we will see what happens when the update goes live. im sure she will be dumped rather quickly by most non FO teams. My real message in the previous post was that a lot of people were complaining that it called an assist too often, and all the phasma defenders were acting like we were crazy for saying so. now it comes out that it WAS calling too often, but we have a "test" here that shows it was procing correctly down to less than 1% accuracy. so how can the data be off that far if the test included all team actions?
Because the OP was calculating tghe % based on all actions (as described in the OP), not all attacks. Everybody knew that Phasma was proccing when healing or other non-attacking actions, so, until the Devs said that wasn't intended, the proper method was to divide the total number of procs, by the total number of actions.
Yes fine but, 16% of attacks in a arena match may be 2 assists. 16% of all actions in a Match may be 4 to 6. So if it was 16% of all actions it was still too high because there are many more actions than attacks.
we will see what happens when the update goes live. im sure she will be dumped rather quickly by most non FO teams. My real message in the previous post was that a lot of people were complaining that it called an assist too often, and all the phasma defenders were acting like we were crazy for saying so. now it comes out that it WAS calling too often, but we have a "test" here that shows it was procing correctly down to less than 1% accuracy. so how can the data be off that far if the test included all team actions?
Because the OP was calculating tghe % based on all actions (as described in the OP), not all attacks. Everybody knew that Phasma was proccing when healing or other non-attacking actions, so, until the Devs said that wasn't intended, the proper method was to divide the total number of procs, by the total number of actions.
Yes fine but, 16% of attacks in a arena match may be 2 assists. 16% of all actions in a Match may be 4 to 6. So if it was 16% of all actions it was still too high because there are many more actions than attacks.
Yeah, because it was and still is bugged. Now it will be fixed in the next update.
My name is cosmicturtle333, aka CT-333, aka Threes.
I find it really funny how everyone ignored this thread for two weeks and now that she is having her bug fixed everyone's lost their minds.
Just to be clear:
Captain Phasma
Phasma's Leader ability had bugs that frequently allowed the bonus Assist attacks to trigger off of non-attack abilities, sometimes even targeting allies instead of enemies. It is unknown to what degree this influenced Phasma's effectiveness, so we will observe the effects of this fix before considering any future balance adjustments.
Fixed a bug that allowed Phasma's Leader ability to trigger on non-attack abilities
Fixed a bug that could cause Phasma's Leader ability Assists to target allies instead of enemies
My name is cosmicturtle333, aka CT-333, aka Threes.
People are "losing their minds" because you and a lot of others claimed 100% there was nothing wrong and refused to consider the other side. Anyone who said so was wrong, they were wrong to question you because you did the math in a flawed GW match where the seed numbers never change, just the actions taken with those numbers change.
People are "losing their minds" because you and a lot of others claimed 100% there was nothing wrong and refused to consider the other side. Anyone who said so was wrong, they were wrong to question you because you did the math in a flawed GW match where the seed numbers never change, just the actions taken with those numbers change.
I battled her multiple times using a different leader and squad arrangement so as to not repeat battles and here's what I got.
If you read the OP and understood how GW battle seeds are determined you'll understand what this means and how you're wrong, again.
My name is cosmicturtle333, aka CT-333, aka Threes.
Yes, and as I said 16% of every attack is significantly less than 16% of every ACTION. You might attack 15 times in a match, but you will have 30 actions. 16% of 30 is greater than 16% of 15. So it was procing more often.
As for the seed numbers, they don't change. When an action needs to be taken it uses the next number in the sequence. If the first number in the sequence is 42 then it's used. It doesn't matter if you use it for attack or heal, it is used and we go to the next number for the next action.
Yes, and as I said 16% of every attack is significantly less than 16% of every ACTION. You might attack 15 times in a match, but you will have 30 actions. 16% of 30 is greater than 16% of 15. So it was procing more often.
Yeah. So go talk about that in a bug thread or create a new thread for it. This thread is about the proc rate at the time of testing it. If it's bugged then it's bugged.
