CHA seems to be impacting charm durations on P99 far more than it did on Live

[Dolalin]

Compare this test on P99 with varying charisma scores on a blue mob:


First, I found a pet who was just on the cusp of charming viability. At level 52 I grabbed a goo hitting for 116. Prior to this we tried a goo hitting for 120 but even with full charisma gear could not keep it charmed with duration good enough to exp reliably. This choice was intentional, because what I really care about is keeping the best mob I can for as long as I can. I'm sure results would be very different for a light blue mob. On every break the mob was tashed and re-charmed. I just pulled all the data out of my log file after our session and crunched it all using excel. results are as follows:

High Charisma dataset (CHA = 224)
Time of trial: 0:40:18 (or 0.672 hours)
Breaks: 7
Breaks per hour(extrapolated): 10.42
Avg Duration: 5.76 minutes
Median Duration: 3 minutes 10 seconds

Low Charisma dataset (CHA = 95)
Time of trial: 0:58:04 (0.968 hours)
Breaks: 25
Breaks per hour(extrapolated): 25.83
Avg Duration: 2.32 minutes
Median Duration: 1 minute 4 seconds

https://www.project1999.com/forums/showpost.php?p=738889&postcount=52


To this test conducted in classic with varying Charisma scores on a blue mob:


Xenti, you said there have been studies done on both mes and charm. I don't ever remember data being posted on charm. That was the whole reason I did this test was to provide hard numbers. Can you give me a link to the previous test?
With that said, in my insanity today I decided to run more tests. This time I ran them on Ssolet Dnaas, an Iksar in Warslick's Woods that is for part of the Veeshan's Peak key quest. He is 50th level and cons blue to a 60th level player. This means that he is most likely in the level range that most people with Boltran's will be engaging. And to test on an even con mob with Boltran's would require doing the tests at 53rd level as after that you can only charm blue cons.

I also decided to extend my tests further this time. I did the first 2 tests again of 75 cha and 205 cha each with a sample size of 25. Then I ran 2 more tests of 25 samples at 75 and 205 cha, but this time I also buffed the Iksar with Resist Magic. Here are the results.

Test 1
Charisma = 75
Mob Buffed = No
Total Resists = 3
Lowest Duration Charm = 6 seconds
Longest Duration Charm = 424 seconds (7:04)
Average Duration Charm = 204 seconds (3:24)

Test 2
Charisma = 205
Mob Buffed = No
Total Resists = 1
Lowest Duration Charm = 6 seconds
Longest Duration Charm = 425 seconds (7:05)
Average Duration Charm = 119 seconds (1:59)

Test 3
Charisma = 75
Mob Buffed = Yes
Total Resists = 11
Lowest Duration Charm = 6 seconds
Longest Duration Charm = 231 seconds (3:51)
Average Duration Charm = 71 seconds (1:11)

Test 4
Charisma = 205
Mob Buffed = Yes
Total Resists = 10
Lowest Duration Charm = 4 seconds
Longest Duration Charm = 143 seconds (2:23)
Average Duration Charm = 36 seconds

Conclusions

The tests without buffing show some strange occurences. With 75 CHA the average duration went up by almost a full minute, and with 205 CHA it went down almost a full minute from my previous test. My guess is that this means there simply aren't enough samples taken yet.

Now it is very obvious that MR has a very large impact. Not only did it cut the max duration charm for both CHA settings in half, it also cut the average durations in half. And on top of that resists went up significantly.

And the most interesting thing with this new set of tests is that all of the tests performed better when run with 75 CHA than they did with 205 CHA. The average durations at 75 CHA where almost twice as long as those at 205 CHA.

So what does this mean to me? After running 150 tests I still see no data showing that high CHA gives any benefit at all. In fact it might be just the opposite. The only thing that does stick out to me is that charm is much more successful when magic resist is lower. So I will continue to keep my CHA wherever it ends up when focusing on other gear, and will use Tash while charming.

Wandatin Dai'Noga

http://web.archive.org/web/20010501121237/http://forums.castersrealm.com/cgi-bin/eq/ultimatebb.cgi?ubb=get_topic&f=9&t=004395


The effect of CHA on his charms was so small that it got lost in the statistical noise.

I suspect P99 is allocating much too high of an impact to Charisma score when calculating its impact on charm duration.

