A look at the Taunt skill

[Paen]

Starting a new thread since I dont really want to post this in the steaming pile of shit the last one has become

All tests done on a mezzed mob, real xp mobs in real xp zones. They are a rather small sample size but they display the issues I have with taunt rather well.

If anyone wants the logs for some reason I have no problem uploading them somewhere just let me know.

Level 1 Snake

108 taunts 43 failures 61% success rate

2 failures in a row 9
4 failures in a row 2
6 failures in a row 1

Level 65 Kyv

235 taunts 94 failures 60% success rate

2 failures in a row 13
3 failures in a row 1
4 failures in a row 3

Level 68 Kyv

210 taunts 145 failures 31% success rate

2 failures in a row 11
3 failures in a row 7
4 failures in a row 4
5 failures in a row 2
7 failures in a row 3
8 failures in a row 1
20 failures in a row 1 (yes 20 failures, not a typo)

A level 1 snake and a level 65 kyv have the same success rate /boggle, which also makes the chances of streaked failures relatively the same.

The problem isnt that success rate plummets on yellow and red mobs, 1 in 3 chance for a successful taunt is fine with me and most everyone else on red cons I would wager, the problem is the streaked failure rate skyrockets when the success rate plummets.

So as suggested in the other thread (here if you really want to read it http://www.thesteelwarrior.org/forum...ead.php?t=8515) I think the best way to solve this issue without a power increase is to scale taunt success rate with each failed taunt.

ie 30% success rate first taunt then if lasttaunt=kyv_eyegouger01 use taunt rate*1.5 to give a 50% success rate that taunt only. or something similar, so it only 'kicks in' if you actually failed a taunt on that same exact mob. I'm not a programmer and I dont know eq's code I just want people to get a clear picture of what I'm proposing to avoid sidetracking.

What this would accomplish is lowering the streaked failure rate without lowering the difficulty of the mobs or trivializing agro control (as if thats an issue with 2 hate weapons and 2 anger 3 augs and WA5 anyway) I could break a mez after 2-3 taunts instead of 8 etc.. Like every other mob yet I would still have a 30% success rate for snap taunts leaving that to the hybrids and their spells as it should be. Hybrids should receive the taunt change as well, their taunt button sucks same as ours does.

Oh I also parsed failure rate out of about 3 months of logs. Using the data above you could guesstimate half the mobs I fight are yellow or red to bring that success rate down to 47% making the issue at hand fairly large, its not like we run in to a yellow or red con a few times a night anymore.

14657 total taunts 7788 total failures 47% success rate

If anyone has any valid concerns or comments, please post!

[Jazun]

Has anyone discovered any insite as to why you can get a successful taunt and the mob not turn when stuns arent being used?

[Sirek Saron]

ie 30% success rate first taunt then if lasttaunt=kyv_eyegouger01 use taunt rate*1.5 to give a 50% success rate that taunt only. or something similar, so it only 'kicks in' if you actually failed a taunt on that same exact mob. I'm not a programmer and I dont know eq's code I just want people to get a clear picture of what I'm proposing to avoid sidetracking.

Would avoid me 6 fails in a row....

[NeoKlux]

succes rate on high lvl mob is a joke

thx for the parse

[Neksnappa]

One reason why a taunted mob won't turn to you is if it's a spell caster. Tank some of the giants in BoT with a group leader who has HoTT... you'll see that even if you remain on top of the list (thus be the mobs target) they will still turn and malos/nuke/etc other groupies, without ever changing from you as target.

And although I dislike getting 5+ failed taunts as much as anyone else, I doubt SOE will change it; red/yellow mobs are supposed to be much harder to taunt, and having a guaranteed success after 3-4 taunts would go a long way towards trivializing certain mobs. I'm in agreement with the majority of the other taunt success rate thread that was floating around recently.

[Timmok]

I agree the success ratio of > lvl 65 mobs suck atm, however the Omens expansion will probably make this issue moot(or at least limit its impact for exp-group purposes).

Once we all are level 70, our success rate on taunt on lvl 68 mobs should be good again. I doubt we will be exp'ing on mobs higher than 70(of course I could be wrong, havent spent time in beta).