This thread wasn't about whether or not the ability wasn't working properly, it was about whether or not she it was proccing more than 16%. It was about how people would say "Phasma procs 60% of the time!!! Nerf her!!!"
My name is cosmicturtle333, aka CT-333, aka Threes.
As for the seed numbers, they don't change. When an action needs to be taken it uses the next number in the sequence. If the first number in the sequence is 42 then it's used. It doesn't matter if you use it for attack or heal, it is used and we go to the next number for the next action.
The seed is re-rolled each time you change leader or squad members. You clearly don't know what you're talking about.
My name is cosmicturtle333, aka CT-333, aka Threes.
Yes, and as I said 16% of every attack is significantly less than 16% of every ACTION. You might attack 15 times in a match, but you will have 30 actions. 16% of 30 is greater than 16% of 15. So it was procing more often.
Yeah. So go talk about that in a bug thread or create a new thread for it. This thread is about the proc rate at the time of testing it. If it's bugged then it's bugged.
This thread wasn't about whether or not the ability wasn't working properly, it was about whether or not she it was proccing more than 16%. It was about how people would say "Phasma procs 60% of the time!!! Nerf her!!!"
You are funny. If it procs on twice as many actions then guess what? It's double the percent. 32% proc on rate of attack.
As for the seed numbers, they don't change. When an action needs to be taken it uses the next number in the sequence. If the first number in the sequence is 42 then it's used. It doesn't matter if you use it for attack or heal, it is used and we go to the next number for the next action.
The seed is re-rolled each time you change leader or squad members. You clearly don't know what you're talking about.
The seed is predetermined before the match starts.
As for the seed numbers, they don't change. When an action needs to be taken it uses the next number in the sequence. If the first number in the sequence is 42 then it's used. It doesn't matter if you use it for attack or heal, it is used and we go to the next number for the next action.
The seed is re-rolled each time you change leader or squad members. You clearly don't know what you're talking about.
The seed is predetermined before the match starts.
When you retreat and change leader the seed changes. Just search 'seed' and you will see this has been mentioned many times before.
My name is cosmicturtle333, aka CT-333, aka Threes.
"That's not quite how it works. For a computer to make random numbers it needs a "seed." The time in seconds from a set time is a common seed because it will never be the same twice (unless you get the seed more than one time per second ). This game uses your team composition to create the seed. So the same team will give the same seed every time. This was probably intentional. AFAIK, this has been confirmed.
I'm not a developer for the game, so this isn't official, but here's how I assume the game works. "
So you go on this guys sssumption?
The AI chooses the target and action. First action of the match isnt set in stone. If the target is a different toon they may choose to special someone else rather than attack whoever replaced JC. Then the seed is all out of whack just from the first move. It can use the exact same sequence and give totally different results.
Nice try making this about RNG rather than proc rate though.
@Widget_gaget
Yeah, I do go on that guys assumption.
I also go by the fact that I used almost every single hero I have available including level 1 heroes, tested this with each squad available to get the seed different every time.
How many times have you retreated on a match? How many times have you seen the exact same thing happen if you switch leaders? Because in the hours of testing I was doing not once did I ever see the same action taken when using different leaders and squad combinations.
Go ahead, go do a galactic war battle, retreat, switch leader and switch your squad to new heroes and you can clearly see that the seed changes every time.
I'm not going to reply to you again. So post whatever you want.
My name is cosmicturtle333, aka CT-333, aka Threes.
we will see what happens when the update goes live. im sure she will be dumped rather quickly by most non FO teams. My real message in the previous post was that a lot of people were complaining that it called an assist too often, and all the phasma defenders were acting like we were crazy for saying so. now it comes out that it WAS calling too often, but we have a "test" here that shows it was procing correctly down to less than 1% accuracy. so how can the data be off that far if the test included all team actions?
