP. (changed in Serpent's Spine?) When PoP launched dark blue range at 61 was 13 levels wide, at 62 it was 11 levels wide, then it stayed at 10 levels wide at 63 up to 65. This range was expanded to 15 levels wide for all PoP levels on June 11th 2003.

I found a 2002 ShowEQ thread discussing con color ranges here: https://www.showeq.net/forums/showth...-your-findings


The September 19 2006 (Serpent's Spine launch) patch changed the con colors in this way:

Green cons changed to gray.
Light blues changed to green.
Dark blues were split into dark blue and light blue.
Yellow cons changed to three levels instead of two.

These changes made sense due to the way the MLM works. See that section for details.


Elaboration on Zone Experience Modifiers

As mentioned previously, every zone has a 'ZEM' number. Sony used this as a way to make dangerous zones grant more experience or to entice players to go to underpopulated zones. Unfortunately they didn't put enough consideration into this, resulting in poor risk vs. reward when it comes to exp in many zones. ZEM values ranged from 75 to 100 in classic. In Luclin some lower level dungeons had large ZEM increases up to 160. 75 was common for outdoor zones.

Sony applied a ZEM override or floor (unsure which but some AK data suggests floor) for new players. The Stratics EXP calculator mentions this: "Varies based on zone and your level. For newbies it is 114 or 100." I can verify this from checking logs of players leveling and estimating their experience gains; they did seem to gain more experience than expected and in some cases 114 made the numbers fit well. However the margin of error for this is quite high due to the level variance of NPCs and the massive difference in exp per kill for each level of NPC at lower levels so I can't give estimates of what the ZEMs were other than the experience gains were much higher than a ZEM of 75 would grant. A Velious era newbie log with solo kills I parsed showed this at level 5 but not at level 8. I couldn't parse levels 6 and 7 because the log wasn't on when the character hit level 7.

Casey Webster was an early ShowEQ developer and he made this comment on March 12 2002:
"At lvl 6 and under, a "newbie" zem of 100 is applied to kills regardless of zone"
https://groups.google.com/g/alt.game...m/m9N5Q__WctAJ

This September 2000 Usenet post says it took 10 snakes for a Human Necromancer to reach level 2, or 20 snakes if the pet did all the damage. That would agree with a 114 ZEM:
https://groups.google.com/g/alt.game...m/SjHmH-BxS14J

Sony also applied some kind of experience bonus for players levels 3 to 9 during Luclin. I doubt that it's another ZEM floor or override as I saw this in Al'Kabor logs of players in Paludal, so it looks like an additional multiplier. This bonus seemed to increase with levels up to level 9 then was gone at level 10. Absor mentions this newbie experience bonus in the "EverQuest Third Anniversary State of the Game part 2" post on 03-15-2002:

Quote:
Originally Posted by Absor
When starting a new character for the first time, it can take quite some time to get into that class' core abilities. In some cases, the style of gameplay needs to shift dramatically. That being the case, we've accelerated the rate at which people gain experience up to level ten.
For a list of early PoP era ZEMs, see this: https://wiki.takp.info/index.php/Current_ZEMs

TAKP's ZEM list is accurate for the time period as it was packet sniffed from Al'Kabor. Sony changed ZEMs several times over the years however, so the remainder of this section will outline and date all of the known ZEM changes.

Dolalin did a lot of sleuthing of ZEM changes up through Velious and I highly recommend reading his thread on it and this post in particular:
https://www.project1999.com/forums/s...0&postcount=24

Here are the patch notes, dev comments and ShowEQ user comments regarding ZEM changes that I've found:

May 25, 2000 (early Kunark) - Patch note
Quote:
Originally Posted by Patch Note
Upon investigation, we discovered a deficiency with the amount of
experience being rewarded in many Kunark zones (mostly dungeons). The
affected zones have been updated, and adventurers should notice the
greater experience reward in return for the increased risk and
challenge associated with these zones.

January 17, 2001 (early Velious) - From Producer's letter
Quote:
Originally Posted by Producer's letter
As of the next patch, you will receive additional experience (per kill) in the following zones:

Droga increased by 12%
Nurga increased by 12%
Solusek's Eye (SolA) increased by 13%
Najena increased by 13%
Befallen increased by 13%
Paw increased by 13%
Permafrost increased by 13%
Kaesora increased by 18%
Qeynos Catacombs increased by 20%
Runnyeye increased by 20%
Kerra Ridge increased by 20%
The Hole increased by 25%

January 2002 (early Luclin) - Comment on ShowEQ's forum:
"Most outdoor zones, ZEM is 75, most dungeons, 80. Then there were all the beefed up dungeons a year or so ago, they have values like 90, 92"
https://www.showeq.net/forums/showth...ill-experience


March 7 2002 (early to mid Luclin) - Comment on ShowEQ's forum:
"ZEM varies based on zone, but in general:
75 Outdoors
85 Indoors
100 Dungeons

Then VI Revamed some of the unused dungeons giving them a ZEM as high as 125 (KerraRidge)."

https://www.showeq.net/forums/showth...e-XP-Modifiers


March 12 2002 (early to mid Luclin) - Usenet comment by Casey Webster:
"ZEM is 75 for most outdoor zones, 80 for some dungeons, 85 for some others, and it maxes out at 100 for underused zones and some especially difficult ones"
https://groups.google.com/g/alt.game...m/7h7DhR63ZmwJ


Another March 12 2002 (early to mid Luclin) - Usenet comment by Casey Webster:
"some zems reached 100 as a result of the patch (the hole is an example), but no zones exceeded it. The only zone at 100 before that patch was Kedge i believe."
Note that he was likely referring to the early Velious patch from a year prior.
https://groups.google.com/g/alt.game...m/m9N5Q__WctAJ


March 15 2002 (early to mid Luclin) - From EverQuest Third Anniversary State of the Game part 2 Absor post
Quote:
Originally Posted by Absor
Experience gained in dungeons designed for players level 30 and under will be increased. For many people, there is nothing more satisfying in EverQuest than adventuring in dungeons. By giving people the tools to help them succeed earlier, and rewards to match, we hope that more players will discover why many people feel this way.

March 19 2002 (early to mid Luclin) - Patch note:
Quote:
Originally Posted by Patch note
Raised the amount of experience gained in lower-level dungeons.
August 6 2003 patch note mentions a nerf for city zones around the time of PoP launching:
Quote:
Originally Posted by Patch Note
- Experience in city zones updated – Some time around the release of
Planes of Power we reduced the experience gain in some city zones. The
experience gained from quests in those zone has been increased.

June 11 2003 (about when the first guilds got into PoTime) - Sony nerfed PoP zone ZEMs when they increased the group experience bonuses:
Quote:
Originally Posted by Patch Note
Planes of Power zones had an increased experience award above and
beyond those of most other zones. Because of these new dramatic
improvements to group experience gains, we will be reducing the zone-
specific experience bonus for Planes of Power zones a bit. As long as
character is grouped with at least one other in these zones, they will
see an improvement in experience gain over the old system. The one down
side to the new system is that there will be a small decrease in
experience for those who do not group, and only for those who do so in
Planes of Power zones.
The new ZEM for most PoP zones was apparently 119 after that patch. This thread lists a lot of ZEMs days later: https://www.showeq.net/forums/showth...?3686-ZEM-list

Here's a 2004 post with more zones:
https://groups.google.com/g/alt.game...m/ZkseuNNwfaEJ
Reply With Quote
  #3  
Old 05-26-2024, 05:12 AM
Torven
Sarnak
 
Join Date: Aug 2014
Posts: 61
Default

Group Splits and Raid Exp

Originally experience was split based on the total experience of the players, not their levels. This was mentioned in the January 2001 Producer's Letter by Gordon Wrinn:

Quote:
Originally Posted by Gordon Wrinn
In regards to the sharing of the experience penalty, it was apparent in beta, before the penalty was shared, that those playing characters without an experience penalty leveled faster than those that did. It was obvious that this would occur, but it was to the extreme that a group of friends, all playing together, would become separated to the point that they could no longer group efficiently in the mid to upper-mid levels. So we chose to distribute experience in the group on the basis of the total experience of each member rather than the level, in order to keep groups together.
This meant that the entire group shared in the burdens of all race and class penalties. It also meant that higher level players would have taken a larger share of the pie than they do now, due to higher level players having a much larger experience sum. Sony would have likely added something to equalize splits at very low levels however, because otherwise the higher level player would end up getting virtually all of the experience; especially since players start at zero exp. Adding an amount to each player's sum (like 10k) would seem to be the easiest solution, but I have no idea what they did.

They changed it to a level-based division in early Velious:

Quote:
Originally Posted by Gordon Wrinn
This means that we must address the penalty differently: basically, for every kill, after all grouping bonuses and zone bonuses are applied, the experience will be split up according to level, rather than experience. For those classes that do not have a penalty, they will then be given that share. Those classes that have a penalty will get their share, multiplied by their experience penalty. Essentially we are creating extra experience to give to those with a penalty after everyone else has gotten their share.
Sony also added 5 to each group member level when doing the division. This was seemingly done so that players would end up with a similar amount of experience as at lower levels in particular the division would award too much exp to the higher level players otherwise. When I checked Darkpaw's servers I found that they still do this, as the math didn't work without adding 5 to levels.

The source for the +5 claim is Casey Webster who was one of the older ShowEQ developers. He discusses it at length in the "Pet Damage, exp. AGAIN!" thread here:
https://groups.google.com/g/alt.game...m/7h7DhR63ZmwJ

In another thread he mentions where he got it from, although it's still done on Darkpaw servers so proof isn't really needed:
https://groups.google.com/g/alt.game...m/tE0KxYd9rFAJ

It's commonly known that groups have an experience bonus. This bonus has changed several times over the years. Originally it was +2% for each group member after the first one, for a total of 10% at six members. This was changed to: 2, 6, 10, 14, 20% in early Velious at the same time they changed divisions to be level based and when they removed class penalties. From the Producer's Letter:

Quote:
Originally Posted by Gordon Wrinn
Our goal is to get people grouping earlier, and provide them enough of a bonus where they do not feel that they are losing ground during the learning process. We feel we can do this by doubling the grouping experience bonus and by scaling it up based on the size of the group. Currently, the bonus is an additional 2% experience per group member, not counting the first one, leading to a maximum bonus of 10%. Following the next patch, the bonus will be as follows:

2 person group - 2% total bonus.
3 person group - 6% total bonus.
4 person group - 10% total bonus.
5 person group - 14% total bonus.
6 person group - 20% total bonus.

This bonus is applied to the total experience reward for killing a creature prior to distributing it to the group.
In June 2003 they massively increased the group bonuses and removed the sixth member from counting in the division:

Quote:
Originally Posted by Patch Note
Prior to this update groups gained a 2% to 20% experience bonus for
having two to six members.

As of today this bonus has increased to range from 20% to 80% for
having two to five members. When a group adds a sixth member, the 80%
bonus remains, but the experience gained is only divided by 5 before
being distributed. The sixth group member no longer causes the
experience gain to be divided by 6.
Prathun (an ex-Darkpaw dev) posted this in March 2020 regarding Darkpaw's bonuses:

Quote:
Originally Posted by Prathun
Was:
1 characters - 1x
2 characters - 1.38x
3 characters - 1.68x
4 characters - 1.92x
5 characters - 2.16x
6 characters - 2.2x

Is now:
1 characters - 1x
2 characters - 1.58x
3 characters - 1.92x
4 characters - 2.22x
5 characters - 2.5x
6 characters - 2.56x
https://forums.daybreakgames.com/eq/.../#post-3885518

I ran tests in 2 and 3 man groups and I can verify that as of 2024 the latter values are still accurate. I don't know when the former values were adopted.

