In stating that the AI cheats. You have to realize that this statement is biased.
Saying you had a game like X - doesn’t help. We all have. Just like we had lucky games. That doesn’t prove anything. Statistically - it has to happen sooner or later.
However, people always tend to take negative luck to heart instead of the good one. Unless you provide data that that this is not the case and out of 1000 games AI really had more luck than you (assuming you didn’t just miss your chance and give it to the AI), we can assume lots of players just feel the phenomenon of bias.
Also, there was something mentioned about the player never surging. Check the Jinx trait available at Banshee, Famine, … You might just be interested how it suddenly makes sense.
I suppose you’re looking into cascades despite it ‘all being in our heads’. That’s good news, thanks salty.
In the UK krud means crap, seems a pretty appropriate name seeing as you talk a lot of krud.
Used same team on all 3 pc accounts of gws yesterday in gw and 1 on my PS4 account…
Jarl
Hellcat
Tds
Gwayne
Pc = 15-0
Ps4 = 2 - 3
And 2 of the losses were honestly hilarious. The cascading on them was out of control. 2nd turn ai had max mana everyyone.
1 of the losses was legit as my opponent was Triple famine and it got out of hand.
The ai was getting insane drops. Every jarl cast in the losses not 1 match.
Why?
Famine is not OP at the moment. It is the stacking of multiple Famine’s that is broken. 2x Famine means 2x Jinx trait, and that’s what’s currently the issue. People are having difficulty filling up their mana to cast spells against Famine because the 2x Jinx prevents most human player mana surges.
So the logical conclusion would be just to either not make the Jinx trait stack anymore, or remove the Jinx trait from Famine.
I think it’s very generous of you to re-run the same tests again to help alleviate some of the console players concerns. I have one request for those tests, if it’s possible to pass it along to the devs, and that’s for a histogram of the cascades difference. As some have mentioned, there seem to be a number of issues confused, specifically the AI decisions, the streaky cascades, and AI favouritism. While the overall averages provided from the previous tests showed parity over time, it didn’t highlight some of the more extreme situations which occurred and show that it occurred for both sides. Instead of simply “Average Player Cascades” and “Average AI Cascades”, a histogram which shows the difference in cascades would make this more visible. Show how often the player got X cascades more than the AI and vice-versa, something that visually confirms some of the players feelings that the AI gets significantly more cascades sometimes, but also demonstrates the same thing happening on the other side (balance/fairness). And once that helps quell the cheating/favouritism claims, it may also provide a better opportunity for conversation about how much they should occur for the sake of “fun” rather than parity (since as some have mentioned, just because it’s legit or fair doesn’t mean it will be fun to play).
Alternatively, and maybe even a better idea if they’ll do it, is to release the test data with the summary. That would allow some of us in the community to do a deep dive and prepare some of that information ourselves, which would free up developer resources . Didn’t think of that until after I typed out the earlier/histogram idea, but I know some of us would love to help out with or do our own analysis
Maybe I misunderstand, are you meaning that the immune for manaburn trait will now also have immunity for mana drain as well? I never heard that any new nerf was in the plans for Famine that just had a nerf, before today.
@Salibu your GW results are in line with what I’ve observed, but not quite as pronounced of difference for top Bracket players. The Unity AI decision making absolutely does make a difference, I documented the differences the best I could with actual results from the top cluster of Bracket 1. (PC vs Console)
PS to other players: I personally don’t believe the AI cheats on its gem drops.
. Didn’t think of that until after I typed out the earlier/histogram idea, but I know some of us would love to help out with or do our own analysis