At any competent level of play, a game like one-round Lost Cities is pretty darn close to a coin flip for who will win. It's not uncommon to have a stretch where you win 8 out of 10 games and later a stretch where you lose 8 out of 10. As a result, my Arena ranking oscillates wildly from day to day, from around #50 down to around #350 and back up again. And as a result of that, who actually IS in the top 50 seems largely meaningless. It is probably mainly players like me who just happen to be at the top part of our swing at that particular moment.
In fact, I noticed a few days ago that among the top 10 players of Lost Cities, not a single one of them had actually kept playing any appreciable number of games once they got their top 10 spot. They get there, then they immediately stop playing, probably because they understand the volatility -- that stopping is the only way to stay there. To me this seems incredibly sad. (I'm not bringing up the topic of rating decay here though.)
Why is it like this? Why not divide the number of ELO points exchanged by 2 or 3 or something? These matches are so short and luck-based that you don't really have a valid basis for exchanging as many points as it's currently doing.
Alternatively, could your ranking be based on your *average* ELO over your last 5 matches instead? (That method has an unfortunate anomaly in that it is possible for your ranking to fall slightly even after you just won a game, which would really piss people off. So you could do a weighted average by recency.)
In fact, I noticed a few days ago that among the top 10 players of Lost Cities, not a single one of them had actually kept playing any appreciable number of games once they got their top 10 spot. They get there, then they immediately stop playing, probably because they understand the volatility -- that stopping is the only way to stay there. To me this seems incredibly sad. (I'm not bringing up the topic of rating decay here though.)
Why is it like this? Why not divide the number of ELO points exchanged by 2 or 3 or something? These matches are so short and luck-based that you don't really have a valid basis for exchanging as many points as it's currently doing.
Alternatively, could your ranking be based on your *average* ELO over your last 5 matches instead? (That method has an unfortunate anomaly in that it is possible for your ranking to fall slightly even after you just won a game, which would really piss people off. So you could do a weighted average by recency.)