TLDR → 16 failed 1LP slams in a row: 1% probability or is something wrong, I’m leaning the latter.
Nearly convinced the Eternity Cache’s algorithm is either broken, or, can actually influenced opposed to what’s stated in the guide. The guide states that the Eternity Cache selects the affixes to be imprinted at random. Assuming that’s true, each of the 4 affixes has an equal 25% chance to be imprinted.
The last 2 weeks I’ve failed 16 consecutive slams on 1LP twisted hearts (all T5-T7 +frostclaw). The probability of failing a 25% chance 16 times in a row can be calculated as such:
75% failure rate = (.75)
(.75)^16 = .0100259575
16 consecutive failures = 1% probability
My hypothesis: The 86% re-roll chance on “very rare” prefixes are in fact influencing the Cache’s algo and skewing the imprint probability of “very rare” prefixes to much much lower than 25%.
Why? → When targeting affixes with Common (0%) to Rare (50%) re-roll chances I’ve had many successful T7 slams, the majority within 1-3 attempts. Whereas when targeting “very rare” affix slams I’m rarely hitting them, if ever.
Is it possible that RNJesus hates me and I’m truly in RNGHell? Surely, more than possible
My hope is this is enough for EHG to actually take a look… because in the current state, CoF is more painful than farming Duriel on D4 MG NEXT CYCLE!
I used to play poker. 25% is a bit less than open ended straight draw. At times, I would be missing my straight draws for days on end. Also, I have witnessed occurrences as rare as one in a million. Rare things happen if sample is large enough.
HOWEVER, Eternity Cache is total b***h. I would love to see some bad luck protection there.
Completing open-ended straight draws by river is 31%, or 17% on each street and there’s no 1-in-a-mil poker hand probabilities, flopping a royal is 1 in 600-something thousand. I played for a living.
Your probability is probably right, but its still just average.
For fun, I tried inputting into a number generator 16 attempts that give 1 at chance of 25% and 0 otherwise (75%).
It took me like 4-5 of these generations until I got straight line of zeroes up to 15.
It took me 20 of these generations until it got me straight line of zeroes up to 16.
As for the rest, it is extremely common to get long lines of zeroes.
That 1% chance for it is based on your 16 attempts. Thats pretty tiny sample. If you were to repeat this 16 runs several times and check probabiltiies for each, there would be large change in the %.
For fun, I ran that random number generator of those 16 attempts several times just like that. At the end, based on those, minimum and maximum I got everything between 0% to 62.5%. Not 25%. Over large number of attempts, the average would indeed go to or increasingly close to 25%. Not with 16 attempts.
For fun, I did the same, but did the number generation up to 100. Similarly the minimum and maximum success chances for those attempts were still between 18% and 31%. Much closer, but still not 25%. At hundred, it would already start to give some numbers that might already have some trust in them since the difference is getting small.
Sounds like you know. I’m curious, whats the odds you have the best hand possible with whats available, while player 2 has the second best hand with what’d be available, while player 3 has the same?
EDIT: To rephrase… whats the most unlikely combination possible among 3 players and probability of it happening? I can imagine 3 full houses with 1 player A, 1 K, 1 Q’s. That sounds pretty unlikely… (more unlikely than the royal flush, idk).
As to the 1% thing, unlucky sucks. It must happen to 1 in 100 players I guess, and it’ll be that guy who posts about it
I’m sorry that you’ve been so unlucky! I’ve yet to have any good LP drops but it must suck when they don’t proc the right affixes when you slam!
That being said, I play a table top board game called Blood Bowl, which is like a fantasy American football dice based game. The amount of times I’ve rolled 1s on a D6 when trying to sprint my players up the field is insane (which makes them trip over). They then take an armour roll to see if they get injured on 2D6, if their armour gets broken they roll a casualty roll on a D20. On a 20 they die…… this has happened far too many times for stats to feel real. When a game hates you, it hates you! In Blood Bowl there is a god called Nuffle. It’s common for people to say that Nuffle hates everyone! it’s also very common for people to laugh these poor dice rolls off as part of the character of the game and yes, I’ve had a player die in the last turn of the game as I have to “go for it” (move an extra square to resemble sprinting and putting extra effort in at the cost of a dice roll of 2+ on a D6), I fail and use my last token to reroll and get another 1!! It’s horrid to loose a game like this as well as a great scoring player to a heart attack or whatever injury you want to make up to fit the theme.
