Mission completion %, how does it work?

[Deleted]

It's per mob and per follower, so 9 rolls for a mission with 3 followers vs 3 mobs.

[Delaios]

It's per mob and per follower, so 9 rolls for a mission with 3 followers vs 3 mobs.

What irrefutable evidence do you have that supports this? Until there is confirmation on this, by either a dev or from a massive collection of mission data, I'll stick with what makes the most sense: If a mission states it has a 70% success chance, it actually has a 70% chance to succeed instead of 70%^(amount of counters).

[Amerissis]

No, we do know. Its been confirmed in a number of places.

Its 86% chance per mob. So 3 mobs = 86%, 86%, 86%. Three separate rolls. Anything under 90% is taking a gamble on 3 mobs. Thats not to say you wont win. You might win every damn time, but its still a gamble. Thats why people say.. if it isnt 100%, fix it.

1 Mob	2 Mobs	3 Mobs
100	100	100
95	90	86
90	81	73
85	72	61
80	64	51
75	56	42

That is utter nonsense. It is 86% per full mission. They only show per mob because it's more engaging. Even without any evidence, logic already dictates that 86% chance of success is 86% chance of success and not 64%.

[chazus]

I noticed people were getting bent out of shape about the whole proof thing. I wanted to clarify a couple things.

Firstly, I didn't feel the need to 'prove' it largely because it didn't affect me. Proving it doesn't help me in any way, and it wasn't (at the time) really worth my time to do my research all over again. At the time (a month or so back) I spoke with a number of people, some whom -had- done several hundred iterations, as well as mod API makers who can get this kind of information more reliably, and we had determined that it was per-mob. I was happy with that answer, as it had come from (again, at the time), a reliable source.

Since people got all upset that I didn't have proof, I really sat down and did some serious investigative work. I googled "garrison mission success math" and the third hit was a post from Simca, whom is in touch with API people and very versed with both the garrison devs and community. I spoke with Simca further on it, and it turns out that it only is one roll (according to that) and the "each follower doing something" is entirely aesthetics. It does not however, explain the math involved as to 'who' fails the mission.

So, there you have it. I was wrong. And I was happy with my answer before because it was reliable (at the time). And now I have a new one (that is also reliable).

Notably, the first hit on google is just math on how to get 100%. The second hit on google is a thread showing that it's 3 rolls. Not that google=truth but still curious. So... take that as you will.

[Chaltione]

One thing I wonder is why the percentage changes for the other two mobs/followers if I change one of the followers in a 3 party vs 3 mob mission.

Example:
I have listed 80% chance.
Now assume that it's 3 rolls.
I upgrade the itemlevel of follower #1.
The listed chance is now 85%.

Why is roll #2 and roll #3 changed?

[Amerissis]

One thing I wonder is why the percentage changes for the other two mobs/followers if I change one of the followers in a 3 party vs 3 mob mission.

Example:
I have listed 80% chance.
Now assume that it's 3 rolls.
I upgrade the itemlevel of follower #1.
The listed chance is now 85%.

Why is roll #2 and roll #3 changed?

Because it's just one roll.

[Ryme]

If it's every roll that's incredibly misleading to players. A mission that claims 50% chance actually means 12.5% chance, I don't think Blizzard would design it to be intentionally misleading.

[aggixx]

It does not however, explain the math involved as to 'who' fails the mission.

Well if I were to program an aesthetic, I would first check if the mission succeeds or fails, and then if it fails RNG 1-3 (or however many "conflicts" there are) and that would be the one that fails it. Doing it any other way is just overly complicating something that doesn't need to have any statistical significance.

[Simulacrum]

So, there you have it. I was wrong. And I was happy with my answer before because it was reliable (at the time). And now I have a new one (that is also reliable).

No, just like it isn't reliable now, it never was reliable, and what gets people "bent out of shape" is the very duchy, passive-aggrssive way in which you defended your blatantly obviously wrong assertion, all the while pretending to authoritatively spread your ignorant, mis-informed opinion.

[Deleted]

No, just like it isn't reliable now, it never was reliable, and what gets people "bent out of shape" is the very duchy, passive-aggrssive way in which you defended your blatantly obviously wrong assertion, all the while pretending to authoritatively spread your ignorant, mis-informed opinion.

Come to think of it, it shouldn't be too hard to prove it at all. Statistically speaking ofcourse, you'd need the coding for 100% certainty.
I've been leveling some alts and their garrisons, and 3 followers on a lvl 100 quest with 3 counters required (those 8k exp missions for example) seem to give 25% chance. Now if it was 3 rolls of 25%, you'd be down to 1.56% chance total. It would take a very very small sample to have a 99.9% confidence interval to disprove the 1.56% chance. If you do this - say - 20 times and you get 3 or more success then you're probably already in the 99%+ range chance of it NOT being 1.56%. Cba to do the exact math but the difference between 25% and 1.56% should be really easy to show.

Edit: rough math shows me that chance of getting 3+ success on 20x 1.56% chance is a mere 0.5%. If you'd get 4/20 then you're looking at 0.1% or lower already.