A Statistical analysis of Charm Duration

[Samoht]

Originally Posted by charleski [You must be logged in to view images. Log in or Register.]

How does this affect standard guidelines? Newbie enchanters should put their points into STR, so they can carry more fine steel back to town to sell.

While I support this sentiment 100% for other caster classes, it's just not true for enchanters.

There have already been people who pointed out that we need CHA for lulls.

And the OP has already learned that they had a flawed assumption that charm lands 100% of the time. CHA is assumed to reduce initial check. As is level. As is MR.

So this whole conversation about whether or not CHA affects charm duration might be cool to figure out, but from the two statements above, it is moot because you will have CHA already. And Wilson might say that the tests from the first post show no statistical difference, but this is magic, baby, and the enchanter sees the difference between the two graphs.

So, ultimately, my statement to Wilson is that I am in agreement with the following:

Originally Posted by kjs86z2 [You must be logged in to view images. Log in or Register.]

who cares?

[Goregasmic]

Yeah.

Coincidentally I've been playing a lvl9 DE cleric and at 65cha I'm getting 50% crit fail on lulls. I'm not even getting resists it either sticks or it crit fails. I remember my ench with ~110cha was getting 10-20% in similar circumstances. At 59 with 255cha in HS north I'm probably somewhere between 5-10% but the consequences are much dire, I'd work toward 300 if it could get me sub 5%. Above 10% it is probably considered unusable if you're doing risky lull jobs.

I'm sure you could get to 60 without lull but you'd be locking yourself out of so many good spots for no particular reason. I can't remember the last time I got a charm initial resist so one could drop cha until they start getting some but that means you'd most likely get AE mez/stun resists and that's a big no no. What for exactly anyway? Sacrifice your main line of defense for an extra 200 hp/mana? You could be a little more efficient but is the extra risk worth it? But then chain mezzing/rooting/blurring adds because you couldn't lull is inefficient.

I'm trying really hard to find a scenario where you wouldn't want near max cha but outside of being a raid C2/haste bot I really can't find one that wouldn't cripple you.

[Snaggles]

I don’t even ench and know that charm will break, crit lulls will happen, it’s the player that either lives or dies.

I also didn’t think it was overly difficult capping CHA buffed, and having a decent amount of mana with all the taps and c2. Most people can pull this off with EC gear and are still better soloers than almost every other class, in most situations.

Personally my Lull classes have a moderate set of CHA gear as the diminishing returns seems to hit hard past the mid 100’s. For messing with light blue cons that’s good enough and you don’t need to swap out 20 pieces of gear.

[Goregasmic]

Originally Posted by Snaggles [You must be logged in to view images. Log in or Register.]

I also didn’t think it was overly difficult capping CHA buffed, and having a decent amount of mana with all the taps and c2. Most people can pull this off with EC gear and are still better soloers than almost every other class, in most situations.

Started my ench from scratch a year ago on green. I got to like 150cha in my mid 30s and only got 205 cha unbuffed in my mid 50s but it never crippled me. I think when you get to mid 40s and about 180 you're pretty good from that point. The thing is it is kind of easy to get cha gear but it often sucks (pieces with only cha) so my +mana and int was abysmal until my late 50s. It's true that ench get good regen but charm and nukes are super draining and all the little spells like root/tash/mez/stun chips at your bar too. It wasn't uncommon to get into trouble and have to gate because I couldn't finish adds I got. Or I started a fight at 50% and I got 4 charm breaks and I was OOM. So regen is nice but intra-fight mana total helps. Apparently it gets worse when you do the end game stuff because you need rapture + boltran which are 400 mana a pop and dictate is 750.

Originally Posted by Snaggles [You must be logged in to view images. Log in or Register.]

Personally my Lull classes have a moderate set of CHA gear as the diminishing returns seems to hit hard past the mid 100’s. For messing with light blue cons that’s good enough and you don’t need to swap out 20 pieces of gear.

Yeah if you want to do HS ent/basement or move around in like... Lguk it is good enough as a crit fail on lull won't be disastrous but if you're trying to break crypt keeper room (hs north, 9 mobs, half of them HTs) it is risky even with 255. It depends on the content you want to do really.

[Samoht]

Originally Posted by Snaggles [You must be logged in to view images. Log in or Register.]

I also didn’t think it was overly difficult capping CHA buffed, and having a decent amount of mana with all the taps and c2. Most people can pull this off with EC gear and are still better soloers than almost every other class, in most situations.

It's not difficult to cap CHA, no.

What is difficult is balancing CHA, INT/mana, HP, and MR all at once while not being encumbered. That's going to be a little harder in EC gear.