[strawman]

I read recently that in the early days of EQ, people believed that lower charisma improved the success of Necromancer charm.

In both of the original poster's sets of tests, lower charisma clearly correlates with higher average Enchanter charm duration.

[Dolalin]

I read recently that in the early days of EQ, people believed that lower charisma improved the success of Necromancer charm.

In both of the original poster's sets of tests, lower charisma clearly correlates with higher average Enchanter charm duration.

I don't think so, because he also ran a test with a light-green-con Seafury Cyclops and it showed no impact of CHA on his charms. If it were a negative impact I would expect it to show something. Rather it just looks like the effect of CHA was so small it gets lost in statistical noise.

From the same classic link:


Ok, first off I don't know anything about statistics. About the only thing I do know is that the larger the sample size the more accurate the results. So maybe if we band together on this some we can all help contribute to a large sample size.

First off, the specifics. I am a 58th level Enchanter. The tests were run on a Seafury Cyclops in the Ocean of Tears. Both tests where run on the same Cyclops. The Cyclops conned green to me, but at the end of the test I killed it and it did give me experience so it was probably 1 level below blue. The tests where run without using Tashanian so as to only see the benefits of Charisma. And each test includes a total of 25 charms. Yes, I know this is WAY too low to have a completely accurate result set, but damn, I can only sit it one place for so long charming the same mob.

Test 1
Charisma = 75
Total Resists = 2
Lowest Duration Charm = 5 seconds
Longest Duration Charm = 428 seconds (7:08)
Average Duration = 156 seconds (2:36)

Test 2
Charisma = 205
Total Resists = 2
Lowest Duration Charm = 7 seconds
Longest Duration Charm = 426 seconds (7:06)
Average Duration = 161 seconds (2:41)

Specific Charm Durations:

Test 1: 164, 310, 352, 47, 185, 169, 115, 102, 390, 178, 65, 108, 34, 375, 5, 20, 22, 213, 11, 428, 22, 70, 90, 426.

Test 2: 260, 111, 26, 426, 80, 426, 56, 143, 226, 112, 10, 103, 185, 280, 7, 343, 42, 9, 212, 308, 132, 27, 51, 425, 31.

Conclusion: Well, I don't even know if I should be making one as this is far too small of a sample set. And feel free to tear me a new one for doing so, or even better yet contribute your own hard data to give a better understanding.

But what I see from this is that CHA makes no difference at all. The reason could be the mob was green. I can't test on an even con as Boltran's doesn't work on a 58th level mob.

Each test had a total of 2 resists. The lowest and highest durations between each test only had a difference of 2 seconds each, oddly with worst lowest on low CHA and worst highest with high CHA. And the average duration between the tests was only 5 seconds yet there was a difference of 130 Charisma.

What this means is that I will stick with my gut feel of 160 played days and go with another stat besides CHA. I didn't believe in it's value before, and I won't change my mind unless I can see some good hard numbers that show me otherwise.

Wandatin Dai'Noga

[strawman]

I don't think so, because he also ran a test with a light-green-con Seafury Cyclops and it showed no impact of CHA on his charms. If it were a negative impact I would expect it to show something. Rather it just looks like the effect of CHA was so small it gets lost in statistical noise.

From the same classic link:

The inverse charisma effect was present in the same magnitude in both sets of tests on the blue mob, and it was absent in both sets of tests on the green mob.

If we're using this data to prove no correlation between charisma and charm time, it seems strange to hand-wave away the correlation it shows between charisma and charm time.

[Dolalin]

If CHA were working to increase mob resistance don't you suppose we'd see it in the result set for the green mob too?

It just seems really improbable that it worked to decrease charm duration. But maybe it was bugged that way. Wouldn't be the first time Verant screwed up.

[strawman]

If CHA were working to increase mob resistance don't you suppose we'd see it in the result set for the green mob too?

It just seems really improbable that it worked to decrease charm duration. But maybe it was bugged that way. Wouldn't be the first time Verant screwed up.

This data says to me that there is indeed a ticking CHA-based resistance throw for charm, but it's always successful on a trivial mob, and it's backwards on non-trivial mobs (succeeding more with lower CHA).

This doesn't necessarily make intuitive sense to the player, and maybe the backwards CHA check is a bug, but Verant did make items like https://wiki.project1999.com/Staff_of_Writhing for some reason.