In Velious after Jan 2001, the exp you would have gotten from a kill in a group is this formula:
exp_from_kill = mob_level^2 * ZEM * class_mod * group_bonus * ((your_level+5)/(group_aggregate_level+(num_members*5)))

If a group member is more than 500 units away from the killed NPC's location then that group member is removed from the division. I measured this distance myself on Al'Kabor (late 2002 era) and on Darkpaw servers; it's still 500 units. If you're in a group with members inside the zone, and you're the only one in range of the kill then you'll get a "You gain experience" message instead of a "You gain party experience" message, indicating that you got undivided exp.

Sony did not check for out of range players until after the group bonus multiplier was applied to kill exp. This resulted in a bug/exploit where players could join a group then intentionally stay out of range of kills to allow those still in range to get the full group bonus on the experience even if all of the exp was awarded to a single player. This wasn't a big deal when the group bonus was only 10 or 20%, but after the June 2003 changes this allowed solo players to gain a very large experience bonus by abusing this flaw. Although when I did this trick in Timorous it didn't work, so there may have been two distance checks with the first check being very distant, which is odd. (I was about 7k units away from the group) I believe this exploit was fixed shortly after the June 2003 changes as I found a post mentioning it being nerfed at that time. Players for sure knew about the exploit back then and were discussing it openly.

Another exploit I discovered on Al'Kabor was that the characters that were outside of the 500 unit range still had their damage counted for kill credit. I abused this on Al'Kabor by splitting my group into two, having one half do damage to a train of NPCs using AoE spells including a powerlevelee in this first group, and doing >50% of the damage to the NPCs. Then I moved all the characters out of range except for the powerlevelee, and used a 2nd group to finish off the train. This would result in the powerlevelee getting all of the exp solo and with the group bonus multiplier. I have a video of this on Youtube. I used this character's log to estimate PoP level hell_mod values. This exploit no longer works on Darkpaw servers.

The Al'Kabor (Macintosh) server had another bug unique to that server: the group bonuses at 4+ members were way higher than intended because Hobart made errors when he tried to implement the June 2003 logic into the server code. He wrote a server patch note saying that he added +20% per member, meaning the June 2003 logic and not this weird super bonus at 4 members, so his intent is not speculation. Gnostica also says he spoke to Hobart personally and confirms his intent in the Zamiel thread. Utdaan makes a similar comment. Al'Kabor was kind of frankensteinish and had some newer code and data spliced into it when Hobart had some free time to add it. Combined with the exploit mentioned above, this bug allowed solo players to get 260% the normal experience value when grouping 4 out of range characters. (that's not including the 20% server bonus) See the Zamiel thread for details.

Raid experience was cut down to 60% of normal and without any sort of group bonus multiplier applied to it. Fester, a ShowEQ developer, says that in February 2003. I did a test on Darkpaw's servers and discovered that raid exp is handled differently now. The division is weighted by player levels and excludes any group bonus as expected but does not add 5 to the player levels and does not have any reduction. It makes some sense for them to have removed the -40% since group bonuses are gigantic now and that's enough to make raid exp much worse. It's plausible that old EQ may not have had the +5s either for raids however but I found no mention of it.

"Raid EXP is 60 % percent of EXP *AND* no group bonuses" (March 2003)
https://www.showeq.net/forums/showth...-50-blue/page2


Allowable Group Ranges; Con Effect On Member Exp

The lowest level a group member can be and still get experience in a group is 2/3rds the level of the highest member.

Of course how the math is done matters for this, so I went on Darkpaw's servers recently to check. I found that:

40 DOES get exp with a 61
39 did NOT get exp with a 60
39 DOES get exp with a 59

So knowing that, I tried to make the math fit. top_level * 10 / 15 in integer math seemed to fit. What Sony/Darkpaw did/does exactly I can't say.

There needs to be a minimum range for the very lowest levels however otherwise level 1s couldn't group with level 3s for example. What this should be I can't say with certainty. I think most likely it was 3, at least originally. The official player's guide says: "At lower levels, which are currently defined in EverQuest as Tenth or under, the level disparity is set to three". Several older Usenet posts mention three as well. Whether that means a level 1 may group with a level 4 or it instead means 1 and 3, I don't know. I have a PoP era log shows a level 10 grouping with a level 6, but it could have been expanded at some point before then.

What happens when the killed NPC is green to the top level member and dark blue to the other members? This is not entirely understood.

You might expect the NPC to grant zero exp to everybody in the group if the NPC is green to somebody in the group, but this is not the case.

Al'Kabor players claimed that not only did the lower level members get exp from the green-to-top-member NPC, but also that the exp was higher than expected. Some comments from the Zamiel thread:

Quote:
Originally Posted by Gnostica
If a mob is green to some members of a group, but light blue through to red for others, yet still within the 1.5x level range, then those people that get exp seem to get some sort of bonus. I'm guessing it's worked out much like the OOR bonus, but it might be some other mechanic. The peeps that are too high a level (they're killing green mobs) can kill like mad, but the younger toons get all the exp and don't have to share the split with any high level toons present. (We noticed that doing some shard camps, that low level toons would really rake in exp in some zone much more than others when in group. The SD shard camp is a great spot to test it.)
Quote:
Originally Posted by Placer
From my experience using Provo to mow down hoppers in DSP with a 43-46 PLed toon. The killing of a green appears to get your group members the exact same (60%) that they would get if they were grouped with another toon of the same size. Now if the mob is light blue, it is much worse for the 43-46.
Placer also said this in 2014 on the TAKP board:

Quote:
Originally Posted by Placer
AK would treat say a 65 and 43 grouped together on something green to the 65 as two 43s for the sake of dividing up the xp. I tested this lots in Velks and DSP when farming mats. The 43 would get 60% of solo xp (100+20% bonus / 2 players).

There is also a Usenet thread from April 2002 titled "Grouping and exp question", and they discuss this phenomenon:
https://groups.google.com/g/alt.game...m/808NbZZRqBgJ

Quote:
Originally Posted by Usenet Poster 1
However, from "my" experience when trying to Powerlevel someone without a druid (/em chuckle) If i have a character that is red to the low person but they still get experience, then they get all the experience from a blue mob that was green to the high character.
Quote:
Originally Posted by Usenet Poster 2
I would agree with this. when I used to hunt with my druid and warrior and druid was some 5 levels lower, his exp would move much faster if the Warrior wasn't getting an exp message, or at least it appeared that way.
Quote:
Originally Posted by Usenet Poster 3
I did just this in Mistmoore with a long-time friend just last week. I'm not sure if I was taking 56% of the experience or not; I do know that my Pally friend went through a large chunk of 46 in the 90 minutes we were there.
This however seems counter to Sony's no-free-rides ethos, so I suspect this was unintended.

At first I thought this might have been introduced later (maybe during Luclin) but doing more digging I found an older comment:

(September 1999)
"if a mob is green to the highest but blue or better to anyone else, usually the exp is divided ONLY between the people who can gain any exp from the mob"
https://groups.google.com/g/alt.game...m/zGqUcwcoRMwJ

Case closed? Nope. Other people dispute this: (I've found more than these)

(April 1999; note this is before group damage was aggregated for kill credit)
"If a member in a group is too high for the group, then the group becomes 2 groups for experience purposes (the player who is too high, and the rest of the group) and experience goes to whomever does the most damage. The group is still counted as one for looting purposes though."
https://groups.google.com/g/alt.game...m/tWUkWxLX9DMJ

"Also, while grouped - sometimes I got xp from blue things and sometimes I didn't." (April 1999)
https://groups.google.com/g/alt.game...m/qMeMFPORe-IJ

"High level characters take a huge share out of the experience even if the prey was green." (April 2000)
https://groups.google.com/g/alt.game...m/A71gx4EkTfAJ

This thread has several people saying the higher level still takes their large share from green-to-them kills:
https://groups.google.com/g/alt.game...m/Ci1e6cams2IJ

So now you can see why I put little faith in player comments. Make your own judgements.

It's worth pointing out that before January 2001, the higher levels in a group took a much larger share of the experience.

Another reason to believe that green powerleveling was unintended is that light blue exp was supposedly reduced for all members of a group, even if the mob was dark blue to some members. Placer says this in his Zamiel thread post and I've found a couple other low credibility posts mentioning this. The evidence for this is somewhat lacking however, although that is the behavior I would have expected Sony to put in. (but then I would have expected greens to grant nothing to everybody too) It makes no sense to reduce it for light blues but not greens.

As for Darkpaw's servers, this exploit/trick still works today. (or something very similar at least) In fact it was/is a problem for me because I don't want to level my characters as I use them to parse things, resulting in my needing to suicide them. The reason I had to suicide them was because the exp would end up in level exp even if I had AA exp on when I killed grays with my higher level characters, so this oddity also suggests that it may not be intended. The exp the lower level player in the group gets is the typical level+5 split (i.e. they do not get it all or any kind of bonus) as any other kill but is otherwise not reduced in any way. I tested on grays and greens. Greens had no extra reduction for the players the mob is not green to like light blues supposedly had in old EQ.
Reply With Quote
  #4  
Old 05-26-2024, 05:15 AM
Torven
Sarnak
 
Join Date: Aug 2014
Posts: 61
Default

Pets

Pets used to take experience if they did the majority of the damage to killed mobs. Mid-Luclin this was changed to pets taking experience only if they did 100% of the damage, but if they did they took even more exp.

I found two possible ways the pet exp taking worked prior to Luclin. There is evidence for both of these ways and this evidence contradicts. It is clear however that when solo, pets took 50% if they did more than 50% of the damage to the killed mob. The question is how it behaved in groups.

Absor seemingly explained how pet exp originally worked here:

(A Developer's Corner post I lost the Wayback link to but it's in my archive)
Topic: Update on pets and group experience
Absor Station Admin posted 03-11-2002 10:16 AM
Quote:
Originally Posted by Absor
I was just informed by Scott Hartsman that I made a mistake about the pet experience thing. Pets only take experience from a group if they outdamage the entire group. If they do, then they gain a share as if they were a group member. If they do not, they take no experience at all.

This is how it's been working the whole time, and it makes a lot more sense. I'm sorry for the mistake.
What "gain a share as if they were a group member" means isn't as clear as it sounds. Pretending like the pet is a player character group member in the math is problematic. Group splits used to be divided in proportion to each members' lifetime experience. That doesn't work well for pets. Also pets generally are much lower level than the caster, and solo owners saw 50% reductions presumably even with lower level pets. The caster's exp or level could have been used in place of the pet's however. Also it's just more difficult to code group splits in a way that makes pets an extra member, as hate lists and group member logic is otherwise very unrelated, which creates some doubt it was handled this way; Sony liked simple solutions to problems. Plus the pet penalty would be rather small in a large group if done this way, which doesn't make much sense if the pet is doing >50% of the damage and outdamaging the entire group.

