I just did 6 different battles against Kerberos he casted 9 times and got 8 devours that can not be a 50% chance. 90% of the time I have faced him before he has devoured.
When you toss a coin it is 50% to get heads or tails but you still can get 10 x heads in a row
10 Likes
I would have agreed with you a month+ back - since I only saw one side - the AI.
However, I’ve been playing now with Kerby on my team for 2 months solid. (even without him being traited).
He has fired off 496 times ( i have been keeping track) and has successfully devoured 267 times
Now I can take a snap shot out of there and show skewered stats
there is a 75 attempt run where he did 100%
there is a 30 attempt run where he did 0% I chocked it up to him eating the mech bombots
I don’t cast devour in every battle but I did use him a lot
7 Likes
Absolutely! Don’t forget that a coin like a nickel has a 1 in 6000 chance of landing on the edge too.
3 Likes
I’ve seen it happen with a silver dollar… freaky btw.
1 Like
So what you show here is there is a problem with the RND algorithm. A streak of 10, 15, heck, even 20 might show up during normal play with the number of players we have. A streak of 75 and then a streak of 30 on the other side, is far beyond normal outliers.
is far beyond normal outliers.
No because I don’t always use Kerby - I have a lot more fights I don’t use him at all so that is why it is called askew stats.
A pure stat would be to always use him in every fight until all opponents are dead, no logging off, no switching teams, nothing to change any condition other than Kerb and 50% devour opportunities.
I am not even sure if a kill isn’t worth counting if it isn’t a devour. (He had a lot of kills but no devours as well)
I’m pretty sure his devour procs before his damage, so that may be affecting your stats.
Also, to keep “pure stats” for this, all you should need to do is record whether the devour animation appears every single time you cast (excluding any times where you cast on a troop that can’t be devoured - although the trait that protects them should still pop up if you’re watching carefully). It shouldn’t matter whether you use other troops, log off, switch teams, or whatever. A coin toss is a coin toss, regardless of what you do between tosses.
He needs to be nerfed. Great maw is total garbage now we have kerberos. Maybe make maw able to cast more than once, possibly add an extra ten mana to his spell after each cast or something?
I personally don’t think so. But you might be in a different place game wise, are you mid game?
Not really, level 1053 all kingdoms at 5,6 and 7*
Interesting, that’s end game. Well I’ll let everyone plead their case for nerfdom.
How can that be end game when there are at least 2 if not more kingdoms, and a ton of troops left to get to 10 star (which haven’t even been released to let anyone get to it.
I would say they are in senior but clearly not end of game 
Naw, that’s end game. People who were level 60 in World of Warcraft were still “end game” when Blizzard released expansions that bumped up the level cap.
And yet despite the streaks, the results were still 267/496, or 54% - almost dead on perfect! This simply proves that the RNG is in fact working correctly. Random systems can be extremely streaky, and that doesn’t mean anything as long as the overall results are as they should be.
Or look at it this way: one post complained that seeing a 50% event happen 8/9 times was a sign that something was wrong. When another post zoomed that out to 496 trials instead of 9, it became much more clear that the 50% event was happening almost 50% of the time. If it were possible to zoom out further and see the last, say, 10,000 activations of Kerberos gamewide, it logically follows that the devour rate would get even closer to 50%.
2 Likes
3 Likes
I have this thing about stats…
1 Like
5 Likes
Both are fine. Devour and any insta-death mechanics are kind of annoying as a whole, but both these are fairly preventable. Kerberos always gives up the turn for his devour attempt, while Maw gets both a success and feeds brown + yellow plus an extra turn most of the time. Pre-buff, his 25% chance was so awful that at the time I mainly used him in a niche Black Beast team to get Ghoul summons and would take pretty much any mana match over firing off his spell (unless said enemy was already in kill range of his spell). Post buff, I’ll hazard casting him for the 50/50 chance of killing a troop that could be a problem. Getting the first attempt for 8 mana on Kerberos with the support of a Forest Guardian is a bit ridiculous, but defense teams he can be stalled/denied and getting a turn 2 Kerberos devour is not reliable enough on offense to use as a central strategy. Also, getting a Forest Guardian cast will not generally feed Kerberos (since he converts out both mana colors used by Kerb), and getting a Kerberos cast off does not board mod at all, unlike Maw teams which are generally set up to start a very painful chain reaction. I personally use Kerberos more for the auto warg summon than the devour chance, and Maw when I want that one threat very dead so I can just run over the rest of their team.
1 Like
Do you have any stats for the ai casting Kerberos?
When I use Kerberos it does feel like it’s devouring 50% of the time but when the ai uses it it’s very rare it doesn’t devour.
It might be bad luck that he devours almost everytime I battle him but I find it weird that everytime I battle him 90% of the time or more he devours on his first cast sometimes he doesn’t on his second cast but most of the time I have won or retreated after the first devour.
Last night was the worst everytime I battled him he devoured my Bombot and if I hadn’t killed him before he got filled up again he devoured Black Manacles and the one time I didn’t retreat he devoured my deep Borer as well.