Ok is my maths wrong? Or is maths not mathing in mcoc?

Supersha7

Fighting Titania in necropolis with deathless she hulk.

On she hulks base kit, it says defender has 8% increased chance to drop a special per trauma.

I had 80 traumas on her.

80 x 8 = 640% chance. Deathless thanos synergy prevents this ability from being affected. So it’s definitely 640% increased chance.

She had sufficient power about 3 seconds after she’d triggered last special, and plenty of chances to fire. I even dropped my guard because the timer started ticking down past half way and was prepared to eat combos or a special to keep my traumas. Traumas last 12s, but I actually had the two extenders (40% extended) up ,so they were lasting around 16.8 seconds.

So let’s say around 13 second, she refused to throw a special with 640% increased chance…..

Eh???

What??

EH?.?

If she only had 10% chance to throw, it would be a 64% chance
I she had 15% chance I would increase to 96% chance

Is my maths right? Or is there something I’ve missed?

DNA3000

Is my maths right? Or is there something I’ve missed?

It is something you missed, or rather without knowing the specifics it is something you haven't accounted for. Interestingly, its an omission Kabam itself used to make a lot, but cleaned up for the most part but you can still see the error in Wolverine's signature ability description:



So Wolverine has an 8% chance to trigger a regeneration. What does that mean, exactly? How would you calculate how many regenerations on average Wolverine would expect to get in a fight? Are the odds 8% to get one in a fight? How about two? Do you see the problem?

Saying Wolverine has an 8% chance to trigger a regeneration is saying when the game rolls dice, there's an 8% chance for those dice to come up regen. But how often does the game roll dice? Every second? Every animation frame? Once per mission? Once per day? Once over the lifetime of the game?

It is when Wolverine gets hit that he has a chance to proc the heal. 8% chance each time struck he has a chance to gain a heal. If you don't know that, the 8% number is essentially meaningless.

So when DSH has 80 traumas and is inflicting a 640% greater chance to throw specials, when is that chance anything other than zero? Does the champion roll dice to see if they are going to throw a special every 30th of a second? Every 60th of a second? What we don't know is how often the champion *tries* to throw a special. When it isn't trying the chance to throw special is zero, and 6.4 time zero is still zero.

If DSH tries to throw a special three times during a fight and the base chance to throw is 10%, then the traumas should increase that to 64% per try, and there should have been a 95% chance to see a special thrown somewhere in there (1 - (1-0.64)^3). But if for whatever reason the circumstances under which you were fighting were not giving the AI the opportunity to decide to throw, using whatever algorithm the AI is using to make those decisions, then the trauma induced special chance increase would have nothing to boost.

Unfortunately, the players currently do not have sufficient specific information about how the enemy AI state engine works to analyze that kind of situation in detail.



Note: there is one other possibility. My understanding of how typical AI state engines work is they work by weights. In other words, the AI isn't programmed with a 10% chance to do anything. It is told you have five different thing you can do, and (for example) assigns 50 to A, 40 to B, 10 to C, 5 to D, and 1 to E. These are the relative odds of doing those things. So the total weights is 106, and thus the odds of doing A are 50 in 106, or about 47%. The odds of doing E are one in 106, or about 0.9%.

When the game says traumas increase the chance to throw specials by 640% we are assuming there's some numeric percentage chance that is being increased by 640%. But it is possible it is the weighted odds that are being increased by 640%. So 10 gets increased to 64, and the odds of B happening increase from 10 out of 106 (9.4%) to 64 out of 160 or 40%. This is often done in such situations to prevent chance-saturating, creating situations where nonsensical probabilities are generated, like 10% x 64 = 640%. That's not a real chance. Weighted random systems guarantee the probabilities are always sensible. In these situations probabilities always change by less than you'd ordinarily expect, because the changes are affected by a diminishing returns effect (in fac DR itself is implemented in a very similar way).