(Upon actually reading the Staff of Writhing wiki page, it drops from a mob that charms. I wonder if that was supposed to be a clue?)

[Dolalin]

^^ I really think that's a massive leap.

Out of curiosity I searched for what TAKP had come up with regarding CHA and charm durations. Torven figured the effect of CHA on charm duration was so small (if at all) that they just don't take it into account at all.

Links:

https://www.takproject.net/forums/index.php?threads/diving-back-in.5487/page-3

https://www.takproject.net/forums/index.php?threads/charm-changes-over-nerfed.2575/#post-13018


Charisma not affecting duration mostly comes from old forum posts of enchanters, plus Kayen's investigation.


I'll message Torven and maybe he can chime in.

[fadetree]

Conclusion: Well, I don't even know if I should be making one as this is far too small of a sample set.

[Tecmos Deception]

Items of note:

- tests done on p1999 (both propo's and loraen's) were done WITH tash, but the test linked to from classic era was done WITHOUT tash.

- Loraen's tests - https://www.project1999.com/forums/showthread.php?t=92423 - show pretty similar charm durations on higher-level charms (loraen at 60 vs a 53) to the durations shown by the classic chanter (58 chanter vs a 50).

- But loraen's tests show MUCH LONGER charms against a level 46 (minimum blue con) than the level 58 chanter's charms against a seafury (exp-awarding green con) did. Again, remember Loraen was using tash and the level 58 guy was not.

- Classic-era test from the 58 chanter showed that MR was having an enormous impact on charm duration. While he never (that we saw) posted a test where he was using tash, he did post a test where he put Resist Magic on the same level 50 mob he'd been charming earlier, and charm outright resists, durations, etc. suffered a lot.


What would be fantastic is finding a test from classic era where someone tested charm durations WITH tash on, both on a higher-blue and on a low-blue or green con. But even without that, I "feel" like charisma needs to have a more minor role in charming but without that change affecting durations of charms on p1999. It seems possible/probable that low-level mobs charm too easily on p99, but I haven't seen any data of classic-era tests done with tash on lower-level mobs to see how that played out.

While lots of folks are currently complaining about charm being OP (and have always been doing so, really), if we assume that MR debuffs affect charm durations in a more-or-less inverted way to the way Wandatin's Resist Magic affected his charm durations... then it seems to make sense for p1999 charms with tash to hold well, and p1999 charms with malaise also to last REALLY well, and p1999 charms with -mr gear to last forever when charming a mid-blue con.

[cd288]

Items of note:

- tests done on p1999 (both propo's and loraen's) were done WITH tash, but the test linked to from classic era was done WITHOUT tash.

- Loraen's tests - https://www.project1999.com/forums/showthread.php?t=92423 - show pretty similar charm durations on higher-level charms (loraen at 60 vs a 53) to the durations shown by the classic chanter (58 chanter vs a 50).

- But loraen's tests show MUCH LONGER charms against a level 46 (minimum blue con) than the level 58 chanter's charms against a seafury (exp-awarding green con) did. Again, remember Loraen was using tash and the level 58 guy was not.

- Classic-era test from the 58 chanter showed that MR was having an enormous impact on charm duration. While he never (that we saw) posted a test where he was using tash, he did post a test where he put Resist Magic on the same level 50 mob he'd been charming earlier, and charm outright resists, durations, etc. suffered a lot.


What would be fantastic is finding a test from classic era where someone tested charm durations WITH tash on, both on a higher-blue and on a low-blue or green con. But even without that, I "feel" like charisma needs to have a more minor role in charming but without that change affecting durations of charms on p1999. It seems possible/probable that low-level mobs charm too easily on p99, but I haven't seen any data of classic-era tests done with tash on lower-level mobs to see how that played out.

While lots of folks are currently complaining about charm being OP (and have always been doing so, really), if we assume that MR debuffs affect charm durations in a more-or-less inverted way to the way Wandatin's Resist Magic affected his charm durations... then it seems to make sense for p1999 charms with tash to hold well, and p1999 charms with malaise also to last REALLY well, and p1999 charms with -mr gear to last forever when charming a mid-blue con.

This seems like a very valid argument to me.