Because the OP was calculating tghe % based on all actions (as described in the OP), not all attacks. Everybody knew that Phasma was proccing when healing or other non-attacking actions, so, until the Devs said that wasn't intended, the proper method was to divide the total number of procs, by the total number of actions.
Yes fine but, 16% of attacks in a arena match may be 2 assists. 16% of all actions in a Match may be 4 to 6. So if it was 16% of all actions it was still too high because there are many more actions than attacks.
I'm not sure I'm following you. If you are suggesting that the % of extra attack in non-attack actions was above average, then I think that's just confirmation bias. Surely didn't feel like that for me. Would love to see any numbers that disprove that.
If that's not what you are saying, then it's irrelevant. Let's do it with an easy-to-follow example. Let's say 1/3 of the 260 actions in the OP were non-attacks. They caused 1/3 of the 43 extra attacks. If those wouldn't exist, you would have had about 28 hits in about 172 attacks, and the ratio would had been 16% regardless. It doesn't affect the numbers at all. It would, if he had used total number of extra attacks (43) and compared it with number of attack actions (172 in this example), but that's not what OP did.
Also: there are way more attacks than non-attack actions. By far. In a 5 turn combat, Phasma uses a single non-attack action (March Orders), vs 4 attack actions (1 fussillade, and 3 normal attacks). Luminara uses 1 heal, vs 1 force blast and 3 attacks. Poggle uses 1 buff, and 4 normal attacks. And so on. Many characters (Qi Gon Jin, FOTP, Geo Soldier, Sid, etc) don't even have options for non-attacks. And those who have options, have cooldowns
Yes, and as I said 16% of every attack is significantly less than 16% of every ACTION. You might attack 15 times in a match, but you will have 30 actions. 16% of 30 is greater than 16% of 15. So it was procing more often.
Yeah. So go talk about that in a bug thread or create a new thread for it. This thread is about the proc rate at the time of testing it. If it's bugged then it's bugged.
This thread wasn't about whether or not the ability wasn't working properly, it was about whether or not she it was proccing more than 16%. It was about how people would say "Phasma procs 60% of the time!!! Nerf her!!!"
You are funny. If it procs on twice as many actions then guess what? It's double the percent. 32% proc on rate of attack.
What? if it procs twice as much in twice as many actions, the proc rate is the same.
Yes, and as I said 16% of every attack is significantly less than 16% of every ACTION. You might attack 15 times in a match, but you will have 30 actions. 16% of 30 is greater than 16% of 15. So it was procing more often.
But.... it's still 16%. The proc rate was, and still is, exactly what said in the tooltip. It was being applied to non-attack actions (something we all knew was happening, just we didn't knew it was't intended). This thread was made to disprove the false notion that AI % was higher. It wasn't. It's the same than player %. It was, of course, being applied to non attacks. So was the Player's proc rate. Players were also proc'ing from healings and buffs and taunts.
But the % was correct, it was the % shown in the tooltip, and there wasn't a disparity between the AI percentage and the player's percentage.
Yes @Triqui . But if you have any assist from a non attack it makes the assist per attack increase. The description says assist on attack. If the assist is on a non attack it just made the proc % go up. So you had 10 attacks and 2 proced. Also had 3 proc on non attack. That's 5 procs on 10 attacks. 50% assists as a percent of attacks. Yes?
Yes @Triqui . But if you have any assist from a non attack it makes the assist per attack increase. The description says assist on attack. If the assist is on a non attack it just made the proc % go up. So you had 10 attacks and 2 proced. Also had 3 proc on non assist. That's 5 procs on 10 attacks. 50% assists as a percent of attacks.
I don't really know how to explain it further for you to understand. It doesn't matter, it doesn't affect the outcome of the experiment. In your example, that's 20% proc on attacks, and 25% proc on actions (assuming the other 3 procs come from 10 non-attacking actions, for a 30% proc ration in non-attacks, and 25% overall for your example).
