Page 1 of 2

Regarding Demolish & blind hit bug - from Blizz

Posted: February 21st, 2014, 10:46 am
by 3wd
https://twitter.com/TheCrafticus/status ... 9464007680

1. Blizz admitted the real hit chance of Demolish is 65%.
2. Blizz claimed in 5.4.7, if hit chance will reach 0 or below, there can not be any hit.

Well, many players tested and still found the hit bug exist - especially under the blind using by the adder.

eg.
Blind - and then got hit by the demolish of the idol.
Blind - and got hit by the suicide bomb of the mech pet.

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 21st, 2014, 12:46 pm
by Dakaf
3wd wrote:https://twitter.com/TheCrafticus/status ... 9464007680

1. Blizz admitted the real hit chance of Demolish is 65%.
2. Blizz claimed in 5.4.7, if hit chance will reach 0 or below, there can not be any hit.

Well, many players tested and still found the hit bug exist - especially under the blind using by the adder.

eg.
Blind - and then got hit by the demolish of the idol.
Blind - and got hit by the suicide bomb of the mech pet.
The suicide bomb says "always hits" doesn't it?

Demolish is really annoying though! Make the tooltip match and I'll make my teams based on them!

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 21st, 2014, 8:44 pm
by 3wd
Dakaf wrote:
The suicide bomb says "always hits" doesn't it?

Demolish is really annoying though! Make the tooltip match and I'll make my teams based on them!
Explode has 100% of hit but can be affected by blind / dodge / evade and even darkness and sandstorm.

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 25th, 2014, 11:19 pm
by Se5s
your wish must have been granted, i just had demolish miss 18 times in a row in 5 battles. there's a difference between a move that's strong, a move that's weak, and a move that just passes your turn automatically. at least 3 of those were without sandstorm or any hit chance modifiers active.

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 26th, 2014, 11:20 am
by Chibimage
My 440 sample size of Demolish running at 72% hit rate disagrees with Blizzard's "admission" lol

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 26th, 2014, 3:45 pm
by 3wd
Se5s wrote:your wish must have been granted, i just had demolish miss 18 times in a row in 5 battles. there's a difference between a move that's strong, a move that's weak, and a move that just passes your turn automatically. at least 3 of those were without sandstorm or any hit chance modifiers active.
I also received many reports from other players whose having similar experience of yours.

Looks like after the hotfix, demolish hit rate really came down a lots.

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 26th, 2014, 6:43 pm
by Genome
I just fought a team with an Idol 3 straight times. Not only did the Idol solo my entire DoT team all three times (I didn't just flee so that I could see how Demolish was working as I haven't really battled much since the patch), but Demolish never missed back-to-back in any of the fights with SS up. NOTHING has changed.

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 26th, 2014, 9:43 pm
by Hootstwo
Well, I haven't been using demolish, but I use crush regularly on Pandaren trainers. I've NEVER had a night where so many crushes have missed, it's been infuriating. I can't speak to demolish being nerfed, but man oh man, I'd say crush was! (Not that it had a spectacular hit rate to begin with!)

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 26th, 2014, 10:04 pm
by Liopleurodon
My understanding was that Demolish wasn't changed, because miss streak protection is fairly 'baked in' to the minigame. The change post-hotfix (ie, yesterday after reset) which was supposed to be in for 5.4.7 was that Demolish's miss streak protection was so good that it overwrote the -100% accuracy debuff from the poison. I'm trying to test it now, but so far it looks like the most recent hotfix worked. It's kind of impossible to prove a rarely occurring incident when there are snakes biting your face off though.

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 26th, 2014, 10:06 pm
by Jerebear
I've been using Crush all night on my Idol without any noticeable number of misses. Still seems to hit the same on average for me. I've never really used demolish. I always felt the idea of it wast terrible, and when I heard it was bugged, I didn't want to take advantage of that in PvP.

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 27th, 2014, 8:02 am
by Luciandk
From the latest hotfixes:

Pet Battle

Resolved an issue that was causing Battle Pet abilities with a high miss chance to hit more often than intended.

Seems they realized the miss streak protection was making demolish ridiculous.

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 27th, 2014, 8:20 am
by Monekop
RNG

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 27th, 2014, 1:38 pm
by Awq
Resolved an issue that was causing Battle Pet abilities with a high miss chance to hit more often than intended.
That is sad news. I considered Demolish/Belly Slide similar abilities too weak for personal use, but I did like that they were available. They were too unreliable. Yes, I lost if someone managed to chain many demolishes in a row, but this did not happen frequently.

I've tried using Belly Slide in a Christmas team and I've used the Robo Cub. Reliable pets always outperformed these RNG pets in win rating. Infected Bear Cub outperformed Robo Cub in almost every situation (despite providing better coverage).

I do like if there are abilities/pets that people could play when they have a low win rating. I do not mind it if these pets hit quite hard or slightly more often than intended.
Maybe I am biased. I was quite looking forward to Elementals countering Anubisath!

