4 ms·
I think matching by level is quite psychologically destructive, and should be avoided. People grind for hours and they see no sense of progress as they keep al
by GistNoesis 3y ago
I think matching by level is quite psychologically destructive, and should be avoided.
People grind for hours and they see no sense of progress as they keep always being around 50% winrate because the stronger they get, the better the opponent get. And inversely they can play drunk chess and still win 50% and burn time pushing random pieces for a little short-term fun.
It probably depends on people but I think the average ratio for people to progress optimally their should be around four wins for one loss. And the loss should be correlated to the way you played. Anything below 80% winrate, means it becomes hard to distinguish the learning signal from the variance of the matching. The brain also need some balance between the carrot and the stick otherwise the ego take a hit.
Rating system also generate some strata artefacts meaning there are levels you can't get past without spending as much time as your opponent on tactics, or openings, and you never get to experience chess on a more strategic level, or deeper understanding level. This mean that most of the fun get extracted out of chess once you get addicted/interested enough.
I think it's much better to play unrated games anonymously against a random pool of anonymous players (eventually with a min level to exclude complete novices), and offer rematch when the games were interesting. You should play to win but your reward signal shouldn't be the win but rather the interestingness of the game for both opponent.
When playing you should try to understand your opponent, his current chess understanding and how it's revealed through his current moves, and construct with the board some situations that will be interesting enough for the both of you to have a good time. Don't play the board. Play the player through the board. A good game should look more like a dance than a fight. You have to infer from a few moves who your opponent is, and how to interact with him in a constructive way.
That's what popular streamer do when they play their viewers (usually friendly and constructive). And that's probably what teens expect when they join chess after viewing a few videos. But instead they get thrown into an adverse destructive pit of soulless addicted grinders.
- gubby 3y agoInteresting point, but of course it’s not possible to create a platform with an average win rate of 80%.
- rcxdude 3y agoNor any actual group of people. This "optimal" is zero sum: someone else has to have a lower winrate in order for it to work.
- sbergot 3y agoIn some video games you may randomly face an AI opponent without the game being explicit about it. That way the average human win rate can be above 50%. I don't know yet if this is good or bad.
- orthoxerox 3y agoAnd yet when you see a professional boxing match, both boxers' win rate is usually above 90%, not 50%.
- hgsgm 3y agoA boxing career has approximately 0 rated fights. A cheap career has hundreds to thousands.
- dayjaby 3y agoI play different elo-rated games for the past 13 years and completely disagree with you. 50% is very healthy psychologically as it keeps you humble and challenged all the time. If your play improves to ~55% win rate, you already feel highly rewarded and most of the reward feelings come from the fact that you now play on higher elo. You don't care about single losses/wins anymore.
- dist1ll 3y agoThat's a very elaborate way of saying that competitive players should delude themselves to cope with their weak mental game. > Anything below 80% winrate, means it becomes hard to distinguish the learning signal from the variance of the matching. Using winrate as a learning signal is your first mistake. You're making a million different decisions per game, and then decide to summarize it with a binary W/L? If you want proper signal, you need to analyze your games on a more fine-grained level. > And the loss should be correlated to the way you played. No, it shouldn't and doesn't. Dealing with variance, and things beyond your control is an essential part of developing a strong mentality. Especially when playing in-person. > The brain also need some balance between the carrot and the stick otherwise the ego take a hit. If your ego takes a hit from this, you need to get back to the drawing board instead of praying that the world will change to accommodate your issues. > you can't get past without spending as much time as your opponent on tactics, or openings, and you never get to experience chess on a more strategic level, or deeper understanding level. This mean that most of the fun get extracted out of chess once you get addicted/interested enough. I can't speak to the "lack of stategy", but what makes a game fun to compete in is highly subjective. > soulless addicted grinders This is the exact kind of language and attitude I would see from players who plateau hard, and looking for external factors to blame.
- oblio 3y agoYou sounds like a Dota player. You do realize that a lot of players play for their definition of fun and don't derive or base their personality around their strength as chess players, I hope?
- Shorel 3y agoYou can set up some AI to let you win 80% of times, and let the game to the real players. That way, you will feel better. I dabble in the realm of simracing. Man, what would I do for a 50% win rate. Every race is hard fought, and the possibility of a victory is far below 15%. The times I have won, I trained the whole week, it was hard-earned, and really, nothing has felt better since. It's a high I can still ride with pride. IMO, if you want to win more, you should train harder, or try something easier.