The OP didn't record attacks, he recorded actions. He included all procs, but divided them by all actions. His experiment is ok, and worked well, and shown the same that the devs claim.The proc rate was correct, 16%. It was also the same for players and non players. It was being applied to non-attacks, for both players and non players. It still makes the point moot, and the claim that Phasma's ablity was over-heated for the AI a false assumption based on confirmation bias.
I'll try one more time. If you only attack 10 times and you have a total of 5 assists called how many assists were called as a percent of attack? 5 per 10 attacks. 50%. it doesn't matter if it was an attack or non attack that triggered it is still 5 per 10 attacks.
I'll try one more time. If you only attack 10 times and you have a total of 5 assists called how many assists were called as a percent of attack? 5 per 10 attacks. 50%. it doesn't matter if it was an attack or non attack that triggered it is still 5 per 10 attacks.
I'd try once more. That's not what the OP did, or recorded, was not what the thread was about. You are wrong, no matter how stubborn you are repeating it. The proc rate was correct at 16%, and it was the same for the AI than the players. It was proc'ing from incorrect sources, for both he AI and the players. The OP methodology was correct, and the experiment got the expected results for it's methodology.
Yes, it was 5 procs per 10 attacks. But it was 5 procks per 20 actions, and the OP was recording actions. The 16% figure was correct, was in line with what was expected, and disproved the false notion that AI has a better proc rate, just like the Devs own (and far more extensive) testing did.
The description says 16% of attacks. Can we agree on that? 16% of 10 attacks is 1.6 assists. Anything over that is more than intended. If you get 5 total it is more that 16% PER ATTACK. the character description has nothing to do with total actions, only total attacks. If only 32 of the 242 total actions were attacks then 43 assists would be over 100% assist on attack.
The description says 16% of attacks. Can we agree on that? 16% of 10 attacks is 1.6 assists. Anything over that is more than intended. If you get 5 total it is more that 16% PER ATTACK. the character description has nothing to do with total actions, only total attacks. If only 32 of the 242 total actions were attacks then 43 assists would be over 100% assist on attack.
I give up. You are more stubborn than me. You are wrong, but more stubborn.
Thanks for putting in the work Turtle, if I can get the time I might try and and add some data to the pool, I'll try and compare my Phasma with AI Phasma as well so there is a comparison on the same seed.
To all the haters learn to read, learn the rules of engagement as confirmed by devs and pay attention to the math.
If you think AI proc rates are actually bugged(I don't think they are) then do your own research, get a screen recorder for proof if you really care that much, otherwise keep you're conspiracy theories for the UFO sites
Replies
Yes fine but, 16% of attacks in a arena match may be 2 assists. 16% of all actions in a Match may be 4 to 6. So if it was 16% of all actions it was still too high because there are many more actions than attacks.
Yeah, because it was and still is bugged. Now it will be fixed in the next update.
Just to be clear:
If you read the OP and understood how GW battle seeds are determined you'll understand what this means and how you're wrong, again.
I didn't have to take that bug into consideration as I was counting every action taken, not every attack used.
Yeah. So go talk about that in a bug thread or create a new thread for it. This thread is about the proc rate at the time of testing it. If it's bugged then it's bugged.
This thread wasn't about whether or not the ability wasn't working properly, it was about whether or not she it was proccing more than 16%. It was about how people would say "Phasma procs 60% of the time!!! Nerf her!!!"
The seed is re-rolled each time you change leader or squad members. You clearly don't know what you're talking about.
You are funny. If it procs on twice as many actions then guess what? It's double the percent. 32% proc on rate of attack.
The seed is predetermined before the match starts.
When you retreat and change leader the seed changes. Just search 'seed' and you will see this has been mentioned many times before.
I'm not a developer for the game, so this isn't official, but here's how I assume the game works. "
So you go on this guys sssumption?
The AI chooses the target and action. First action of the match isnt set in stone. If the target is a different toon they may choose to special someone else rather than attack whoever replaced JC. Then the seed is all out of whack just from the first move. It can use the exact same sequence and give totally different results.