[Phantron]

The order a mob casts a spell on isn't necessarily dependent on the aggro list. It seems to be kind of similar to rampage and sometimes it can skip the top person.

[Sirek Saron]

Originally Posted by Phantron

The order a mob casts a spell on isn't necessarily dependent on the aggro list. It seems to be kind of similar to rampage and sometimes it can skip the top person.

It's only on debuff as i noticed, Cancel magic, slow, root, snare, resist debuff and such, they allways DD on the person at the top of the aggro list.

[Paen]

Originally Posted by Neksnappa

And although I dislike getting 5+ failed taunts as much as anyone else, I doubt SOE will change it; red/yellow mobs are supposed to be much harder to taunt, and having a guaranteed success after 3-4 taunts would go a long way towards trivializing certain mobs. I'm in agreement with the majority of the other taunt success rate thread that was floating around recently.

They could just cap success rate at 60% like other mobs, like do 30% then if failed 45% then if failed a second time 60% until success, and once you got a success the failure rate would reset since you didnt have a failure the previous taunt. Even a 75% cap wouldnt change anything other than lowering frustration I dont think.

I just dont see a point in having so many consecutive failures on any mob excluding bosses, I dont rely on taunt for sustained agro, I use it for initial agro. Sure I hit it if a mob hops on a caster but I'm not counting on it to work, I count on my procs, incite, ae taunt and /shield to keep mobs off caster, not taunt since I have control over those for the most part. I definitely wouldnt intentionally try to get two failed taunts just so I could have a 60% success rate instead of a 30% success rate, maybe youre right and some people would though I dunno. If they just attacked the mob during that time they would have agro if they wanted it anyway though, and they should be using AE taunt on MT switches for bosses so I dont see how it could be exploitable.

Or if they want it to be a 1 in 3 success rate then make it a true 1 in 3, random ofcourse, but make 1 of every 3 taunts land making the most failures in a row 4. The chances of someone getting 4 failures in a row knowing that next taunt would work would be very, very slim, even if they tried. Not to mention anyone who did attempt it would die trying to get this elusive guaranteed taunt from gaining agro when they werent supposed to when they could have just maintained their position and used AE taunt when they were supposed to.

And again, the success rate is fine imo, the consecutive failure rate when the success rate lowers that much is the problem and theres not many ways to fix it without making taunt extremely overpowered with the way the game works today (hate procs mainly).

[Darkgecko]

That's some great info, I had never really seen the numbers like that. I typically only use taunt when breaking mezzes, or when a raid mob is loose. And even with chain stunning, there's plenty of times when a raid mob will smack an enchanter long enough for me to fail 5+ taunts in a row.

[Battleblade]

Paen,
Thanks for the parses. They provide a lot of insight into this area.
********
Philosophy - skip this if you only want to focus on nuts and bolts.

I wanted to point out that methods of insuring long series of skill failures don't occur can also insure long series of successes don't provide an overall increase to success percentage. That doesn't mean that such code is practical to impliment in EQ.

The question is does data nudging/feedback of previous results or pure untampered RNG provide a better simulation of a skill.

There are two competing views IMO. In one, Taunt is supposed to have a random chance of success with the chances of a "long" (from the viewpoint of some players) string of failures/successes as a natural and desired consequence.

In the other view, it's the Taunt Skill with implications like a player focusing more on their mechanics as multiple failures approach catastophe and paying less attention as multiple sucesses lull them into deviations in mechanics leading to greater likelyhood of failure in furture attempts.

The 10d die method might be acceptable in a handheld game with 1 K of object code and 256 bytes of ram, but might not be in a cutting edge multi-GByte product.

There is no denying that at least some players find the current streakiness to be a poor approximation of player skills - perhaps because you can only bet 2-4 quarters one at a time at the very slow Coin Toss Table in SoE/Vegas before your raid potentially wipes.

Please return to yer regularly scheduled statistcal analysis.

BB

[Sassinak]

The order a mob casts a spell on isn't necessarily dependent on the aggro list. It seems to be kind of similar to rampage and sometimes it can skip the top person.

That only applies to spells that add/remove icons on the target. (I think the rationale is that those are all crowd control spells.) Instantaneous spells always go on the person at the top of the aggro list.