Also, I'm sort of confused how the original evidence presented in this thread shows anything. It shows a bunch of samples that are inconsistent with each other. Not sure what conclusion is supposed to be drawn from that.

[Tecmos Deception]

Eh, yeah, I guess I didn't look closely enough at his high vs low charisma tests. His tests against the level 50 mob were odd. His tests against the green con earlier in that thread (not specifically linked here) show no difference in duration or resists with 75 vs 205 charisma though.

His data definitely seems to suggest charisma doesn't HELP charm durations. It isn't entirely clear if it somehow hurts them I guess. It'd be interesting to see if the TAKP people found different sources besides these and all the classic-era anecdotes.

What more interested me was his average and max charm durations on mobs of different level and magic resistance, and implications of those for p99's usual tashed, blue-con charms when solo and tashed, malod, blue-con charms when grouped.

[Dolalin]

No classic data set is going to be perfect, or this controversy wouldn't still exist 20 years later.

What we're looking for here is a signal in the noise that says "CHA matters!" and it's not there, at least not from these runs.

I messaged both Torven and Haynar and maybe they will chime in with a TAKP perspective if we're lucky.

[cd288]

No classic data set is going to be perfect, or this controversy wouldn't still exist 20 years later.

What we're looking for here is a signal in the noise that says "CHA matters!" and it's not there, at least not from these runs.

I messaged both Torven and Haynar and maybe they will chime in with a TAKP perspective if we're lucky.

I'm definitely not suggesting that any data set is going to be perfect.

I think we need more data points though. What's the MR of the mobs charmed in each study. What's the level of the Enchanter versus the level of the mob being charmed in each instance (the P99 test may have disclosed the player's level, but I didn't see it in the quote)? Both of those are established as two of the saves you have to make for a Charm break. IIRC, CHA is supposedly the third one and the least important of the three. That could explain why the CHA in the Live tests doesn't seem to make as much of a difference versus the duration, and perhaps the P99 tests MR and level difference is creating the long duration.

[Dolalin]

Agreed on more data points, but I'm having to go out of the classic era now to provide them.

Anyways here's a post from 2003 testing whether CHA affects charm duration, verdict was that it did not. BUT it seems to have affected initial resist chance:


Yandie (TD3)
A Crystalline Golem #1
str 146 Start End Duration Running Average
sta 153 23:52:34 23:53:41 0:01:07 0:01:07
agi 129 23:54:01 23:55:56 0:01:55 0:01:31
dex 112 23:56:27 23:58:03 0:01:36 0:01:33
wis 207 23:58:38 23:58:52 0:00:14 0:01:13
int 315 23:59:15 23:59:28 0:00:13 0:01:01
cha 305 23:59:44 23:59:58 0:00:14 0:00:53
0:00:16 0:00:35 0:00:19 0:00:48
PR 75 0:00:50 0:05:21 0:04:31 0:01:16
MR 184 0:06:06 0:06:52 0:00:46 0:01:13
DR 70 0:07:15 0:08:05 0:00:50 0:01:10
FR 156 0:09:47 0:10:11 0:00:24 0:01:06
CR 116 0:10:27 0:11:24 0:00:57 0:01:05
0:11:40 0:15:28 0:03:48 0:01:18
Buffs 0:16:03 0:16:35 0:00:32 0:01:15
SoW(Potion) 0:16:46 0:17:05 0:00:19 0:01:11
VoQ 0:17:18 0:18:24 0:01:06 0:01:11
AR 0:18:59 0:20:13 0:01:14 0:01:11
SoM 0:20:29 0:28:15 0:07:46 0:01:33
Air Elem 0:28:32 0:30:22 0:01:50 0:01:34
OS 0:30:33 0:31:41 0:01:08 0:01:32
2 55ish 0:32:01 0:32:18 0:00:17 0:01:29
cleric hp 0:32:29 0:33:19 0:00:50 0:01:27
buffs 0:33:41 0:36:21 0:02:40 0:01:30
0:36:40 0:38:35 0:01:55 0:01:31
Average 0:01:31
Casts 24
Resists 0
Resist %0