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 27th, 2014, 5:22 pm
by Genome
Awq wrote:
That is sad news. I considered Demolish/Belly Slide similar abilities too weak for personal use, but I did like that they were available. They were too unreliable. Yes, I lost if someone managed to chain many demolishes in a row, but this did not happen frequently.

I've tried using Belly Slide in a Christmas team and I've used the Robo Cub. Reliable pets always outperformed these RNG pets in win rating. Infected Bear Cub outperformed Robo Cub in almost every situation (despite providing better coverage).

I do like if there are abilities/pets that people could play when they have a low win rating. I do not mind it if these pets hit quite hard or slightly more often than intended.
Maybe I am biased. I was quite looking forward to Elementals countering Anubisath!
While I see what you are saying, the Idol just has too much going for it to be allowed free-reign with an ability like Demolish. This was actually posted in a locked thread earlier this week, but long story short, Idol is very much out-of-control right now, and this has to do with the hit rate of Demolish. He is the best defensive team-pet in the game, coupled with very high health. Add Demolish not missing like it should and he has also become one of the prime damage-dealing pets in the game as well, with zero cooldown issues using said ability, and also makes the "negative" aspect of Sandstorm moot for the Idol player. In essence, he can go toe-to-toe and win against many pets that he shouldn't be able to.

If players with "low-rating" or mediocre players were the only ones using him, what you say about him being a "level-the-playing-field" type of pet may be true, but he is not being utilized by just these types of players. Put him in the hands of a skilled player and I'm sure you can see the problem :).

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 28th, 2014, 9:36 am
by Awq
Genome wrote: While I see what you are saying, the Idol just has too much going for it to be allowed free-reign with an ability like Demolish. This was actually posted in a locked thread earlier this week, but long story short, Idol is very much out-of-control right now, and this has to do with the hit rate of Demolish. He is the best defensive team-pet in the game, coupled with very high health. Add Demolish not missing like it should and he has also become one of the prime damage-dealing pets in the game as well, with zero cooldown issues using said ability, and also makes the "negative" aspect of Sandstorm moot for the Idol player. In essence, he can go toe-to-toe and win against many pets that he shouldn't be able to.

If players with "low-rating" or mediocre players were the only ones using him, what you say about him being a "level-the-playing-field" type of pet may be true, but he is not being utilized by just these types of players. Put him in the hands of a skilled player and I'm sure you can see the problem :).
I get what you are saying! And I agree to some extent.

I feel that Demolish might specifically be a problem on Anubisath idol. 50% hit abilities generally do not influence battles on pets other than the Anubisath idol. It would be more logical to change the Idols move set/breed/stat allocation, unless you feel that the 50% (65%) ability is a problem on the other pets as well. One could replace Demolish with Punch for example.

I do not think that these nerfs were needed, at least to this extent.
I would rather see a damage increase on Demolish-like abilities rather than an incorrect tool tip though. While I am not a fan of the 5.3 hit changes, I do feel that Demolish could scale better. It current does the same expected damage as 100% moves, there is no reward for choosing RNG unlike with other accuracy losses. I would not mind to see 40 Base Attack on 55% abilities. 80% abilities have 22.4 expected damage, I do not understand why a 50% ability should have lower expected damage. It currently does the same damage as bite, with the handicap of unpredictability!

There is no need for 26 (65%) or 30 (75%) expected damage though!

50% hit abilities are stuck with pre-5.3 damage. They have not received a buff in expected damage, unlike other RNG abilities.

Edit inspired by Poofah's explanation:
The problem that I am having with 50% moves is that there is no incentive to gamble, unlike with other RNG moves. It does 20 average damage, the same damage that Bite does. Once you get below 75% the rewards for gambling are diminished, which in my opinion is quite illogical. I believe that the Basic Attack should increase by more than 2 when you get below 75% accuracy.

100% accuracy gives 22 base dmg (20 avg)
95% accuracy gives 22 base dmg (20.9 avg)
90% accuracy gives 24 base dmg (21.6 avg)
85% accuracy gives 26 base dmg (22.1 avg)
80% accuracy gives 28 base dmg (22.4 avg)
50% accuracy gives 30 base dmg (20.0 avg)

The base damage gets increased by 2 for every 5%. 75% provides the biggest incentive for gambling. The benefit of gambling becomes smaller past 75%, because the graph is symmetrically shaped. There is no additional incentive for gambling at 50%

This is all hypothetical:
100% accuracy gives 22 base dmg (20 avg)
95% accuracy gives 22 base dmg (20.9 avg)
90% accuracy gives 24 base dmg (21.6 avg)
85% accuracy gives 26 base dmg (22.1 avg)
80% accuracy gives 28 base dmg (22.4 avg)
75% accuracy gives 30 base dmg (22.5 avg)
70% accuracy gives 32 base dmg (22.4 avg)
65% accuracy gives 34 base dmg (22.1 avg)
60% accuracy gives 37 base dmg (22.2 avg, instead of 21.6)
55% accuracy gives 40 base dmg (22.0 avg, instead of 20.9)
50% accuracy gives 45 base dmg (22.5 avg; instead of 20)

Because of the bug 50% abilities were hitting more often, this resulted in an increase of avg dmg increased. They might have done 26 for 65% hit rating or 30 for 75% hit rating. These numbers are not in proportion with other average hits. I however find it quite illogical that the rewards for gambling diminish from a practical point of view, to the point where this benefit gets removed. Why would you discourage RNG once you get below a certain hit %?
I would like a solution in between these two numbers.
If you make the scaling less symmetrical, the average damage would stay the same after the 75% mark. The benefits of gambling would stay similar after this point. If you give 55% hit moves 40 base attack, the average damage would be in proportion with other RNG moves.

