40/40 Review of Keepers of the Flame | One of the best leagues I've played in a while
" Lmao yeah, I'm just enjoying this league. Waiting on my 17 mil crop mirror shard shipment and then I can gamble some stacked decks. Its so nice to get enough gold to run Kingsmarch from just random alch&go/8-mod t16 maps. |
|
" Blight is my go-to strategy for farming gold. Can avoid that expensive allflame ember with it. PoE players: Our game has a wide diversity of builds.
Also PoE players: The [league mechanic] doesn't need to be nerfed, you just need to play a [current meta] build! And the winds will cry / and many men will die / and all the waves will bow down / to the Loreley |
|
" I would farm the funny big blight chest strat(I think its a scarab that makes one huge chest instead of the normal many chests?) but my PC wouldn't be able to handle that many monsters. I definitely wanna try blight out in the future though. |
|
" I cannot believe I'm saying this but.....I agree with Toforto here. I mean, thes are two players I vehemently disagree with on almost all occasions on nearly everything BUT since the two of you are on opposite sides here I have to agree with one..... @D: Your "retention" data as you present it is false. You cherry picked the rolloff from PEAK, which isn't really "retention" when it comes to PoE. This is ALWAYS the start of a false data narrative. PoE retention is, after the initial spike down from peak, how STABLE the average players remains. THAT is retention. Without fail, there is a downward spike from peak almost IMMEDIATELY (like...first weekend) of at least 30-40%. It is absolutely unfair to use the absolute pinnacle "peak" when trying to argue "retention" in any sane data-driven argument. What really matters is incredibly difficult to calculate using steamcharts (but not impossible), and that is the RATE at which the peaks fall on average over a longer period of time. If we calculated that, ultimately we could KEEP the highest peak within our dataset.....but ONLY if we average it with all other peaks like that. The "average rate of change" from peak to peak, day to day, is the TRUEST measure of retention we have access to. Not "Day 1 minus Day x". And really.....the best thing is to 'correct' for the outlier value (the highest peak) by removing it across the board before comparing rates of change. Why? Because the HIGHEST peak always has extemporaneous data associated with it: the stability of the launch, the day of the launch, the time of year of the launch, the other game launches surrounding or overlapping the launch, etc. A league that has a highest peak of 30k compared to this league's 185k might prove to have better RETENTION because the highest peak is essentially meaningless. Consider this scenario: I am a teacher, with a class of 25 kids. In a given 20 school day month, 19 of those days I teach 25 kids. But ONE of those days I happen to invite another class to join them and end up with 50 kids. If I compare that 50 to any other day of 25, you would say I lost 50% of my class, retention of 50%.....but that doesn't tell the real story now does it? If I average that day with all the others, I end up with 26.25 students on average every day.....which is also not the real story. Remove the outlier, 25 students for 19 days, average of 25 students per day. THAT is the truth. Rate of change (TRUE retention) would be [(0x19) + (-.5 x 1)] / 20, which is not totally accurate but a hell of a lot closer to the truth than -50% Here's one that suits the PoE steamcharts better: Day 1: 50 kids Day 2: 30 kids Day 3: 25 kids Day 4: 24 kids Day 5: 24 kids Day 6: 23 kids Day 7: 23 kids Your math? Day 1 - Day 7 = -27 kids for a retention of 46% for the week. While this is TRUE......it also isn't for the argument its being used for. Rate of change math? Day 1 - 2 = 60% retention, Day 2 - 3 = 83% retention, etc. Average all together? Actual week retention rate of nearly 90% (statistically corrected for an outlier value) THIS is the true value you need to compare league to league to determine which had better "retention" across a length of time. Namely: out of the CURRENTLY PLAYING people, how many are likely to quit if given the historical data of the league? Starting anew....with PoE 2 Last edited by cowmoo275#3095 on Nov 14, 2025, 8:42:58 PM
|
|
|
^
It goes without saying....but the retention rate calculated this way STILL doesn't technically tell the whole story or the 'true' story.....but it is the absolute BEST we can do if we are analyzing steamchart data. Things it leaves out: 1) New player versus returning player retention 2) Ratio of new players joining versus old players leaving peak to peak. 3) Gameplay time: a LOT of that initial peak is players who maybe didn't even kill Hillock, or ever engaged with the league mechanic. Should they REALLY count in LEAGUE retention data? Among other things. Because when all is said and done......peaks are terrible TERRIBLE data to use for nearly any argument. Starting anew....with PoE 2 Last edited by cowmoo275#3095 on Nov 14, 2025, 8:55:42 PM
|
|
|
You know, I don't think it has to be this complicated. It can be as simple as taking a look at steamcharts, and if line goes up and down smoothly day after day, not having a huge dropoff then the player retention of a game is good. Because a consistent number of people still play it every day. But it is what it is I guess I dunno lmao
|
|
" What I described is the exact mathematical "effect" you are seeing on the up-down pattern. Math always SEEMS more complicated.....until you take the time to understand it. (very much like in-game "more vs increased vs %-as-x vs. DoT multiplier vs. flat vs. effectiveness" blahblahblah.) They are one and the same, except one has an easily digestible numerical answer and the other is a very subjective "this is what I see". When someone is trying to come at you with deliberate and incorrectly used "numbers", they are going to expect and require measurable "numbers" that debunk in return, not a gut feeling. That is what is required of data usage. But we are agreeing.......its called the "rate of change" that both you AND I are talking about. Starting anew....with PoE 2 Last edited by cowmoo275#3095 on Nov 14, 2025, 9:23:54 PM
|
|
" You are wrong: https://poedb.tw/us/League#ConcurrentPlayers " Claim based on wrong argument. " You miss the point of bringing that data to this thread. I encourage you to reread why this data was posted. I refuse read further. I would be embarrassed to defend false claim at the same time making false claims. On Probation Any% Last edited by Dxt44#4050 on Nov 15, 2025, 3:12:09 AM
|
|
" What false claims though? The retention is good cus there's 80-100k curve daily, its really that simple. A large amount of players keep logging in, that's really good. And like, comparing a % of the same day where 3 different leagues had 3 different numbers of peak players is pointless? That % is a different number of players for all 3 lol, like yeah ofc Settlers has the highest % cus it had the highest peak ever for the game so far at 228k. You can't be comparing leagues to other leagues when the player peak is different for each, just makes no sense. |
|
|
>Someone actually has fun
>Shitstorm breaks out in the comments Fun isn't allowed apparently lol. League is good, although there are too few changes in ENDGAME systems for me. Atlas tree is old as hell and needs serious revamps. Because every league same shit happens - we invest everything into quantity, map mod effect and scarab drop rate. Generic crap. -We didn't get any petal skill changes in T16,5 maps - they are still lackluster -Sentinel, affliction wildwood and Kalandra reflection is still rare, untargetable and has no support on tree or scarabs at all. It just sorta exists, that's it. -League farms are outdated. They are still trying to make tormented ghosts work which either ends up overpowered as hell, or completely useless. |
|