Yandie (TD3)
A Crystalline Golem #1
str 108 Start End Duration Running Average
sta 140 0:45:29 0:47:25 0:01:56 0:01:56
agi 129 0:47:37 0:51:40 0:04:03 0:02:59
dex 109 0:51:55 0:54:13 0:02:18 0:02:46
wis 188 0:54:48 0:56:14 0:01:26 0:02:26
int 288 0:56:29 0:56:38 0:00:09 0:01:58
cha 130 0:57:06 0:01:58
0:57:14 1:02:07 0:04:53 0:02:27
PR 63 1:02:26 0:02:27
MR 162 1:02:34 1:05:03 0:02:29 0:02:28
DR 58 1:05:22 1:08:12 0:02:50 0:02:30
FR 138 1:08:43 1:09:43 0:01:00 0:02:20
CR 130 1:10:01 1:10:13 0:00:12 0:02:08
1:10:26 0:02:08
Buffs 1:10:33 0:02:08
SoW(Potion) 1:10:41 1:12:57 0:02:16 0:02:08
VoQ 1:13:13 1:13:28 0:00:15 0:01:59
AR 1:16:42 1:18:24 0:01:42 0:01:58
SoM 1:19:07 1:19:25 0:00:18 0:01:50
Air Elem 1:19:33 1:19:55 0:00:22 0:01:45
1:22:25 0:01:45
1:22:33 1:23:32 0:00:59 0:01:42
1:23:55 1:24:58 0:01:03 0:01:39
1:25:38 1:27:42 0:02:04 0:01:41
1:27:55 1:28:12 0:00:17 0:01:36
1:28:35 0:01:36
1:28:43 0:01:36
1:28:50 1:30:56 0:02:06 0:01:38
1:31:11 0:01:38
1:31:18 1:33:35 0:02:17 0:01:40
1:33:47 1:34:05 0:00:18 0:01:36
1:34:16 0:01:36
1:34:34 0:01:36
1:34:41 1:34:53 0:00:12 0:01:32
1:35:00 1:36:25 0:01:25 0:01:32
1:36:36 1:37:50 0:01:14 0:01:31
1:38:02 0:01:31
1:38:09 1:38:51 0:00:42 0:01:29
1:39:04 0:01:29
1:39:11 0:01:29
1:39:19 1:39:45 0:00:26 0:01:27
1:39:55 1:40:40 0:00:45 0:01:26
1:41:01 1:41:47 0:00:46 0:01:24
1:42:13 1:42:53 0:00:40 0:01:23
Average 0:01:23
Casts 43
Resists 13
Resist %30.23


http://web.archive.org/web/20041126094504/www.therunes.net/forums/viewtopic.php?p=10785

[Tecmos Deception]

I'm definitely not suggesting that any data set is going to be perfect.

I think we need more data points though. What's the MR of the mobs charmed in each study. What's the level of the Enchanter versus the level of the mob being charmed in each instance (the P99 test may have disclosed the player's level, but I didn't see it in the quote)? Both of those are established as two of the saves you have to make for a Charm break. IIRC, CHA is supposedly the third one and the least important of the three. That could explain why the CHA in the Live tests doesn't seem to make as much of a difference versus the duration, and perhaps the P99 tests MR and level difference is creating the long duration.

You didn't go to the source, eh?

The chanter, Wandatin, was level 58 the day he made the thread and was testing charm against a green-con-but-exp-giving seafury (so like... level 42-43?) with no tash or other MR effects on the target. The next day, iirc, he posted in the same thread saying that he tested charms against a level 50 NPC with no tash, and then again with no tash but WITH Resist Magic (+40 MR). He presumably was still level 58 in the later tests he did since it was less than 24 hours later, even though he doesn't actually spell it out.

[derpcake2]

The first question shouldn't be how stats impact charm on p1999 or how they did on live, it should be "does p1999 charm resemble classic under similar conditions".

Not entirely sure how parses from PoP are useful, they offer numbers which can't be compared to p1999.

[Tecmos Deception]

The first question shouldn't be how stats impact charm on p1999, it should be "does p1999 charm resemble classic under similar conditions".

Not entirely sure how parses from someone having 305 cha are related to the situation on classic, or p1999.

It's not grand slam evidence, but it is something worth considering.

I mean... if charisma doesn't boost charm durations at 305 charisma, then it seems pretty fucking likely that it wasn't going to boost charm durations at 250, 200, or anything else either, right? That it wasn't actually in classic era just means it isn't a key bit of evidence.