Incidentally Absor's "mistake" appears to have been this comment from a week earlier:

Quote:
Originally Posted by Absor
A pet will take a fair share of the experience (as if it was an additional party member) IF the pet does more damage than all other group members (individually). If any party member outdamages the pet, the pet only takes 1% of the experience.
https://web.archive.org/web/20030203...ML/000511.html

2000 era Usenet posts mention that ShowEQ was used to confirm that half exp was taken if pets did >50% of the damage while solo at least:
https://groups.google.com/g/alt.game...m/fSNeb3GQi1YJ

In that Usenet thread a player mentions having ran several pet exp tests. (without ShowEQ) His solo tests resulted in 10 snake kills to reach level 2, or 20 kills if the pet did all the damage, as expected. 10 kills would indicate a ZEM of 114. His group test however resulted in 27 snake kills in a group of two necromancers (with only one pet out) to reach level 2 with the one pet having done all of the damage. Knowing the ZEM, we can estimate the experience share the pet took in his two person group, which is about 30%. That is less than expected. (not 50%, not 33%)

Another post in that thread asserts that pets took 50% in groups as well however, which disagrees with the snake guy. I sometimes have to judge the credibility of a claim based partially on how well it's written, and the snake guy doesn't inspire confidence in that regard, however his number is hard to write off particularly when the solo numbers fit so well. He also posted again with "I've seen them take as much as a 23% of the experience in a 2 person group" and that the ShowEQ math didn't work in groups with pets. Since this was pre-January 2001, splits were done from the lifetime sums of the group members instead of their levels, so it's possible that was the reason the split wasn't 33% somehow. Anyway, snake guy's claim is much closer to the 'group share' theory.

Another poster in a different thread two months later disputes the 'group share' theory:

"Pets do not take any experience unless they outdamage the entire group (including other pets in the group)... if pets outdamage the party, they take 50% experience, and the other 50% is divided normally amongst the group." (Nov 1, 2000)
https://groups.google.com/g/alt.game...m/cTn_rCz2rZAJ

The June 5th 2002 patch changed pet experience. The day before, Rich Waters (EQ Lead Designer at the time) posted a message on the Developer's Corner forum, which included this:

Quote:
Originally Posted by Rich Waters
Pets no longer take experience in most situations
In the past, pets would take a full share of experience if they did the most damage to a creature. We've changed this rule as follows -

* Pets take ZERO experience from a creature, unless no player does damage to that creature.

* Pets take 75% of the experience from a creature - if no player does damage to that creature.

This means that your pet takes no experience from you or your group unless your pet kills a monster with no help from players. As long as you or your groupmates do damage to a monster, your pet will take ZERO experience.
https://web.archive.org/web/20040910...w.jsp?id=51142

So after that point, players got full exp if they did a single point of damage. It's still that way on Darkpaw servers.

Confusingly his first sentence description of how pets used to work sounds like Absor's "mistake" comment.

That same patch however gave Dire Charm pets different experience logic. Waters said this in the same post:

Quote:
Originally Posted by Rich Waters
Due to the high potential power of dire charmed pets, they will continue to take a share of the experience. In the worst case, dire charmed pets will take a full share of experience - this is the same as it's always been.

The amount of experience taken by a dire charmed pet scales based on how much damage the pet does. If the dire charmed pet is doing the majority of the damage, it's experience share gets larger. If you or your group are doing reasonable damage to a monster, a dire charmed pet will not take much experience. The more damage you do, the less experience a dire charmed pet will take. In most situations, dire charmed pets will take less experience than they did previously.

As with all pets, if a dire charmed pet does the majority of damage, and no player does any damage, the dire charmed pet will take 75% of the experience from the kill.
Rich Waters then wrote another post to clarify things further.

Quote:
Originally Posted by Rich Waters
Pets and Experience –

As there's still some confusion about how experience works with the new pet enhancements, I'd like to talk about it a bit more so that everyone understands. Here are the basic rules, along with some examples -

Situation 1 - You are soloing and you have a summoned pet or a charmed pet (not including Dire Charm)-
* If you do any damage to the monster at all, you get all of the experience.
* If you don't do any damage to the monster, the pet gets 75% of the experience.

Situation 2 - You are in a group and you have a summoned pet or a charmed pet (not including Dire Charm)-
* If you or your group do any damage to the monster at all, your group gets all of the experience.
* If you or your group don't do any damage to the monster, the pet gets 75% of the experience.

Situation 3 - You are soloing and you have a Dire Charmed pet-
* If you do half of the damage to the monster or more, you get all of the experience.
* If you do less than half of the damage to the monster, the pet gets between 25% and 50% of the experience, depending on how much damage it did.
* If you don't do any damage to the monster, the pet gets 75% of the experience.

Situation 4 - You are in a group and you have a Dire Charmed pet-
* If you or your group do half of the damage to the monster or more, your group gets all of the experience.
* If you or your group do less than half of the damage to the monster, the pet gets between 25% and 50% of the experience, depending on how much damage it did.
* If you or your group don't do any damage to the monster, the pet gets 75% of the experience.


How is this different than the old way for non-Dire Charmed pets?

In the old scheme, if a pet did more than half of the damage to a monster, it took half the experience reward.
In the new scheme, pets take zero experience unless no player does damage. If no player does any damage, then the pet takes 75% of the experience reward

This means that it's much easier to make sure a pet doesn't take any experience from you or your group. In most situations where a pet would have taken half the experience before, it now takes no experience at all.

How is this different than the old way for Dire Charmed pets?

In the old scheme, if a Dire Charmed pet did more than half of the damage to a monster, it took half of the experience reward.
In the new scheme, if a Dire Charmed pet does more than half of the damage to a monster, it takes 25% - 50% of the experience reward, depending on damage done.

This means that things are about the same as they were for Dire Charmed pets before, except they take less experience than they used to in most cases. As long as your pet doesn't do a lot more damage than you, you'll get more experience than you used to. In the worst case (unless you don't do any damage at all), your pet will take half the experience just like it used to.

The one exception to this is - If you don't do any damage to the monster, the pet gets 75% of the experience.


What's the bottom line?

With the new rules, players get more experience than they used to with a pet in almost every situation. The only way you can get less than before is if you or your group don't damage a monster. As long as you do any damage at all, your pet will never take the 75% experience share. Make sure you do a bit of damage, and your summoned or regular charmed pets won't take a single point of experience.

For Dire charmed pets it's better also - If you do half the damage to a monster, the Dire Charmed pet takes no experience. If you do less than half, your pet takes as little as 25% of the experience when it used to always take half. It can still take up to half the experience if you don't do much damage, but if you're contributing to the damage you'll get more than you used to.

Rich Waters
Lead Designer, EverQuest
Sony Online Entertainment
http://web.archive.org/web/200206151...ML/000594.html

The patch note said this:

Quote:
Originally Posted by Patch Note
The exception to the above rule is Dire Charmed pets, which will take a share of the experience scaled by the amount of damage they do. In most situations (with an active party, for example) the Dire Charmed pet will not take experience. Like all pets, they will take 75% of the experience if no PC does damage to the target. As long as a player does damage to a creature, a Dire Charmed pet will never take more experience than it used to.
In that lengthy clarification post, Waters says flat out "if a pet did more than half of the damage to a monster, it took half the experience" followed by "from you or your group". Then goes on to say that dire charm pets worked the same way before the patch. He also calls this kind of experience taking a 'share' even though it's not splitting exp like a pseudo-member of the group.

So I would say that pets splitting exp like a pseudomember of the group was less likely than taking 50% before the group split is done but I'm far from certain of this.

As for dire charm exp scaling, it seems clear that the exp returned = exp_amount * (1.0 - (pet_dmg / total_dmg * 0.5))

I did a single test on Darkpaw servers and the dire charm penalty seems to be gone. I couldn't google a patch note mentioning it however.


Charm Exp Nerf

In Planes of Power charm became very strong and resulted in a meta where the only good exp was when using charm. Sony made some rather meager nerfs to charm about 5-6 months after PoP's launch on April 8 2003:

Quote:
Originally Posted by Patch Note
** Charm Changes **

- Healing a charmed pet now generates an appropriate amount of hate for the healer.
- Charmed pets now take up to one third of the experience for each NPC killed. This amount scales down based on the percentage of damage to the target that the pet does. Dire charm pets still take the same experience they always have.
- Charmed pets are no longer selected as a monster’s preferred target if there are many players available for the monster to attack instead.
- ALSO (we forgot to mention this before), the resist modifiers on several charm spells (such as Beckon, Call of the Arch Mage, Command of Druzil and Word of Terris) have been removed, making them a bit easier to resist.
Baelish from EQCastersRealm spoke to an unnamed developer right after the patch and got this clarification regarding the charm exp nerf:

Quote:
Charm Experience has been changed so that NPCs will take a more appropriate amount of experience per kill. If the NPC does 100% of the damage it will take 33% of the experience of the kill. This is the most that can ever be taken from the kill. If it does 50% it will take 16.5%, if it does 10% it will take 3.3% etc. It’s a sliding scale based on how much damage the Charmed NPC does to the monster killed.

This new system is independent of any grouping. The NPC will take Experience with this system whether the Enchanter is solo or not.
https://web.archive.org/web/20030516...628&FORUM_ID=2

That doesn't seem to need any further explanation and is probably this: exp returned = exp_amount * (1.0 - (pet_dmg / total_dmg / 3.0))

I don't know if Darkpaw servers still do this but since charm is heavily nerfed on Darkpaw servers I wouldn't be surprised if they reverted this exp nerf.

None of these explanations answer what happens when you have multiple pets. Was all pet damage added up into a single pet damage pool? Or if you had multiple pets did the penalties diminish because the penalties only took effect based on the highest damage pet?


Death Experience Loss

I was able to determine death exp with a high degree of precision. My sources for this are a client decompile provided to me by Kicnlag, the 2001 producer's letter and Darkpaw's Live servers. The following outlines my current understanding of how it worked in old EQ but this may omit some minor details:
  • Death exp was semi-normalized to that of a warrior of the same race as the player, such that earning it back took the same amount of kills for everybody of the same race.
  • Death exp starts off at 12.5% of the amount of exp in the previous level of what a warrior of the same race would lose, with modifications from there.
  • Death exp used to be 25% but was cut in half about nine weeks after the game's launch in 1999.
  • Death exp was capped at 6 million in old EQ. This results in the exp loss for levels 57 to 65 being the same amount numerically. (except level 61)
  • Death exp and the cap were both multiplied by the class_mod after January 2001, which was a necessary step in the class penalties removal.
  • Exp loss under level 25 was less than the full amount. It was linear scaling from 0% at level 0 to 100% at level 25.
  • No experience loss under level six. A Luclin patch made this level 11.
  • HBMs are most likely NOT factored into experience loss. More about this in a later section.

How Darkpaw servers differ:
  • All races and classes have the same experience gains and losses now. Everybody might as well be a Human Cleric as far as exp is concerned.
  • The linear scaling reduction under level 25 now starts from six instead of one.
  • Death exp is capped at 5.4 million instead of 6 million up to level 65. This results in 10% less exp loss for some levels. Unknown why it is reduced.
  • The first level of experience loss is six again.

All sources agree that exp loss was a percentage of the previous level's experience. The basics of death exp on Darkpaw's servers is also mostly the same, with the loss being exactly the same as it was in old EQ for most levels. This is how I can verify that my math is correct. For example, the exp loss at level 25 on Darkpaw servers is 10.351% (in-game only has 3 decimals of precision) and it's 10.3505% in my spreadsheet. Using death exp loss also allowed me to verify that Darkpaw's experience required for levels is still the same as well.

The percentage was 25% when EverQuest launched in March 1999. It was halved to 12.5% in the May 24th 1999 patch. Imagine how bad dying was near launch.

We can know that death exp was based on the Warrior's experience table for everybody because the January 2001 producer's letter says so. Also the Trilogy client logic has a 10% reduction that otherwise has no reason to be there.

