I agree somewhat with what you’re saying. In none of the 10 attempt stretches that you calculated did you have a rate lower than 40%. However, the point where you started counting was arbitrary. If you shifted your strings of 10 attempts from your 41st to 50th, 51st to 60th, etc., to the 45th to 54th, 55th to 64th, etc., you might get different results. In fact, you would gett different results, I just can’t say what those results would look like.
That kind of analysis tends to not be very helpful for anything since it only illustrates that random results are random, and further subdividing an already small sample won’t help matters.
Increasing your sample to some larger value might see your average continue to drop, or it might not. That’s the thing with random. 19 times out of 20, a sample of X will be representative of the population, but yours might be that 20th sample that doesn’t work.
That is exactly why I said I didn’t draw conclusions from these results and that more results from more persons would help. I simply said that my 50 last events proved to be consistent.
That’s what 63 hits out of 110 attempts, with an ostensible 40% hit chance, looks like. You’re way on the right side - like, way, way on the right side. 0.02% of getting 63 or more hits means that it’s certainly possible, but pretty sketchy. You “should probably” have gotten between 34 and 55 hits (± 2σ).
It’s not enough to say something’s wrong, but it’s certainly enough to say something seems weird.
That’s cool. More people doing more samples will either support the hypothesis that the probability is 40%, or it won’t. There have been a few times over the last couple years that people doing statistical tracking of game results showed without much doubt that there was a bug in the game, which was subsequently resolved. Your data on its own is not convincing to me, but if it inspires other people to rigorously track their own results, and that data also shows that devour rates are consistently greater than 40%, then it will be worthwhile.
Well i already suggested more data and the op said " the % was never under 40 so it can never be 40" thus alleviating himself of any further testing… I suspect because he thinks im right and is lothe to prove me so
I don’t really see the point of coming here with false numbers. Anyone can gather numbers by its own, that would be stupid. And by the way, let’s don’t move too far from my first post. I’m the first one to ask for more numbers. I’m not saying mine are true, just saying they’re showing something. I’m not even complaining. Want to add something to the discussion, then go gather you own numbers so that I’ll also be able to be not entirely confident in the data collection. Don’t let the discussion go in a direction it’s not supposed to go just because of that guy.
No, you still don’t get it, you’re saying useless things since the beginning. I’m the first one who asked for more data, it’s the last sentence of my post. You don’t understand the purpose of my post but you keep speaking. You’re taking things out of their context just to keep talking. Please give us a break, go gather you own numbers.
I dont need to gather numbers… Im just a stupid gorilla… But dont fret pimpin i am leaving this thread for good now and awaiting the resulting cries to further nerf an already useless kerby and nerf kraken into obscurity as well…
Whether intentionally or as a result of a mechanical error, it appears at least on a glance, something is not working as intended. I don’t believe I notice too much of a difference with devour rates and haven’t collected the data others have. Respawn rates on dragon soul and infernal king are noticeably askew with a higher rate than represented.
Sounds like it’s time for someone to put a SWAT team together to systematically collect data on random chances across a larger pool of players. I’m with @Vangor that human error is a serious problem. One or two instances of failure to record correctly can really skew the data in a small sample size. Particularly when it can be something like counting a failed devour b/c the troop died before it could be devoured or b/c it hit an Impervious troop. Having more people collect data should help smooth out any of those inconsistencies.