If you start with the most important stat as high as possible, though, it gives you more wiggle room to get more HP and MR.

[charleski]

Originally Posted by Goregasmic [You must be logged in to view images. Log in or Register.]

There's a good chance those repeated tick numbers was the hard cap on cajoling whispers for that level. Were you L50 for the pawn?

Ah. Yes, I think you have the answer there. The orc pawn data was collected when I was level 44, the data with a cap at 155 when I was 48 and the data capped at 161 ticks when I was 49. So you're right, this is probably simply an effect governed by caster level. Though the increase doesn't seem linear (+12 ticks from L44 to 48 vs +6 ticks form 48 to 49).

[Goregasmic]

Originally Posted by Samoht [You must be logged in to view images. Log in or Register.]

If you start with the most important stat as high as possible, though, it gives you more wiggle room to get more HP and MR.

Yeah I read guides saying cha is easy to get so go full sta or cha/int split, I went for a 15/15 split and regret it. Not a fatal mistake but I wish I went full charisma. Not all slots have interesting charisma options so going full cha frees up a bunch of those for more interesting gear.

Originally Posted by charleski [You must be logged in to view images. Log in or Register.]

Ah. Yes, I think you have the answer there. The orc pawn data was collected when I was level 44, the data with a cap at 155 when I was 48 and the data capped at 161 ticks when I was 49. So you're right, this is probably simply an effect governed by caster level. Though the increase doesn't seem linear (+12 ticks from L44 to 48 vs +6 ticks form 48 to 49).

That means a noticeable amount of pets will make it to the end, that's nice to know. Never noticed since I kill my pet after every fight for leveling but when you reach farming status it is sometimes more interesting to keep one around for a bit.

[DeathsSilkyMist]

I made a streak calculator:

https://drive.google.com/file/d/18gz...usp=drive_link - minified

https://drive.google.com/file/d/1eT1...usp=drive_link - unminified

You can copy the code into https://playcode.io/javascript and run it. The results will display in the "console" window. In the "web view" window there is a green play button you can use to run the calculation again. I'd suggest using the minified version, as playcode.io starts to ask for money if the line count is greater than 8.

If you take a look here, this is the minimum chance for a charm break in the EQEMU code:

https://github.com/EQEmu/Server/blob...ells.cpp#L5546

//Average charm duration agianst mobs with 0% chance to resist on LIVE is ~ 68 ticks.
//Minimum resist chance should be caclulated factoring in the RuleI(Spells, CharmBreakCheckChance)
if (CharmTick) {

float min_charmbreakchance = ((100.0f/static_cast<float>(RuleI(Spells, CharmBreakCheckChance)))/66.0f * 100.0f)*2.0f;
if (resist_chance < static_cast<int>(min_charmbreakchance))
resist_chance = min_charmbreakchance;
}

CharmBreakCheckChance is set to 25. When you do the math the chance comes out to 12.1212.

The roll against the resist_chance is here:

https://github.com/EQEmu/Server/blob...ells.cpp#L5562

//Finally our roll
int roll = zone->random.Int(0, RuleB(Spells, EnableResistSoftCap) ? RuleI(Spells, SpellResistSoftCap) : 200);

if(roll > resist_chance) {
return 100;
}

As far as I know the SpellResistSoftCap (255) is disabled. This means you are rolling a d200 and seeing if it beats 12.1212 when you have the best chance to avoid a charm break.

The streak calculator rolls millions of dice and does the if(roll > resist_chance) check on each roll until you get a fail. Each time you fail is a streak. So if you rolled 5 successful rolls in a row and then 1 fail, that is a streak of 5.

The average charm duration is based on what the resistance chance is. The higher the resistance chance, the worse your average.

You can see that the calculator data has the same pattern as OP's data:

[You must be logged in to view images. Log in or Register.]

TLDR: When rolling a d200 with a minimum resistance chance of 12.1212, the average charm duration is 40 ticks when your max charm duration is 190 ticks.

If you want the average based on when the data tapers off, the average charm duration is 27 ticks when the data tapers off at 82 ticks.

The average charm duration is based on what the resistance chance is. The higher the resistance chance, the worse your average.

The reason why we can average the Successful Dice Roll Streaks is because the basis for the calculation is simply millions of dice rolls, which approach a normal (gaussian) distribution via the central limit theorem.

[kjs86z2]

all that matters is FT8

[bcbrown]

Originally Posted by DeathsSilkyMist [You must be logged in to view images. Log in or Register.]

I made a streak calculator:

This is an excellent example of using Monte Carlo methods to generate an exponential or geometric distribution. Good work!