Quote:
Originally Posted by Gordon Wrinn
There is a problem, however, with this 'new' formula. Death penalties are currently based off of the level before your current one. Secondly, everyone suffers the same numeric experience loss on death as anyone else of their race, regardless of class or class-based experience penalties. What this means is: if I am a cleric, and you are an SK of the same level and race, we both die and lose the same numeric value of experience (Example: 100,000 experience points). When we go back to recover from death, you as the SK will get your 100,000 points back faster than I will as a cleric, since all of the experience you get is multiplied by your class-penalty (1.4). Essentially, I lose and gain experience at 1.0, but you lose at 1.0, and gain at 1.4.

This is a balance issue we decided was also necessary to address. If we are going to make the statement that class experience penalties should not exist, we then have to do it on both ends (with exception to the two classes that we've decided to leave as-is). As such, rather than losing the same numeric value, loss on death will ALSO be multiplied by the experience penalty. Since everyone currently loses experience as if they are a warrior of their own race, we do not want anyone to lose more relative experience (e.g. experience such that recovery from death is more difficult). Hence, we further multiply the experience loss on death by the class experience modifier for warriors (0.9).

In our example above, my cleric would lose 90,000 XP on death at my level (Same as before since clerics do not have an XP penalty), but your SK will lose 126,000 XP (Same as before, plus something to offset the experience gain bonus). Death is, however, still easier to recover from for both classes since we create experience out of thin air for every kill.
In the January 2001 patch Sony eliminated class exp penalties by giving most classes bonus exp on kills, which was done by multiplying gained experience by the class_mod for these classes. The death exp also had to be increased otherwise hybrids and casters would end up with a much reduced penalty from death. The Trilogy client decompile shows us how this was done. For the most part Sony just stopped substituting the player's class with Warrior when getting the previous level's exp. Here's some pseudocode of the client logic:

Code:
if (level < 25) then
	levelFactor = level * 0.01
else
	levelFactor = 0.25
end

LastLevelExp = GetTotalExpForNextLevel(level - 2, class, race)
ThisLevelExp = GetTotalExpForNextLevel(level - 1, class, race)

xpLoss = (ThisLevelExp - LastLevelExp) * levelFactor / 2
maxLoss = 6000000

if (class != ROGUE and class != WARRIOR) then
	xpLoss = xpLoss * 9 / 10
	if (classMod > 10) then
		maxLoss = classMod / 10 * 6000000
	end
end

if (maxLoss < xpLoss) then
	xpLoss = maxLoss
end

if (xpLoss < experience) then
	experience = experience - xpLoss
else
	experience = 0
end
That should be easy to follow except for maybe the block that excludes Rogues and Warriors. Sony needed to reduce the exp loss for most classes there by 10% because Warriors had a class exp bonus in the form of 10% less exp required to level than the baseline. (baseline would be priest classes) And since all classes previously used Warrior exp to determine death exp, an adjustment was needed. With that 10% reduction the effective rate of refilling the exp meter ends up the same as before for all classes. Note that the client logic matched Mr. Wrinn's exp example of an SK losing 126,000 compared to a cleric or warrior's 90k, so that is verification of the client's logic matching the server. (90k * 1.4 = 126k)

The Trilogy client decompile is where I found the 6 million death exp cap. This cap fit the observations. It was commonly known that exp loss at level 60 was about 1 yellow bubble and exp loss at level 65 was about 5-6%. A 6 million exp cap fits this with the results being 18.83% and 5.43% for most classes. I have video capture of myself dying at level 65 and losing 6% on Al'Kabor. (as a Halfling) A 5.4m cap is too low to match observations and the Trilogy decompile did not look like the cap was reduced. Other than Darkpaw's servers using a 5.4m cap, the exp loss for levels 25-65 seems unchanged since 2001 but I can't rule out edge possible cases.

"At level 65 I lose about 5.5% of level upon death (and get most of it back when rezzed, of course)"
https://groups.google.com/g/alt.game...m/0Wu9sa5l-ckJ

"It's more like 16% to 20% per unrezzed death in 60, somewhere in the neighborhood of 4 to 5 blue bubbles."
https://groups.google.com/g/alt.game...m/SaSgilV2_xkJ

"Level 60 is a short level but a death is a 19% loss"
https://web.archive.org/web/20040707...x?PostID=47517

Because death exp was based on the previous level's experience, it resulted in some levels having double experience loss, as hell levels required twice the experience. In Absor's "EverQuest Third Anniversary State of the Game part 2" post from March 15 2002, he mentions the removal of the double experience loss in levels after hell levels:

Quote:
Originally Posted by Absor
the "post-hell-level" experience loss in levels 31, 36, 41, and 46 was driven from the face of Norrath as well.
Sony removed the double exp loss at the same time they smoothed out 30-50 hell levels. Sony just divided the exp loss in half for those levels, which I was able to verify on Darkpaw's servers. The evidence strongly suggests that Sony never multiplied death exp by the HBMs. See the hell level removal section for details.

Deaths in hell levels also showed an anomaly in that the visible experience loss in the UI meter was half of normal, but in actuality the experience loss was numerically typical and the visual difference was entirely from the level requiring twice the exp to complete, so making the exp back took just as long.

The 25% number along with the May 1999 halving are easily seen in the client logic. I did find an old Usenet post mentioning 12.5% so the ShowEQ guys seemed to have discovered this back then as well or a Sony rep mentioned it. After death exp was cut in half in May 1999 and accounting for the Warrior bonus, typical death exp loss (as visibly shown in the exp bar) for priests would have been about 10.5% for most levels 25-50, 19.5% to 21.8% for post-hell levels, and 5% in hell levels. For the first nine weeks or so of EQ in 1999, dying at level 46 as a priest class would have cost the player 43.5% of their experience bar. The reduced experience loss in levels under 25 also kind of explain why the resurrection spells weren't obtained until later levels before Luclin since experience loss is fairly small until level 20 or so.

Not included in the pseudocode is a level check to skip death exp if the player is under level six, but it was in the decompile. I found old posts going back to 1999 mentioning the level six threshold as well. They reverted the Luclin change to raise the first exp loss level to 11 back to six in 2006.

I found this post by Rashere (Sony dev) from 2006:

Quote:
Originally Posted by Rashere
When EQ first launched, XP loss started out low and ramped up smoothly from level 1 until level 25. At level 25, the full death penalty is in effect and from then on the amount of XP you lose on a death is constant.

When the change was made to make players not lose any XP up to level 10, that scaling wasn't taken into account. So, when you started losing XP at level 11, you were starting out with 11 levels of scaling already in place. This made you go from not losing any XP to taking a sizeable hit immediately.

After the changes we just made, the scaling has been adjusted to start at level 6 and ramp up from there. So when you first start to lose XP at level 6, it will be just a small amount and will smoothly increase from there until level 25 when the full effect is in place.

Basically, we just smoothed things out a bit so they're not so jarring when you first start to lose XP. The full death penalty is unchanged and still kicks in at level 25.
https://web.archive.org/web/20060609...e=1&format=all
Reply With Quote
  #5  
Old 05-26-2024, 05:17 AM
Torven
Sarnak
 
Join Date: Aug 2014
Posts: 61
Default

Quest Exp

Quest experience is simple since it is merely a constant value per quest (but with a cap) and is sent over the wire to the client in packets. Better still, the recent addition on Darkpaw servers to display precise experience gain percentages allows us to obtain a quest's experience value rather easily without packing sniffing, which allows non-technical people to collect this information.

Robregen packet sniffed quest exp for some quests on Darkpaw servers back in TAKP's early years. This was very helpful not just to make quests accurate but also to verify that I could deduce quest exp from using the detailed experience gain messages that were recently added. By assuming that level exp is the same now as it was in 1999 -- accounting for the changes to penalties and bonuses and assuming everybody's exp is now like an old Human Cleric -- one can do simple math to find the quest exp value. An example:

Canloe Nusback in Kaladim is the Crushbone Belt turn-in NPC. On a level 4 Human Cleric played on the Vox server I got 12.432% from completing the quest. A Human Cleric (again the race/class does not matter anymore on Darkpaw servers) needs 37,000 experience to reach level 5 from the start of the level. 37000 * 0.12432 = 4600 exp granted from the quest. Robregen obtained that same value when he sniffed it from packets.

There are two caveats to this method however that must not be ignored. The first is that there is a cap on quest exp and this cap is 14.284%, which is one seventh the amount of exp to complete the level. (old EQ's quest exp cap was 25%)

The second consideration to be aware of is that Darkpaw added some kind of bonus exp to quests starting at level 8, which grows 1% per level for 5 levels then grows 0.5% per level up to around level 21 where it stays at 10%. The highest level character I used for this was level 30 so I cannot say how this might change after that. This bonus is added AFTER the quest exp is capped. (it also applies to kill exp) Why they added this odd little bonus I don't know, but I've done dozens of quests to know that it's there. It must be accounted for when computing the quest's experience value. To be very clear: this bonus did NOT exist in old EQ and is some newer addition. ('new' being relative; it could be 17 years old for all I know)

For example, the Captain Hazran Deathfist Belt quest granted 0.422% at level 29, which is 10,284 exp. The quest granted 10.275% at level 6 which is 9,350 exp. The actual experience gain from the quest is therefore 9,350.

My experience spreadsheet lists the experience rewards for 50 quests. This includes the usual experience farming newbie quests like bone chips and orc belts. Sony did nerf the exp for some of the quests however, so I had to do some detective work to try and estimate what the quests granted pre-nerf. I also have a large quest exp research and notes file. Links to both are listed below in the 'sources' section.

There's too much data to include in this write-up so I'll just discuss some of the major findings. It's worth mentioning however that I only found evidence for a few quests being nerfed (reduced exp) and the vast majority seemed unnerfed. Also it is difficult to impossible to estimate quest exp based on old user comments, so it's mostly a handful of 'this might have been nerfed'.

One major find was an Allakhazam comment under the Gunlok Jure NPC which listed in detail how many turn-ins were required to level up to level 16. That NPC is the Kaladim bone chip quest giver. This quest was notorious for granting too much experience. It was heavily nerfed on Darkpaw servers years ago, and maybe more than once. You can buy infinite bone chips on Darkpaw servers now, and hand in quest items stacked, so naturally they had to. The nerfs seemed to have occurred after 2004. The detailed Allakhazam comment can be found at this link: https://everquest.allakhazam.com/db/...04956364492208

Right away from looking at his Gunlok Jure data you can clearly see the old quest exp cap was 25%. The quest exp seemed to be 4600, but I couldn't make the data fit expectations very well for some reason so there is uncertainty to that estimate but if it's wrong it's not very far off. 4600 is a stupid amount of exp for 4 bone chips. On Darkpaw servers this quest grants 250 exp.

I had found log chatter of a guy on Al'Kabor saying his Hero bracer quest turn-in granted 25% at level 1, indicative of the cap.

Captain Tillin gnoll fangs grant 4500 exp on Darkpaw servers, but older comments suggest that it was higher, about 7-8k. I found seven comments agreeing with 7-8k and only one with a much lower estimate. The outlier could have easily been a guy getting less exp because he was grouped. The Halas quest awards 8000 exp and it only had one old comment under it which indicated that it was also nerfed and awarded more in the past.

The goblin ears quest in Highkeep did not award experience or it was virtually zero. Two players in the Zamiel thread (Al'Kabor players) said they had given ears to level 1 characters to powerlevel them and the quest awarded nothing, despite giving an experience message. Allakhazam is full of comments saying the same thing, although one said it did but it's very outnumbered. On Darkpaw servers it currently grants 25 exp. Similarly, the orc scalps in Highpass granted no exp but on Darkpaw's servers now give 25. I also noticed that at some point they changed the coin reward for these.

Moonstones, both greater and non-greater, award 10k exp on Darkpaw servers. Comments suggest that this is unchanged, or at best it was only slightly more.