Nice try making this about RNG rather than proc rate though.
Yeah, I do go on that guys assumption.
I also go by the fact that I used almost every single hero I have available including level 1 heroes, tested this with each squad available to get the seed different every time.
How many times have you retreated on a match? How many times have you seen the exact same thing happen if you switch leaders? Because in the hours of testing I was doing not once did I ever see the same action taken when using different leaders and squad combinations.
Go ahead, go do a galactic war battle, retreat, switch leader and switch your squad to new heroes and you can clearly see that the seed changes every time.
I'm not going to reply to you again. So post whatever you want.
I'm not sure I'm following you. If you are suggesting that the % of extra attack in non-attack actions was above average, then I think that's just confirmation bias. Surely didn't feel like that for me. Would love to see any numbers that disprove that.
If that's not what you are saying, then it's irrelevant. Let's do it with an easy-to-follow example. Let's say 1/3 of the 260 actions in the OP were non-attacks. They caused 1/3 of the 43 extra attacks. If those wouldn't exist, you would have had about 28 hits in about 172 attacks, and the ratio would had been 16% regardless. It doesn't affect the numbers at all. It would, if he had used total number of extra attacks (43) and compared it with number of attack actions (172 in this example), but that's not what OP did.
Also: there are way more attacks than non-attack actions. By far. In a 5 turn combat, Phasma uses a single non-attack action (March Orders), vs 4 attack actions (1 fussillade, and 3 normal attacks). Luminara uses 1 heal, vs 1 force blast and 3 attacks. Poggle uses 1 buff, and 4 normal attacks. And so on. Many characters (Qi Gon Jin, FOTP, Geo Soldier, Sid, etc) don't even have options for non-attacks. And those who have options, have cooldowns
But.... it's still 16%. The proc rate was, and still is, exactly what said in the tooltip. It was being applied to non-attack actions (something we all knew was happening, just we didn't knew it was't intended). This thread was made to disprove the false notion that AI % was higher. It wasn't. It's the same than player %. It was, of course, being applied to non attacks. So was the Player's proc rate. Players were also proc'ing from healings and buffs and taunts.
But the % was correct, it was the % shown in the tooltip, and there wasn't a disparity between the AI percentage and the player's percentage.
Don't bother. He is either trolling or doesn't understand how simple game mechanics work. Either way he's not worth your time.
I don't really know how to explain it further for you to understand. It doesn't matter, it doesn't affect the outcome of the experiment. In your example, that's 20% proc on attacks, and 25% proc on actions (assuming the other 3 procs come from 10 non-attacking actions, for a 30% proc ration in non-attacks, and 25% overall for your example).
The OP didn't record attacks, he recorded actions. He included all procs, but divided them by all actions. His experiment is ok, and worked well, and shown the same that the devs claim.The proc rate was correct, 16%. It was also the same for players and non players. It was being applied to non-attacks, for both players and non players. It still makes the point moot, and the claim that Phasma's ablity was over-heated for the AI a false assumption based on confirmation bias.
I didnt insult you, please don't insult me. I've been very civil to you.
@FashionFett
Yes, it was 5 procs per 10 attacks. But it was 5 procks per 20 actions, and the OP was recording actions. The 16% figure was correct, was in line with what was expected, and disproved the false notion that AI has a better proc rate, just like the Devs own (and far more extensive) testing did.
Thank you
I give up. You are more stubborn than me. You are wrong, but more stubborn.
Thanks for putting in the work Turtle, if I can get the time I might try and and add some data to the pool, I'll try and compare my Phasma with AI Phasma as well so there is a comparison on the same seed.
To all the haters learn to read, learn the rules of engagement as confirmed by devs and pay attention to the math.
If you think AI proc rates are actually bugged(I don't think they are) then do your own research, get a screen recorder for proof if you really care that much, otherwise keep you're conspiracy theories for the UFO sites