Early on in the ELO era, a lot of hand wringing centered around some player intentionally losing games, presumably to inflate the ELO of those games. So, if I understand correctly: a) the same player tanking the same game repeatedly does not impact the ELO of that game beyond the first loss; b) randomly losing a string of games in quick fashion might impact the ELO of those games, but would not likely affect the guilty party's ELO. (I am assuming that the perp in this case was playing under some "burner" nickname, not their normal name used for serious play.)
If i understand well, level 5 is the result of a card change from an original state. This change consists in moving low cards to get them more accessible. The consequence is that most of the games become easier but not all
Originally, ElGuapo, I was asking how a 7x3 game that was 1/59 could only be rated 1906. I know the math is what it is (but you confirmed all those are probably underrated), but that just seemed fairly stinky. So my question really had to do with a) how many times somebody beat on it to take it down; and b) thus what ratio of plays/players that 59 represented (likely unknowable, or essentially so, in practical terms).
And then this comment by joey: "Almost all of those plays were on the replay, and so didn't count." ........confused me further. So I was looking for clarification for what "counts" and what didn't - in order for the 'count' to get to 59.
Ok, but you are clarified now, right? Games-that-count-for-streak get ELO treatment (and cause secret sauce ELO adjustment as well.) No other games affect ELO at all.
*editted*
(If secret sauce is still happening, that is. If it is, I think it only affects previously unplayed games but wouldn't swear to that either.)
I think so - but not certain. And I still don't know the most likely explanation for how the 1/59 got arrived at. But realize that's almost certainly unknowable.
Let me re-phrase the 'question'...........
If that game was already, say 1/50 when the new world arrived, then its elo rating makes more sense, because 'we' didn't know how to rate individual games, right? (Altho wasn't some consideration given to a game that was, say, 2/10 vs. one that was 2/40, in the same variant? Don't know that, but I *think* that was baked in?) But if it (the 7x3) was only, say, 1/5, and then since the new world arrived someone played and lost it 50+ times, trying to bludgeon it to death, then those extra 50 attempts do not 'count', right - but show up *only* for the game stats, and NOT for the game elo, or the player's elo. But if 50 unique players played it and lost, then there would be points exchanged each time. Hope that's straight. Thank you all for your patience.....
I should have clarified my earlier response was to BuzzClik's question. I thought joey covered the rest and agree with his responses.
Also, @joey, yes the secret sauce is a permanent part of ratings and is ongoing. It affects all games in the level and is the reason the 7x3-5 ratings are so low (too low by my subjective opinion). We get anomalous ratings anywhere people play at superhuman levels (i.e., never recording a loss), but it's more pronounced in game levels like this where mere mortals never tread.