Note:
I would propose 55% accuracy with 40 base attack. 50% accuracy with 45 base attack might make this move too strong in combination with Uncanny Luck.

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 28th, 2014, 12:53 pm
by Badpathing
Awq wrote:Lets say that you play a team that wins 90% or 95% of the matches. This is not unreasonable, I play multiple teams with such win rates.
Awq..your post....just....*shakes head*

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 28th, 2014, 2:57 pm
by Awq
Badpathing wrote:
Awq wrote:Lets say that you play a team that wins 90% or 95% of the matches. This is not unreasonable, I play multiple teams with such win rates.
Awq..your post....just....*shakes head*
I read it again. I probably should have left that first part out of my post ^^ I came across in a way that was not intended, it distracts from the point that I am trying to make. I will edit my post! Thank you for the feedback.

I find it too difficult to explain why RNG moves have higher expected damage (except 50% abilities). I do not know enough factors and I am terrible at explaining it. RNG does interact with winratio.

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 28th, 2014, 3:08 pm
by Badpathing
Awq wrote:
Badpathing wrote:
Awq wrote:Lets say that you play a team that wins 90% or 95% of the matches. This is not unreasonable, I play multiple teams with such win rates.
Awq..your post....just....*shakes head*
I read it again. I probably should have left that first part out of my post ^^ I came across in a way that was not intended, it distracts from the point that I am trying to make. I will edit my post! I do play a lot of fun teams as well! Thank you for the feedback.
My comment was supposed to be in good humor. I see a lot of posts claiming those sorts of %s, and while I myself enjoy pretty good win ratios, I seriously doubt anyone can fabricate a team with those sorts of results. At least, not any team you are using regularly.

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 28th, 2014, 3:23 pm
by Awq
Badpathing wrote: My comment was supposed to be in good humor. I see a lot of posts claiming those sorts of %s, and while I myself enjoy pretty good win ratios, I seriously doubt anyone can fabricate a team with those sorts of results. At least, not any team you are using regularly.
I know. But you were right, despite joking!
A lot of people would dismiss my arguments because of how I started my posts.
Using statistical effects are not a good argument either, not when there are better arguments in favor for a damage buff.

You can. At least on my BG. It would not be possible if there was more variety. People sadly do not use a wide variety of teams. One often encounters the same enemies over and over again. If you can form a team that has good self synergy and that does well against most FotM pets, you will win a lot.
I generally test good teams for 50 matches. These teams won't stay great for long though, because the meta game keeps evolving :P
My winratio was a lot lower in 5.3. The current meta game is a lot less luck based without abilities like Moth Dust and Horn Attack dominating the meta game.

Re: Regarding Demolish & blind hit bug - from Blizz

Posted: February 28th, 2014, 4:20 pm
by Poofah
Awq wrote:I find it too difficult to explain why RNG moves have higher expected damage (except 50% abilities). ... RNG does interact with winratio.
In short, it is to promote gambling, and therefore give an advantage to the underdog. Look at it this way:

suppose we have player A and player B, playing a game with no RNG. Player A is better at the game and wins 90% of the time; player B only wins 10% of the time.

Now suppose we add a new ability called win20 to the game: roll a D5, on a 1 you win, on 2-5 you lose. Objectively this is a pretty bad ability, and most people wouldn't use it. Player A certainly would never use it. However, player B only wins 10% of the time anyway. So player B will use this ability every game and his win % will go up to 20%.

This is what the low-accuracy attacks are, albeit their effect is more subtle and plays out over a longer term. I can accept risk to gain a small but significant advantage in terms of damage-per-turn. If I'm winning 90% of the time anyway, I probably wouldn't like to accept any significant risk to gain extra damage. But if I'm winning 10% of the time, then I have every reason to gamble.

100% accuracy attacks are 20 base dmg per turn; 95% accuracy gives 22 base dmg (20.9 avg); 90% accuracy gives 24 base dmg (21.6 avg), etc. If 20 base dmg is 'fair' (ie 20 base dmg is win50), then 95% accuracy attacks are win52, 90% accuracy attacks are win54, etc. We are being given an advantage if we're willing to gamble; and the lower my win percentage, the more willing I should be to gamble. With the current system, it makes sense to gamble even if you think your chances are 50%.