The Crushbone shoulderpads quest in Kaladim awards 29400 exp. This is quite a bit and old comments seem to agree with that value suggesting that it's unchanged.


Alternate Advancement Points

AAs were only 15 million exp, however Sony reduced the experience that went into AAs by 20%, so the effective exp for an AA was 18,750,000. The -20% is a bit strange and I can't tell you why they just didn't have a higher amount without a reduction. I don't know if AAs are the same on Darkpaw servers.

I found a few sources which made these claims, and many posts mention 15 million. The 15m-20%/18.75m number also fit when I tested it with observations and math using kill data from logs and player comments.

"updated alt exp slightly, it appears 15,000,000 is the exact value per ability point" -- Fee, ShowEQ developer (February 2002)
https://www.showeq.net/forums/showth...22-2002&p=3048

"the raw exp value is still sent. According to the everquest.h in CVS the aaxp has always been sent in the charInfoStruct as a value from 0 to 15,000,000" (June 2003)
https://www.showeq.net/forums/showth...st-Patch/page3

"don't forget, if your using aa-exp value changes, that theres a 20% penalty to aa exp" (March 2002)
https://www.showeq.net/forums/showth...e-XP-Modifiers

"I was never able to figure out why my XP wasn't what I thought it should be, until I saw that AA is penalized 20%, and then it worked out perfectly" (February 2003)
https://www.showeq.net/forums/showth...ed-for-50-blue

"I have found that AAXP IS nerfed by 20%. I.E. if you figure out how much XP a mob should give, then multiply by .8, that is how much AAXP you get. And, yes, I have AA set to 100%." (February 2003)
https://www.showeq.net/forums/showth...-50-blue/page2

You can see my adding up kills and calculating experience estimates from them in my spreadsheet. I need to point out that there was a PBAoE reduction on Al'Kabor and that's why the math doesn't work for many of the tallies. I go into detail about this in a later section.

There was a common assumption/claim many years back that one AA point was about as much exp as level 51 took to complete. This is actually true only if you account for the hell level balance multiplier, which granted about 30% more exp for kills in the level, effectively reducing the amount required. In that case the numbers actually come out quite close. However that claim predates the level 51-59 smoothing, which occurred in late Luclin, so it appears to be a coincidence. If there were ever a dev quote making that level 51 claim I've not found it. I did find some player comments refuting the claim before the level smoothing, with them saying AAs went up faster.

AA exp is not modified by race or class penalties or bonuses. But there was an exception to that rule. Zamiel says that there was an old bug which was if your AA exp division was not 100% and it was split between level and AAs, then the racial modifiers did apply. I did not verify this but I found a few old pages that mentioned it, albeit not very strong claims. That it was even mentioned at all is evidence. I do not know when this was fixed, but it was active on Al'Kabor. (a circa November 2002 server) Hence any race with a penalty would have been ill advised to be between 100% and 0% AA exp. Zamiel claimed that Halflings however could gain bonus AA exp (slightly less than 4%) by setting their AA exp to 99% using the slash command. This bug seems strange to me because of the way exp worked: penalties were manifested by increasing the experience required to level, not by reducing the experience gained. So if it existed (seems likely) then Sony did some weird math.

Sony actually forgot to account for the hybrid penalty removal when distributing experience to AAs when Luclin launched. They fixed this mistake one week later.

This same patch also made it harder to gain AAs. Perhaps this is when they applied the -20% to AA exp. Either that or they raised the amount required to 15m then. Perhaps it was just easier for them to reduce gains by 20% than change the total amount of exp required?

Quote:
Originally Posted by Patch Note
- Fixed a bug with the “Alternate Advancement” system. An artifact from
the removal of the class-based experience penalty several months ago
resulted in hybrids gaining AAXP much faster than they should, and
warriors and rogues gaining it slower. AA-point cost was designed to be
the same regardless of class.

- Increased the experience cost per AA-point overall after analysis of
point accruals accross the game.

Hell Level Removal (HBM)

In Luclin Sony effectively removed hell levels using some quote "mathemagics". The solution they used was to do for hell levels what they did for class penalty removal, which is apply multipliers to experience gains for levels 30-60. In my experience spreadsheet I call this "HBM" which is short for hell (level) balance multiplier which is a term I borrowed from some old ShowEQ threads.

In order to try and figure out what these were, I added up the kills I needed to level on Al'Kabor and estimated the experience gains at each level. I mostly used my solo druid logs for this, so party splits were not a factor. The estimates I came up with for the HBMs were:

30: 1.65
35: 1.6
40, 45: 1.5
31-34, 36-39, 41-44, 46-49: 0.9
50: 0.75
51: 1.3
52: 1.35
53: 1.05
54: 1.15
55: 0.9
56: 0.95
57: 0.95
58: 0.835
59: 1.1 or 1.2
60: 0.75

These are estimates with a margin of error. Some levels are more accurate than others due to the mobs killed in my log. Levels 56 and 57 however I can confirm were 0.95 as old ShowEQ threads (http://www.showeq.net/forums/showthr...-6-patch/page5) had packet sniffed experience data. My estimates for these levels ended up 0.952 and 0.953, so this is one way to confirm that my methods for computing exp are correct.

I'm extremely confident in level 60's 0.75 HBM in old EQ because the old ShowEQ threads showed this and my AK data shows it with no room for doubt. Darkpaw's servers seem to have a 0.5 HBM for level 60 and probably a significantly higher HBM than 1.1 for 59 although it's hard for me to be sure since I don't know the current exp gain formula. In the 2002 ShowEQ threads, Scott Hartsman says the HBM at 59 was 1.2 although not in a very assertive manner. ("something to the effect of 20%" he said) My estimate for AK's level 59 HBM was 1.1 but this could be wrong, however 1.2 makes the exp required for 59 slightly less than 58 if 0.835 is accurate so I question if 1.2 is correct. Of course the 58 estimate is likely not precise either.

I have discovered what appears to be a bug/exploit with how HBMs worked on Al'Kabor while researching for this document. It looks as though the HBMs were not applied on Al'Kabor if the exp was split between AAs and levels. I have ironclad evidence for the level 60 HBM being 0.75 on Al'Kabor (video and logs) but one of my videos shows my bard gaining exp from kills in level 60 as if the HBM reduction were not working. My bard had AA exp set to 80% and level exp set to 20% which I could deduce from the rate of AA exp gain. The level gain was going up at the rate expected if the HBM were not there. In another video I have a level 60 with all exp going into the level and it precisely showed the 25% reduction. So knowing this, one could have theoretically set their AA exp gain at 1% in levels with a HBM less than 1.0 and they would have gotten significantly more exp in those levels. Likewise they'd have lost exp in levels with an HBM greater than one doing that. This would make some sense since split exp was also factoring in race penalties to AAs erroneously so Sony must have had two code paths for exp and they neglected things in the path that handled exp when split.

The most notable Sony posts about hell levels and their removal were in the "EverQuest Third Anniversary State of the Game part 2" thread by Absor which was posted on Sony's offical forum. That thread no longer exists, and the Wayback copy of it also no longer exists. (Wayback has been losing pages) I did however save the Wayback copy of it to my EQ archive before it was lost.

For convenience I'll provide another Wayback link to a website that copied the Absor post: https://web.archive.org/web/20020515...cID=1547.topic

For people seriously interested in this topic, I recommend getting the copy from my archive however since it includes more commentary. I will quote most of it next however in case the links break.

Quote:
Originally Posted by Absor 03-15-2002
After applying advanced mathemagics, the commonly known "hell-level" experience gain effect in levels 30, 35, 40, and 45 has been largely done away with. Further, the "post-hell-level" experience loss in levels 31, 36, 41, and 46 was driven from the face of Norrath as well.
Scott Hartsman's response was also in the thread:

Quote:
Originally Posted by Scott Hartsman
We don't plan on releasing detailed mechanics - Some things people should really discover for themselves as part of normal gameplay. The goal was to smooth the path from 30 to 50, not to make it require less overall gameplay, or happen faster. The intent wasn't to get people from 30 to 50 any faster.

The overall path isn't any easier to reach level 50 -- a retroactive experience boost wouldn't really be called for. Down the line of what I said at the last fan faire, "If we did something with hell levels, we would have to find a fair solution that didn't affect anyone's existing experience or progress."

We found that solution.

Times change, people change, opinions change. The overriding opinion on the team is that hell levels pre-50 really aren't something that we like having in the game. This is also consistent with at least a handful of statements from the past: "If we could fix it, we really would." Those levels were a large source of frustration that really didn't have any context in the game world.
Hell levels were smoothed in two different patches. The first patch smoothed out classic levels, and the second patch five months later smoothed Kunark levels. The classic level smoothing was mentioned in the March 19 2002 patch note:

Quote:
Originally Posted by Patch Note
The experience curve from levels 30 to 50 has been smoothed.
A ShowEQ user posted that he noticed classic hell levels removed on March 18th 2002: (maybe a Test server player? or patch note date is off by a day)
"Pre-50 hell levels = gone.
I've seen it with my own characters... on a side note, the level after hell (ie: 31, 36, 41, 46) seems to be slower than previous. My thought is that they probably evened out the curve quite a bit."

https://www.showeq.net/forums/showth...?675-EXP-Chart

Quote:
Originally Posted by Patch note September 4 2002
- We have smoothed out level progression from 50-60. This should
mitigate the “penalty effect” that occurs in levels 51, 54 and 59. Note
that it will cost the same experience to level from 50 to 60 as it did
before. Also, death experience loss will appear different in some
levels. This is an unfortunate side effect that must remain to prevent
some experience exploits.
I can say that they did cut the post-hell exp loss (for levels 31, 36, 41, 46) exactly in half as it is that way on Darkpaw servers.

I would have expected Sony to have multiplied death exp by he HBMs however but this is NOT the case on Darkpaw servers as of 2024. The patch note sounds clear that death exp was modified in some way, and the obvious thing to do here is to multiply it by the HBMs. That will semi-normalize death exp in most classic levels to around 9.5% for non-hell and around 7.5-8.5% for hell levels, which you would want to do since the levels were smoothed out. Failing to do this will result in hell level deaths being greatly reduced in severity, deaths in non-hell levels being about 10% harsher, and deaths in levels 50 and 60 being 25% harsher.

There are consequences to multiplying death exp by the HBMs however. It worsens an exploit where players could store exp in corpses then resurrect them after leveling such that they would benefit from the hell level HBM multiplier outside of the hell level. Also if you die immediately after leveling in one of the classic hell levels then you'll lose nearly twice as much exp. This may be why Sony opted not to do this or removed it later. More logic could be added to prevent these issues however.

I found two player comments which indicated that Sony didn't multiply death exp by HBMs for the reasons I mentioned, but they didn't provide a source so I wouldn't consider this super credible:

"Also of note is that when the Hell Levels were "smoothed out", they did not adjust the death penalties. They stated that they felt that if they mucked with it, it would introduce exploits where getting rezed would grant experience above what was lost. So you still lose more or less exeperience based on whether the level used to be a hell level or not." (Dec 8, 2002)
https://groups.google.com/g/alt.game...m/MuH-7xCfD4cJ

"My recollection was they specifically stated death experience wasn't touched with the adjustment formula, just like any bonuses in place weren't touched." (Aug 15, 2003)
https://groups.google.com/g/alt.game...m/bKmrC3b-FBcJ

I did go looking for player comments in logs which mentioned how much experience they lost from death. The results from this were highly inconclusive. Some comments matched no HBM multiplication, and some did, and some didn't fit either case. Some contradicted each other. People are frustrating unreliable. One comment stood out though as being extra significant:

Quote:
Originally Posted by Usenet Poster
"I thought that when you died you always lost the same amount of exp. 10%.
It seems like this is about right except last night I died in Crystal
Caverns when I was mobbed by spiders, but I only lost 5%. I know exactly
how much it was because I have a custom UI that gives the amount of exp in
percentages and I had 12% before I died and 7% when I came to at my bind
point. Are there some zones that give you less exp penalty? or was it
because I was killed by greens? Or maybe the level I died at? Do you lose
only 5% when you die after 50?"
https://groups.google.com/g/alt.game...m/_Pz9Tt9wLykJ (Dec 7, 2002)

That guy was level 51. The expected exp loss at level 51 is 4.829%, or 6.28% if you multiply the HBM. This is more credible than the other comments because he mentions his before and after exp amounts using a UI that shows percentiles, and it matched the expected exp loss without HBM multiplication.