Staff aren't dummies. They aren't going to base mechanics on non-classic era unless they absolutely have to. But if some classic-era evidence suggests charisma didn't affect charm duration, and there's no patch notes or other mention of changes to these mechanics, and post-classic-era evidence suggests that even more charisma still didn't affect charm durations... you see where I'm going with this.

[derpcake2]

It's not grand slam evidence, but it is something worth considering.

As I said before, the wrong question is being answered.

We don't know if charm on p1999 lasts longer then it does on classic, everything else being the same.

Proving that CHA has more of an effect on p1999 then it has on live, would most likely lead to nerfing the impact CHA has on p1999.

Doing this without answering the first question is not how this process should work.


Staff aren't dummies. They aren't going to base mechanics on non-classic era unless they absolutely have to.

I know staff is trying their very best and appreciate their effort a lot, but given the mage pet situation, I'd rather prevent people pushing for "fixes" which result in more staff work to correct them afterwards.

[Dolalin]

Another possibility is that CHA did factor into the per-tick break check, but there was an effectiveness cap that was fairly low, like the vendor price CHA cap, which I think is 100-130ish (so says EQTraders).

That would actually align pretty well with past EQ dev posts on the subject vs these results.

[strawman]

^^ I really think that's a massive leap.

What we're looking for here is a signal in the noise that says "CHA matters!" and it's not there, at least not from these runs.

The signal is there - you're just ignoring it. It's plain as day in the first data set. Negative charisma means longer charms on non-trivial mobs.

The other data sets posted here are 1. an out-of-era test on an out-of-era level 62 mob and 2. a test on an extreme green-con (level 1 decaying skeleton).

Neither addresses what the first data set is showing us - which is that in two separate test sets, the average duration of charm is twice as long for the enchanter with 75 CHA than it was for the enchanter with 205 CHA.

[Torven]

Charm's precise mechanics are not 100% certain, but I know this:

I had found some old enchanter posts of people who had done charm tests at varying CHA levels and the results indicated either no change or so small as to be within the (albeit significant) margin of error.

I did a lot of charming on animals on my druid on Al'Kabor, so I have a lot of data from it. Charm duration was ~5.5 minutes average using 20 minute charm spells. This was the case on both green mobs and dark blues. (using animal tash as well) Halfling druids have low CHA.

A charm test (an overnight log with automated casting) I did on Live using a level 25 enchanter with significant CHA on a level 1 decaying skeleton in 2015 resulted in a very similar average duration. (~55 ticks)

Daybreak recently stated that charm (and root, blind, fear) have a minimum 5% chance to break per tick. So that explains why the spell can break on level 1s with max CHA. However if it were 5% chance per tick every tick, it would break far too often compared to the observations. So that leads me to believe that the Daybreak dev omitted another roll that determines if the save throw is even checked. Currently this roll is 25% on TAKP. (total domination AA reduces it)

The Daybreak developer was explaining what charisma does for the player:

how much merchants will charge you
bard fizzle chance
Chance of an NPC aggro'ing you when you cast pacify
Chance for SPA 63 to successfully blur your target
Chance for DI spells to heal for their full amount
Chance for SPA 22(charm), 31(mez), 34(confuse), 63(blur) to be resisted, you hit this cap at 200 CHA
Note SPA 3(movement rate?), 20(blind), 22(charm), 99(root) have a minimum 5% chance of breaking every tick regardless of how much CHA you have. (only 22 gets a CHA bonus)

note: I added the ()s except the last one; that's copied from our code comments that I put above our logic.

So that post says that charm, mez, confuse and blur resist rolls have a CHA modifier when the spells land. Then it suggests (I'm not 100% sure on that, it's a little ambiguous) a bonus for charm tick saves.

I think the most likely conclusion given the available data is that CHA merely helps (if at all) charm hold when the chance to break is above the 5% minimum-- i.e. the NPC's MR is high enough to resist above 5%-- such that if the target is zero MR then CHA does nothing to help. Achieving that minimum is also easily accomplished with tash on the target for common NPCs and tash is not even needed if the NPC is a low dark blue.

Incidentally I made the chance 6% on TAKP since 25% * 5% seemed too good in my simulations vs. the data I had. 25% is a guess anyway and it's possible rounding errors or comparison operators (> vs >= etc) might explain the discrepancy.