Whilst I appreciate that our experiences are very different thematically in the 2 different games the TL:DR is that the old saying about statistics is true. ‘Lies, damn lies and statistics”. It’s not nice to hear but it’s true. I do hope Nuffle shines on you and you get your slam next time matey.
Silly and completely unrelated note: when I play Blood Bowl (also a computer game Blood Bowl 3), I play some seriously risky moves that min/max type players look at me very oddly when I suggest the move!!! You should see their faces when I pull it off though. Some of them hate playing me as they don’t know how to defend against crazy play styles and they prefer to play against other min/maxers as the game is more predictable for them. I play erratic because Nuffle shines on me when I play the way he intended the game to be played. Sure, my players die more often than they should but this is what the crowd want and enjoy. All hail Nuffle!
THIS. Real RNG are too streaky to feel good in games. While 16 failed slams is only 1%, 10 failed slams in a row is 5.6%. That’s a pretty good chunk of the playerbase who will fail over and over with no reward.
That doesn’t even take finding the unique and rare affix exalt in the first place. T5 class affixes don’t grow on trees for CoF players and it’s not as prophecies help much.
@Reticai - Adding a 3rd player into a statistical scenario (3 highest ranking hands) like you mentioned would push initial required hole card distribution probabilities into the billions I would imagine, assuming this was on a full board (all 5 cards).
To keep it simple and answer your question, If I remember correctly the odds for 3 players to all flop a set (each has a different pocket pair and flops one of their 2 outs) is around 1 in 1,900. In that specific scenario those 3 players would have the 3 highest ranking hands possible at the time (only 3 community cards).
@yawan - Haha I gotcha… Flopping a set 3 times in a row with the same two hole cards has to be in the hundreds of millions, if not over 1B.
I’ve flopped a set (different pairs/hole cards) in back to back hands many times, but never 3 in a row unless I’m forgetting. But then again, a whole lot happens after playing 10’s of millions of hands. Have never flopped a royal though!
1000 looks like a good sample for something believed to 25%. Even 100 would be good. From statistics courses i had i remember 30 hits to be the magic threshold/rule of thumb for relevance. That’s when your “invalid complaint” becomes valid.
Maybe eternity cache is random, but weighted and that little detail didn’t make it to the guide?
@MorriganGrey - Agreed. I’m personally having a blast on LE (350+ hours since launch) and will never go back to #D4bad. That said, CoF gearing / end game progression feels unbelievably hindered by RNG. Worse than Duriel rotas, never thought I’d say that…
Selecting CoF seems to also inherently mean you should consciously pick between:
Brain numbing farming requirements to obtain / slam the strongest & most widely used items in the game (boss drops, rare uniques, 2+ LPs)
Drastically narrow your build selection to only include those without rare items (aka ignore 90% of S & A tiers / strongest builds)
If CoF had a mechanism which enabled us to use our 100’s of Base / 1LP uniques towards progression, that would be a start. For instance, if I could actually make use of my 25+ non-LP hearts and craft all 25 into say 5 hearts with LP… that would drastically improve CoF experience and our hordes of stashed items would actually = progression.
MG NEXT CYCLE!
it’s also very common for people to laugh these poor dice rolls off as part of the character of the game and yes, I’ve had a player die in the last turn of the game as I have to “go for it” (move an extra square to resemble sprinting and putting extra effort in at the cost of a dice roll of 2+ on a D6), I fail and use my last token to reroll and get another 1!! It’s horrid to loose a game like this as well as a great scoring player to a heart attack or whatever injury you want to make up to fit the theme. 