Kicnlag found this old Allakhazam comment regarding sacrifices:

"On Al'kabor you can only sacrifice level 60 toons to 46th. It took me 77 deaths to go from all but dinging 66th to 60th. (The exp bar was capped. We're capped at 65th on Al'kabor) I then sacrificed myself 138 times to reach level 45"
(source: https://everquest.allakhazam.com/db/...55409398759577)

The post does not mention the class he was. His death numbers fit rather well for a warrior or rogue if you don't multiply the HBMs. My expectations for a warrior would be about 140 deaths to go from 60.99 to 45.99, and about 81 deaths to go from 65.99 to 60.99. If however he was not one of those two classes then my estimates would be 156 and 90. Multiplying the HBMs makes the 45-60 number fit worse.

So the evidence is strongly in favor of not multiplying death exp by the HBMs. It's possible they did something to modify how death worked in edge cases after the smoothing but I've found no evidence of it, other than perhaps one case in an AK log where the experience loss was apparently much greater (he says 19%) than expected when it caused a player to lose level 55. I should mention that when I checked loss on Darkpaw servers that I noted exp loss on death but did not resurrect any corpses nor paid attention to delevels.
Reply With Quote
  #6  
Old 05-26-2024, 05:20 AM
Torven
Sarnak
 
Join Date: Aug 2014
Posts: 61
Default

High Level Kill Bonus Experience (MLM)

Near the end of Luclin, Sony added a very large experience bonus to higher level kills. This radically changed the meta for efficient leveling after level 50. The ShowEQ guys called it the "Mob Level Multiplier" in the old threads, so I refer to it as the MLM for short.

Sony implemented this in two patches which were two days apart. The first was on September 4 2002:

Quote:
Originally Posted by Patch Note
- We have rebalanced the rewards for killing NPCs for characters over
level 50. NPCs that are considerably weaker than you will give you less
experience. NPCs give more experience the higher level they are, and
for the best results you should fight the most challenging monsters you
can reasonably manage.
This patch also included the hell level smoothing for Kunark levels, which cut the experience gains at level 60 in half. It also reduced the experience for killing low blue cons, which angered the playerbase considerably, causing Sony to revert the blue con nerf two days later, on September 6th:

Quote:
Originally Posted by Patch Note
** Experience Balancing **

The combination of the NPC experience changes and the post-50 "penalty
level" rebalancing at once had a stronger effect than we anticipated.


We've altered a few of the changes to more accurately reflect our intent.

- The intent is to promote grouping, not to inflict a harsh nerf on our more
casual players who frequently do not have the time to invest in finding and
forming a group.

- However, we do not believe that going solo should be the single most
optimal way to gain experience in EverQuest, which is what it had become
prior to these changes.


That said, we'd like to take a moment and explain what happened.


On "Penalty Level" Rebalancing (Levels 51, 54, and 59):

- Many people saw much smaller experience gains at level 60. This was due to
the "penalty level" rebalancing effect more than anything that was done to
NPC experience values.

- In the previous patch, Level 60 became a full level, when it had been half
a half level the day before.

- "Penalty levels" have once again been rebalanced such that level 60 is
much closer to the half level that it was before, and experience was instead
redistributed throughout the earlier levels.

- The penalty effect is still mostly mitigated, however the distribution
should be more fair to those who are already level 60, yet do not have a
full bar in that level.


On Grouping and Soloing:

- From the point of view of a level 60 character, a steady stream of level
45 NPCs represented the single best advancement in EverQuest, when the score
is kept in "Experience Per Hour."

- With the advent of Luclin Alternate Abilities, nearly every sufficiently
advanced level 60 could eventually and easily solo these "optimal" NPCs, and
benefit more than they could in any type of grouping.

- We understand that forming a group, travelling to your destination, and
other inconveniences have an effect on progression.

- With that in mind, we've undone the penalties for killing low blue NPCs,
and instead added an additional group bonus to those who fight NPCs that are
reasonably close to the group's level.

- The idea is that people should be willing to take risks, fighting the
hardest NPCs that they possibly can, to reap the best rewards in EverQuest,
as opposed to the blue NPCs that represent the lowest risk.

- For both groups and soloists, we strongly recommend trying out harder
overall zones and the more difficult areas of zones that you might already
be used to.

- We're confident that you're going to find that the additional risk pays
off *extremely* well.
This second patch set the HBM for level 60 to 0.75. It was 0.5 for two days.

This patch note calls the high level kill bonus a "group bonus" which is a bit odd as the MLM applies even when solo. The group bonus at this point in time was still the Velious era one that went up to 20% with six members, and the June 2003 patch note says as much. Since the MLM would only kick-in on mostly unsoloable mobs near the player's level (at least in Luclin), perhaps that is why they worded the patch note that way. Notes are often written by people who didn't do the actual code and this also frequently causes some miscommunication. In another bullet point it also says "for both groups and soloists", so even this patch note contradicts itself.

The MLM Sony added was actually quite massive, resulting in upwards of three times as much exp for kills. It needed to be large because of how pointless it used to be killing level 55+ NPCs. This is one of the few changes to the game since Luclin launched that I feel was good for the game, although I might have implemented it a little differently.

The MLM worked like this:
  • It only applied if the player was level 51 or higher.
  • It only applied if the killed NPC was within 5 levels of the player. (i.e. on level 55 mobs but not 54 at level 60)
  • This bonus multiplier scaled significantly with player level from 51 to 60 and was fairly small for players near level 51.
  • At level 60+ the multiplier is 2.6 for a white con. So killing a white con granted +160% more exp than before.
  • It reduced or increased depending on the level difference of the player to the mob, resulting in up to 3x (+200%) exp on red cons at level 60+.
  • Checks were done to prevent grouping players well under the group average level getting this bonus.
My sources for this are the ShowEQ forums and exp data from observations/logs. Zamiel's thread in particular had some good observations. Soloers in PoP would get something like 8% of an AA or more for a single kill which allowed for these observations to be meaningful.

This thread in particular has the bonus scaling from level difference:

"Level_Modifier = (260 - 13*(level - mob_level))/100" (December 2002)
https://www.showeq.net/forums/showth...ed-for-50-blue

At level 60+, for each level difference, the bonus grows or shrinks by 0.13. So the multiplier for a level 55 mob killed by a level 60 player would be 1.95, which is nearly double the exp gain. I believe it may have capped at 3.0 because Zamiel's thread mentions this: "It is known that there is no marginal benefit for killing a mob over 3 levels higher than you"

Fester, a ShowEQ dev, mentions some restrictions on how the MLM is selectively applied in October 2003:

Quote:
Originally Posted by Fester
Cases where you don't get bonus:

1) group with a group where the average level is greater than 5 levels above you.
2) killing mobs greater than 5 levels above you.
3) killing mobs greater than 5 levels below you.

This really only harms power leveling with a low level player in a high level zone.

There also seems to be a few other odd cases and I suspect this is one:
1) group with someone >5 level above you and they do all or most of the damage.
2) group where the average is less than 5 levels above you but the highest member is more than 5 above you.

In short, I havent found any way to kill with a higher level and get bonus on the lower EXCEPT killing ungrouped with the lower level and use higher levels for healing, buffing, nuke down from 20% etc.
https://www.showeq.net/forums/showth...-50-blue/page2

He was responding to a poster that claimed his two person group of a level 62 and a level 51 resulted in no bonus exp for the level 51 while killing level 55 mobs. Sony apparently had some strict checks in place to prevent higher level players from powerleveling lower level players. I doubt these checks involved damage or hate lists however as this would punish grouped healers. It's possible, even likely, that he actually WAS getting the MLM on the level 51 however, but the MLM is so low at level 51 that he didn't notice and assumed it wasn't there. The MLM scales with player level from level 51 to 60.

Zamiel's thread has data from soloed NPCs at various levels, and this is where I got the scaling from. The thread discusses the best level to grind AAs extensively. The PoNightmare gargoyle data in particular is noteworthy:

Quote:
Originally Posted by Zamiel
while soloing gargs in PoN (who are level 57):

24 kills per AA @ level 52
21 kills per AA @ level 53
19 kills per AA @ level 54
18 kills per AA @ level 55
17 kills per AA @ level 56
16 kills per AA @ level 57 (with 2% AA left over)
16 kills per AA @ level 58 (with 5% AA left over)
16 kills per AA @ level 59 (with 6% AA left over)
15 kills per AA @ level 60
16 kills per AA @ level 61 (with 0% AA left over)

in this case the optimal level to kill level 57 mobs is actually level 60 (a +3 differential), which seems counter-intuitive.
Zamiel and the Al'Kabor folks didn't know about the MLM. They just noticed some oddities when trying to figure out the most optimal level to grind AAs in. Here Zamiel notices that the best level to kill gargoyles for AAs is actually three levels above the mob's level, which is very unexpected. The explanation for this is that the MLM scales by player level and is smaller the lower under level 60 you are.

The MLM is 2.6 for white cons at level 60 through 65. It doesn't increase after 60. Since the MLM is a multiplier, another way to read it would be +160% more exp. Zamiel's thread also mentions that killing white cons in Plane of Nightmare at level 59 resulted in an MLM of about 2.43, which is 0.17 less than 2.6, and this is within margin of error of 0.16. The patch note said that the bonus applies to characters over level 50, so it fits that the base MLM for whites would scale linearly by 0.16, or +16%, for every level after level 50 up to 2.6 at 60. This is why arguably the best level to grind AAs in was level 60.

Working on the assumption that the MLM is +0.16 for each level above 50 for white cons, the next task is to try and figure out the scaling for blues and yellows/reds. The gargoyle data gives us the final MLM values for those levels, so it's just a matter of finding for X. Here too I also simply tried adding 0.013 for each level above 50 in a linear fashion, and this worked on the blue kills but not the yellows and reds. I found that if I divided this in half then it fit the yellows and reds fairly well. However for kills at or after level 60 the scaling should not be divided in half.

The final model I came up with for the MLM is this:

Code:
level_diff = mob_level - player_level

if ( player_level >= 60 ) then
	mlm = 2.6 + 0.13 * level_diff
	if ( mlm > 3.0 ) then
		mlm = 3.0
	end
else if ( player_level > 50 ) then
	base = 1.0 + (player_level - 50) * 0.16
	scale = 0.013 * (player_level - 50)
	if ( level_diff > 0 ) then
		scale = scale / 2
	end
	mlm = base + scale * level_diff
end
If you read Zamiel's thread for exp data, it's important to understand how I calculated the experience. Al'Kabor had a 20% server bonus, and players there often abused the out of range group bonus bug, so to calculate experience on AK with 4 members in the group out of range, you would do this:

exp = mob_level * mob_level * ZEM * 1.2 * 2.6 * MLM

PoNightmare's ZEM was 145 and the 2.6 is the bugged group bonus on Al'Kabor with 5 or 6 members. (it also being 2.6 is seemingly a coincidence) I used 18,750,000 for an AA point. Also note that the sixth man on AK was NOT free when dividing among the group.