[derpcake2]

Another possibility is that CHA did factor into the per-tick break check, but there was an effectiveness cap that was fairly low, like the vendor price CHA cap, which I think is 100-130ish (so says EQTraders).

That would actually align pretty well with past EQ dev posts on the subject vs these results.

You've seen my previous post, and since you didn't refute it, I'll be waiting for the answer to the first question.

You already broke magicians, without a second thought beyond achieving your goal, the community can do without your agenda-feedback.

I understand you said "sorry" for that, and I don't care, because you obviously haven't learned anything from the situation.

Did you consult Haynar before going onto a crusade which resulted into no mage pets? You claim to have sources you can rely on, but went on a 50 hour quest which ended up by staff having to undo previous work?

You should at the very least be flagged for providing feedback that should be taken with some buckets of salt.

[Dolalin]

Thanks Torven.

Given I just saw Haynar had made some changes to Enchanter charm in 2015, maybe P99 is already using rolls similar to TAKP?

In which case the first test I posted from that p99 Enchanter might be obsolete.

[cd288]

Charm's precise mechanics are not 100% certain, but I know this:

I had found some old enchanter posts of people who had done charm tests at varying CHA levels and the results indicated either no change or so small as to be within the (albeit significant) margin of error.

I did a lot of charming on animals on my druid on Al'Kabor, so I have a lot of data from it. Charm duration was ~5.5 minutes average using 20 minute charm spells. This was the case on both green mobs and dark blues. (using animal tash as well) Halfling druids have low CHA.

A charm test (an overnight log with automated casting) I did on Live using a level 25 enchanter with significant CHA on a level 1 decaying skeleton in 2015 resulted in a very similar average duration. (~55 ticks)

Daybreak recently stated that charm (and root, blind, fear) have a minimum 5% chance to break per tick. So that explains why the spell can break on level 1s with max CHA. However if it were 5% chance per tick every tick, it would break far too often compared to the observations. So that leads me to believe that the Daybreak dev omitted another roll that determines if the save throw is even checked. Currently this roll is 25% on TAKP. (total domination AA reduces it)

The Daybreak developer was explaining what charisma does for the player:



note: I added the ()s except the last one; that's copied from our code comments that I put above our logic.

So that post says that charm, mez, confuse and blur resist rolls have a CHA modifier when the spells land. Then it suggests (I'm not 100% sure on that, it's a little ambiguous) a bonus for charm tick saves.

I think the most likely conclusion given the available data is that CHA merely helps (if at all) charm hold when the chance to break is above the 5% minimum-- i.e. the NPC's MR is high enough to resist above 5%-- such that if the target is zero MR then CHA does nothing to help. Achieving that minimum is also easily accomplished with tash on the target for common NPCs and tash is not even needed if the NPC is a low dark blue.

Incidentally I made the chance 6% on TAKP since 25% * 5% seemed too good in my simulations vs. the data I had. 25% is a guess anyway and it's possible rounding errors or comparison operators (> vs >= etc) might explain the discrepancy.

So just to clarify these are all tests from an emulated server in the PoP era, Live EQ today, and statements from Daybreak as to how Charm works on their servers today?

[Torven]

No, my druid charming was done on the Al'Kabor server which was a Sony server that was essentially EQ circa November 2002 as far as mechanics goes. Not to be confused with The Al'Kabor Project which is the emu I help to develop. I'm sure p99 devs took a lot of data from Al'Kabor when it was up; it's the most legitimate source the emus had for a long time.

Unless Al'Kabor's server code is leaked, you're not going to get definitive answers. We can only make best guesses. (even then charm could have been changed between 1999 and 2002) I spent a lot of time parsing AK logs to get average charm durations to compare with logic that we've implemented on our server. I favor log data above comments and even dev posts.

I spent a great deal of time working on EQ mechanics and collecting data on Live servers. I can say that much of the underlying logic was left intact and that Sony tended to build on top of it instead of scraping it and redoing it. (although lately they tend to be changing more of it) It wouldn't surprise me if charm's CHA modifiers had remained the same all these years-- particularly when the dev says the cap is 200 still.

[Wurl]

Thanks a lot for the good research in this thread, Dolalin and Torven.

[cd288]

No, my druid charming was done on the Al'Kabor server which was a Sony server that was essentially EQ circa November 2002 as far as mechanics goes. Not to be confused with The Al'Kabor Project which is the emu I help to develop. I'm sure p99 devs took a lot of data from Al'Kabor when it was up; it's the most legitimate source the emus had for a long time.

Unless Al'Kabor's server code is leaked, you're not going to get definitive answers. We can only make best guesses. (even then charm could have been changed between 1999 and 2002) I spent a lot of time parsing AK logs to get average charm durations to compare with logic that we've implemented on our server. I favor log data above comments and even dev posts.

I spent a great deal of time working on EQ mechanics and collecting data on Live servers. I can say that much of the underlying logic was left intact and that Sony tended to build on top of it instead of scraping it and redoing it. (although lately they tend to be changing more of it) It wouldn't surprise me if charm's CHA modifiers had remained the same all these years-- particularly when the dev says the cap is 200 still.

So the test was done on a server that is Luclin era mechanics? (Luclin was December 2001)

[cd288]

Dol, also have an additional question for you on this. How do we square this with the statement from Brad McQuaid on the Wiki about Charisma (copied below for reference)?

"CHA: affects amount you will be paid for goods by NPC merchants, and how much they will pay you; affects the saving throw on certain bard and enchanter spells (charms in particular)"

[Dolalin]

Dol, also have an additional question for you on this. How do we square this with the statement from Brad McQuaid on the Wiki about Charisma (copied below for reference)?

"CHA: affects amount you will be paid for goods by NPC merchants, and how much they will pay you; affects the saving throw on certain bard and enchanter spells (charms in particular)"

I've been mulling this over and I think Torven's interpretation, namely that the CHA save on charm breaks existed but was only significant against e.g. even cons where you weren't already at the MR resist floor, does fit the data. We wouldn't have expected to see any CHA bonus for green and blue con charm break tests, but that's what everyone seems to have been testing against because let's face it, it was safer.

The problem really is that there's so little data on even con+ charms and we're never going to get any more.

Maybe Haynar has already coded up something resembling TAKP's charm logic anyways on P99 and this thread has nothing new to offer. Always good to discuss though.

[Keebz]

This looks like PoP(?) era, but this guide (http://www.therunes.net/rguidehighlevel2.htm) suggests "Charm does an initial check that consists of a level of caster vs level of mob modified by MR.
Charisma affects this by adding a modifier to the MR approx 10 percent of your total charisma.. so a 300 chr enchanter gets a bonus neg 30 check. After the initial charm land..theres a per tick check based off the server clock."

It is unclear to me how noticeable the difference between say 100 and 200 CHA would be if this were mechanism and whether it jives with the log data.

[DMN]

Those dev comments are strange, were they made pre or post AA's? Did snares suddenly start getting chances to break every tick sometime ater velious? Druid and necro snares certainly never had random durations through velious(during vanilla there was a brief period of time ensanre did have a tick based break chance) which nakes me think that something was changed-- perhaps to make AA's more balanced and/or appealing. Would also suggest they might be messsing around with the % and break chance formulas for similar reasons.

Getting back to the charm duration and testing, sadly, like I mentioned already. the base randomness of charm is so high that trying to find any signal amongst the cacophony of noise would certainly take many hundreds of hours of data, maybe even over a thousand, and in this case, for any factual conclusion to be drawn, we'd need them both from then AND now.

[Dolalin]

I've shelved this issue for awhile to be honest. I agree DMN, there's just not enough evidence to say one way or the other. Maybe someday I'll stumble on some though.

[xmaerx]

It's likely just that the formula has a small window for Charisma to be useful. Mobs significantly lower level probably get squashed by the caster's CHA + resist mods versus their level, mobs with innately higher resist probably squash the CHA mod entirely, and anything outside of a specific range is subject to the goofy resistibility mods P99 deploys for anything with a stat-based resistibility modifier.

Then again, it seems like charm (and pacify) have initial-cast modifiers outside of the intended "secondary" check, with charm being virtually irresistible, and pacify resisting like mad.

I feel like resistibility in general on P99 is wonky, and different level spells should have different resist rates (low level stun versus high level stun, etc) per the "resist mods" in the spdat file. Higher level spells are Magic(0) according to the SpDat, but lower spells are almost all Magic(25).. except for charm, which is all over the place resistibility wise.