When Sony changed the con colors in Serpent's Spine they (I think) made dark blue 5 levels wide for players above level 50. The MLM being +/-5 levels may be the reason they did that.

I had found two ShowEQ threads which discussed the MLM right after the patches which have explicit exp data, so I was excited to find them:
https://web.archive.org/web/20021014...&threadid=1870
https://www.showeq.net/forums/showth...ember-6-patch/

The threads show that the MLM was calculated differently in the first patch and then changed two days later to what I outlined. At first it scaled linearly from the lowest dark blue up to (presumably) red cons, including a nerf for the low dark blues, but at two different scale rates. The big bonus level range looks like it was +/-6 instead of 5 levels wide, and the levels under that got a bonus or penalty scaled by 10% instead of 13%, such that a level 50 NPC killed by a level 60 player granted the same exp as before and a level 45 NPC granted half the exp. This is academic since I doubt any server would want to implement this but it's interesting from a historical standpoint I suppose.

In these threads you can also see several HBMs: level 60's HBM went from 0.5 to 0.75. Level 56 and 57 were 0.95. Scott Hartsman chimes in and mentions that level 59's HBM was 1.2, but that was not what I calculated on Al'Kabor, where it was 1.1 but my estimate could be incorrect.

Also, importantly, one guy calculates MLMs for a level 56 character killing level 51-53 mobs in the second thread, and his numbers match my model's results precisely.

I cannot say how the MLM changed after Planes of Power and the above should be assumed to apply to PoP era EQ only. I can say however that on 2024 Darkpaw servers, the exp scaling is very different and looks more like the September 4 2002 patch scaling as it's very gradual. Exp gains on Darkpaw servers seem radically changed and I have little understanding of how it works.


Individual NPC Multipliers

Some NPCs in Planes of Power award more or less exp when killed than would normally be awarded by the standard formula. This is easy to see from the amount of movement in the experience bar compared to other NPCs of the same level with the same ZEM. Sony seems to have added some kind of individual experience modifier to NPC types in PoP. I'm unaware of any cases in prior eras, so I think it's probably a PoP addition but I can't be sure.

The level 68 mobs in Plane of Fire had a large exp bonus applied to them. In the Zamiel thread, Al'Kabor players mentioned these NPCs and how much exp they got from them, which was 21% of an AA solo or 55% solo with 4 OOR in group. The standard exp for these would otherwise be 13.273% on Al'Kabor.

68 * 68 * 150 * 1.2 * 2.99 = 2,488,637 / 18750000 = 13.273% (remember AK had a 20% server bonus)

So the level 68 PoFire NPCs were awarding 60% more exp than expected on Al'Kabor. The reasoning behind this was probably because level 66+ NPCs all quad attack, and these did have more hp and hit hard, but they didn't summon and were snareable which was a questionable decision given the exp awarded.

I checked well over 100 NPCs on Darkpaw's servers and I did find that these individual multipliers still exist, but some of them appear to have been nerfed. This doesn't surprise me since some of the multipliers were rather large and it probably made other places too unattractive. Particularly after Sony removed the flag requirements to enter into elemental planes.

I discovered that Darkpaw's level 67 flamehead mobs in the second field of PoFire awarded more exp than the level 68s did. The level 66 bridge golems granted the most at +75%, above that of the 67s and 68s. That makes sense because they are the hardest common mobs the zone. What I suspect happened is that Sony or Darkpaw nerfed the bonus exp on the 68s at some point but left the others alone. The 68s now award +25% more instead of +60% more, and the 67 flameheads award +40% more as they probably always have.

The level 65 flameheads and djinn NPCs in the first fortress area in PoFire have a 25% bonus to them on Darkpaw servers. I was able to check one of my old personal July 2003 logs where I was exping there, and I can confirm that they were 25% back then as well. The second field mobs also have significant modifiers on them but these modifiers vary.

This thread from May 2003 (about half a year after PoP launch) also verifies the +25% and +60% bonus multipliers: https://web.archive.org/web/20050320...php/t-475.html

I also noticed a reduction in Plane of Water: the level 68 hraquis NPCs actually award less exp than the baseline which is especially unusual for a water zone, and the regrua only award 12 or 17% over the baseline exp. Darkpaw's Triloun Champions actually had a larger bonus at 25%. This very much looks like Sony or Darkpaw nerfed at least the deeper area mobs at some point. Several mobs in the zone have very not-round multipliers which look like they may have been reduced algorithmically and that may also explain the hraquis giving less than baseline. My Al'Kabor logs show that level 66+ PoWater NPCs had bonus exp similar to the doomfire flameheads, and possibly as much as +100% for the regrua. A comment in the above link also agrees with this. I think old EQ's hraquis mobs may have awarded less than the regrua like Darkpaw's but still elevated way above baseline.

An AK log shows no bonus Triloun area mobs, but an early 2003 log seemed to indicate (it's just a group comment in a log, which are unreliable) the front area mobs having a bonus. I could not find old forum posts to corroborate a front area bonus however so this is very uncertain. On Darkpaw servers the large piranha actually had 25% LESS exp than baseline, which is another mob in a water zone giving less than the typical experience amount. This also suggests to me that Sony/Darkpaw nerfed the exp for many of the NPCs in the zone at some point and perhaps made some errors in the process.

I checked several rare/named NPCs in both PoFire and PoWater and all but one had larger bonuses than the nearby common NPCs and sometimes much larger. The level 61 trap mobs in PoWater awarded reduced experience as you might expect, as I got exactly half the baseline exp from them. These weaker trap NPCs may have been a motive to add this individual NPC type exp adjustment option.

Darkpaw's Plane of Earth A mobs all seem to have +20% modifiers, except the level 46 spawns from armadillos. I checked AK logs to see if this zone awarded more exp than expected and I can confirm they did award a lot more. An AK log I have clearly shows kills granting 7% in a 5 man group. The best fit multiplier estimate from this log data is +50%. Darkpaw's Plane of Earth B mobs in fact have 50% multipliers, so it appears as though they once both had 50%. I also found three old Steel Warrior forum threads mentioning EarthA's higher exp rate:

https://web.archive.org/web/20030317...cID=2911.topic
https://web.archive.org/web/20030710...cID=1844.topic
https://web.archive.org/web/20050313...hp/t-2315.html

On Al'Kabor it was known that the level 60 frogs in the Plane of Storms were unusually good exp. Zamiel's thread has exp data for these mobs, and it shows a 30% experience bonus for them. However there are two different kinds: the forest area frogs (loroks) and the center cave frogs (tempest toads, amphans). The lorok frogs don't summon and are weaker but they had this 30% bonus along with the cave frogs in old EQ. On Darkpaw servers only the center cave frogs have this 30% bonus multiplier. So Sony nerfed the loroks at some unknown point. The level 61 versions in the zone-out cave are weaker and have no bonus.

This post from June 2003 seems to indicate that they still had the bonus then since it says 6% with no charm reduction: https://thedruidsgrove.org/archive/eq/t-475.html

I found contradictory evidence of the giants in PoStorms having a bonus multiplier of 40%. Darkpaw's PoStorms level 62 giants have a 40% bonus on them and the above link also seems to indicate that this was in from early PoP as it says 10-11% for a solo giant kill, which is quite a bit and the math comes out to about 10.78% if you apply a 40% bonus. However I found a comment in a AK log of a group of 5 getting 5% for a giant kill, and 4.8% is what they would grant there with no bonus. Another AK comment says "two blue bars" (8%) solo. Also the lack of any log chatter about them being good exp suggests that they weren't on AK. There is lots of chatter about the frogs. It's possible Sony was adjusting these multipliers throughout the expansion and that's why AK lacked some.

The only Darkpaw mobs outside of elemental planes and Storms that I discovered with a bonus multiplier were some Crypt of Decay mobs. Tactics mobs did not have any that I found. I only checked about 140 NPC types however.

Supposedly the 'a metallic slime' mob in Plane of Innovation awarded a lot of exp at some point. I found this comment on Allakhazam: (Prathun was a Darkpaw dev)

"Prathun in the eqplayers forum about easter eggs in everquest said this slime is worth way more xp than normal mobs."
https://everquest.allakhazam.com/db/...77286247065402

But when I killed a metallic slime on Darkpaw servers I did not notice any bonus experience from it. They seem to have nerfed it but I can't see why they would have.

My spreadsheet lists the NPCs I checked for bonus exp on Darkpaw servers.


The PBAoE Reduction/Nerf

When I was trying to figure out EQ's experience logic for TAKP, a few things threw me off for awhile and delayed my understanding it. One of these things was a PBAoE nerf Sony put in some time during Luclin. TAKP had the incorrect AA exp amount for a couple of years because of this. We incorrectly had it at 24m because that is level 51's exp required (a false rumor that was often repeated) and my Fungus Grove logs (I AoEd that zone A LOT) had AA estimates well above even 24m which didn't make sense, and I couldn't fix our AA amount without the numbers making sense. So my adding up kills wasn't giving me correct answer to the AA amount. Later I stumbled upon the old ShowEQ threads which allowed me to complete the puzzle.

I made videos of my boxed AoEing which are on YouTube. One video shows a kill count and the amount of exp gained. If you do the math, the exp progress in the bar is less than expected. An AA being 18.75m exp is proven beyond any doubt so the only possibility is some kind of reduction for PBAoE. I checked for this in other zones I had PBAoEd in and I also found it in my Grimling Forest, Karnor's Castle and PoNightmare logs so it wasn't specific to Luclin zones.

However my Plane of Valor logs did not have any reduction. Also my Grimling Forest logs showed no reduction until level 42, and it was small at 42 but the reduction increased for each level gained, ending in a 60% reduction at level 50. Grimling Forest miners were level 35-39. This suggests that the reduction did not kick in for mobs near your level. My Plane of Nightmare log showed a smaller reduction of about 27% compared to Fungus Grove at the same level, which was 40%. My Karnor's log at level 50 had a smaller reduction (23%) compared to the Grimling Forest log of the same level. This suggests that the reduction increased the lower the level of the mob from the player.

I found another exception to this nerf: when I powerleveled my magician on Al'Kabor, I didn't group him with anybody and did not abuse the out of range group exploit. He was entirely solo. His experience was not reduced when I PBAoEed with him. It seems this nerf was only applied in groups.

I also have personal AoE logs from Acrylia Caverns in January 2002 and Fungus Grove after GoD launched in 2004. Both of these logs showed no reduction in exp. So presumably Sony nerfed AoE some time after January 2002 then removed the nerf some time during PoP. Either that or this was a nerf custom to Al'Kabor which seems unlikely.

I could not find any mention of a AoE exp nerf in Usenet or eqarchives.org. A 40% reduction in exp gained would theoretically be noticed and mentioned somewhere, but the archived posts are a tiny fraction of the posts that existed at the time so it's certainly possible they existed and were lost. It's possible this nerf was put in late Luclin and the increased exp from high level kills took over as the best exp so players just didn't notice or care about the nerf. I've seen players miss things that were more obvious than this so I wouldn't be surprised if they hadn't noticed it. Still I find it troubling that I have found no mention of it. The math is irrefutable that it existed on Al'Kabor.

I did find some comments on Project 1999's forum but they carry very little weight as they are 12 years after the fact and very vague:

"aoe groups where nerfed 3 months into luclin, and mainly in luclin areas. shit is classic sadly."
"Fungus was harder due to the exp nerf and the mobs who shroom you"

https://www.project1999.com/forums/a.../t-156004.html

Sony did nerf AoE in other ways however, and this shows motive. Days after Luclin launched, Sony patched The Grey to specifically nerf AoE groups:

Quote:
December 12, 2001 12:00 am

- Made several gameplay changes to The Grey. A single group having the
ability to simultaneously engage and prevail against 35+ experience-
giving creatures is not in line with the desire for risk vs. reward and
the promotion of balanced groups.
Sony increased MR on the NPCs, increased their HP and put in some magic immune mobs in that patch.
This old thread explains what the nerfs in The Grey were and includes Sony dev commentary: https://web.archive.org/web/20020715...cID=1160.topic

Later Sony raised the respawn timers in Acrylia Caverns and put magic resistant mobs in there as well so color stuns would not work on them. This May 2002 thread mentions the nerfs to AC: https://web.archive.org/web/20020602...rt=81&stop=100

So Sony had a history of nerfing AoE during Luclin. This makes an AoE exp reduction very plausible as Sony's other nerfs were not curbing the practice very much. It's also plausible for them to not have mentioned it because I can't find mention of the AC nerfs from Sony sources and they got so much flak after nerfing The Grey. I mention this because I got a lot of flak for nerfing AoE on TAKP and players simply wouldn't believe me that the reduction existed even with the numbers put in front of them and videos of AoE groups on Al'Kabor showing the exp rate, as if Sony wouldn't have done such a thing.
Reply With Quote
  #7  
Old 05-26-2024, 05:23 AM
Torven
Sarnak
 
Join Date: Aug 2014
Posts: 61
Default

Sources and Research

My experience spreadsheet which has a lot of data (e.g. known exp gains) used to figure out Sony's formulas; also Quest Exp:
https://docs.google.com/spreadsheets...f=true&sd=true

My Web archive of old EQ website pages which have important details about game mechanics:
https://drive.google.com/drive/folde...usp=drive_link

My exp notes file which includes exp data and player comments I used to reach my conclusions:
https://drive.google.com/file/d/1DH_...usp=drive_link

My quest exp notes and research file can be found here:
https://drive.google.com/file/d/1WD3...usp=drive_link

A copy of Zamiel's Al'Kabor Experience Thread: (note that my web archive has the original thread which is now offline and includes more info)
https://www.takproject.net/forums/in...mechanics.948/

I saved the most important ShowEQ forum pages to archive.is in case the forum goes down: (my archive also has these)
https://archive.is/www.showeq.net

I recorded video on Al'Kabor of my killing many NPCs in a short amount of time and this provided very useful data:
https://www.youtube.com/@Torven99/featured
Reply With Quote
  #8  
Old 09-04-2024, 02:44 AM
Torven
Sarnak
 
Join Date: Aug 2014
Posts: 61
Default

I will be posting addendums when I encounter evidence worth documenting. I discovered this about charm recently:

Before the May 8 2001 (mid-Velious era) patch, charmed NPCs and their targets kept their hate lists on breaks. This resulted in probable corpse poofing and no experience gain if either the broken charm pet or the enchanter's target were not blurred prior to killing them.

Sony had coded their NPCs to not pursue each other after charm had broken even while maintaining each other on their hate lists. This is not the only circumstance a NPC will keep an entity on their hate list and not pursue, as Feign death is similar in that the feigned player is not actually removed from the hate list while the player is feigned, and the chance to wipe is done on standing up. (on Live servers the Extended Targets window shows the NPCs still having the player on their lists while feigned)

The evidence for this:

Casters Realm thread, "Charmed Pet EXP" (November 2000)

Quote:
"I have charm soloed for over 20 levels. If your ex-pet does over 50% of the damage and is still alive when your target dies the corpses poofs. Always. Unless you blurred or mezzed it and got a blur off the mez.

Not to hard to test this if you really haven't seen it. Just invis, root your ex-pet, and nuke your target. This is assuming you have not blurred or mezzed the target. The corpse will poof and you get 0 exp."
Quote:
"Alandrel is completely correct. If your charm breaks, your ex pet is an npc. If that npc has done 50% of the damage to the other monster, and you kill it, it poofs."
Quote:
"The only way the creature your killing poofs is if the mob who did most of the damage is alive and NOT charmed at the time of death. A mob can break 10 times and as long as you have it charmed when whatever you are fighting finally goes down you will get the exp."
Quote:
"Your pet, charmed or summoned, is acquiring kill credit as it damages the creature. When the creature dies, kill credit is determined.
If your pet gets the kill, then it takes half the experience and your group splits the rest.

If your pet was charmed and goes wild before the creature dies, then it keeps that kill to itself, counts as an NPC kill, and the body goes poof. I had that happen to Gullerback once. (It was a funny fight, 53 shaman, 49 paladin and me, 50 enchanter trying to kill Gullerback. Everyone was OOM, so the paladin was running in big circles, with Gullerback chasing him and my pet chasing Gullerback while the shaman and I medded. I had to recharm about five times.)"
https://web.archive.org/web/20010114...ML/002835.html


Casters Realm thread, "Igbert's Guide to Charm-Solo'ing" (February 2001)

Quote:
Originally Posted by Igbert
"Poofs:

Poofs suck. That is, you blast a mob for final kill and it disappears!
No corpse, no xp. This happens because some other (non charmed) mob did more
damage than you and is still alive. Scenarios:

1. Your pet and target mob fight until almost both are completely dead.
Then charm breaks. They both wander off because they are both at a sliver
of health. They have both done 99% damage to each other. If you blast
EITHER, it will poof. If you blast both, the first will poof, the second
will not (because you killed the first who had the most credit for the
kill!). The answer? Recharm your pet, blast the second (or let your pet
kill it), then cast invis and blast your pet. You'll get both kills.
2. Both your pet and mob are rooted. A moment before the target mob dies,
your pet breaks charm. Because they are rooted next to each other, they
still fight. If the target mob dies then, it poofs. Recharm immediately if
possible.
3. Your pet fights some mob almost to the death then breaks charm and
chases you. Roots are resisted for awhile. By the time you get your pet
under control, you are halfway across the zone and have no idea where the
original target mob is. One battle later, you blast your own pet and he
poofs. The old target mob (across the zone somewhere) still did more damage
than you and got credit. The only solution here is to memblur your pet
a few times and hope it works.

Eventually, you'll lose a nice kill to a poof. Again, its part of the game,
just say "oh poop" and move on."
https://web.archive.org/web/20020527...c&f=9&t=003747

Interestingly the 500 range exp credit limit apparently did not apply to NPCs or didn't exist back then if Igbert's statement was accurate.


Sony Developer's Corner thread, "Charm Change" (May 9 2001)

Quote:
"-Prior to the 5/8 patch when I recharmed a mob it immediately went back to it's target.

-Since the 5/8 patch I always either have to order the pet to attack the previous target, or take a hit.

-Is this intentional, or a bug?

Mercuria Cestlavie
55 Troubador, The Nameless"
Quote:
Originally Posted by Absor
"That was intentional, and most certainly something we should have mentioned. Sorry about that.

Yes, when charm breaks and you re-charm the target, you will have to command it to attack again. We're changing this as part of something larger that is, basically, an exploit. The extra step of hitting the /pet attack macro is something that we had to do to make something else work right.

Sorry that I can't be more precise, but we're still working on it, and exploit fixes aren't always things we announce.

Alan"
https://web.archive.org/web/20010515...ML/000165.html


Bards complained about losing charm DPS from this change, so soon after Sony increased Bard charm to 3 ticks:

"- Increased duration on Bard Charm songs to 18 seconds from 12 seconds." (EQ Patch note, May 30 2001)


Casters Realm thread, "The New Charm and Kiting" (July 9 2001)

Quote:
"Okay, it's official. I like the new charm.

The retargetting thing is a minor annoyance, but the other benefits- extended duration and the memory blur- far outweigh the drawbacks.

The memory blur is probably the greatest thing about the new charm. As has been suggested on this board already, when the charm wears off, the mob remains aggressive toward you, but the damage counter which keeps track of whichever thing did the most damage to it is reset. What this means is that you charm a mob, you sic it on a target until the target is nearly dead, then you finish the target yourself and you get the FULL xp"
https://web.archive.org/web/20010727...&f=74&t=000924


After the May 8 2001 patch, charm seemingly behaved as it does today in regards to hate lists. I presume the change was made to prevent faction wars, but faction wars were still possible until a 2004 fix.
Reply With Quote
  #9  
Old 09-06-2024, 04:04 AM
Torven
Sarnak
 
Join Date: Aug 2014
Posts: 61
Default

I found some old claims regarding pet experience in groups that I think are credible and worth documenting here.

A guy using the handle "TedtheChanter" posted on enchanter class forums and he claimed to understand how pet exp splits worked. He also says in another thread I've found that he was either examining network packets or otherwise was using ShowEQ.

Quote:
Originally Posted by TedtheChanter in 2001
In a group, pets take exp as a group member of the caster's level, but only if they get more than half the total damage.

You can search on this BBS on my alias for numerical test results with level 1, 2, and 3 characters that clearly show the pet taking experience in groups.
https://web.archive.org/web/20010716...c&f=9&t=004834

Unfortunately his post or thread containing his "numerical test results" was seemingly not captured by Internet Archive. IA did capture another, later, post of his however:

Quote:
Originally Posted by TedtheChanter in 2002
OLD VERSION, GROUPED:
-Pet does more than half damage in a group of N players, it takes 1/(N+1) exp. Group splits the remaining N/(N+1) of exp. Depending on how many are in your group, 1/(N+1) ranges from 14% to 50%.
- Pet does less than half damage, pet takes nothing.
Quote:
Originally Posted by TedtheChanter in 2002
3. There are two small nerfs to DC in large groups.

A. DC pets used to be capped at 14% exp suck in a 6 man group in normal combat. Now they can take 25% to 50% of group experience. Obviously a nerf, but not a big one in a good group.
https://web.archive.org/web/20020718...c&f=9&t=007328

This obviously bolsters the "takes a share as if they were a group member" argument. (the opposing claim being pets take half even in groups) Until further evidence is discovered to the contrary I will lean in the direction of "takes a share".


He also claimed to understand how multiple pet splits worked post charm nerf.

In response to somebody asking the question: "1) If the group consists of 2 enchanters w/DC pets and 1 slacker druid (for snaring mostly), and the damage distribution is DC pet 1 does 45% and DC pet 2 does 45% and gimp druid does 10% damage, does the group get 100% exp for the kill or will there still be a penalty?"

he replies:

Quote:
Originally Posted by TedtheChanter in 2004
Only the last pet to engage counts, and only damage done after it is engaged is counted. Dire charm > regular charm > summoned pets, so engaging a summoned pet late will still use the Dire Charm pet rules.
https://web.archive.org/web/20040122...hp?p=11923&amp


He does make some other statements regarding experience gains that I know to be incorrect, so I wouldn't take this to be definitive proof, but it may be the best evidence that has been preserved. I would say however that I think it likely that Sony did not add up damage credit from multiple pet sources, so his claim aligns with my intuition at least.
Reply With Quote
Reply


Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

   

All times are GMT -4. The time now is 11:29 AM.


 

Everquest is a registered trademark of Daybreak Game Company LLC.
EQEmulator is not associated or affiliated in any way with Daybreak Game Company LLC.
Except where otherwise noted, this site is licensed under a Creative Commons License.
       
Powered by vBulletin®, Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Template by Bluepearl Design and vBulletin Templates - Ver3.3