Forums

Amount of games played vs rating?

Sort:
ddave2150

How much does the amount of games correlate to rating? Is someone with way less games and the same rating a more talented player than someone with way more games? 

I see some players with 10,000 to 50,000 games on this app but are still 800-1500??? I just dont understand that at all. Does that mean that their brains just dont work with chess? Are they not smart enough? It's genuinely puzzling to me. On the other hand I've seen players with 1000 games that shot up to 2000 plus. 

Personally I've had maybe 9000 games across all my accounts since I've stared and my peak rating is 1950 something I think. I think maybe 3000 of them were bullet though. I know it's hard to improve playing that exclusively like I did for a while. 

tygxc

@1

"How much does the amount of games correlate to rating?" ++ Not at all.

"Is someone with way less games and the same rating a more talented player than someone with way more games?" ++ The one with less games learned more from them.

"I see some players with 10,000 to 50,000 games on this app but are still 800-1500?"
++ They repeat the same mistakes.

"Does that mean that their brains just dont work with chess?"
++ They do not learn from their mistakes.

"I think maybe 3000 of them were bullet though." ++ Bullet and blitz are worthless for progress.

TheJobavaSicillian

If you want to do the work and find out it's easy. It just takes a long time. I did this on Rocket League a few years ago. You just randomly select players on the leaderboard and take their number of games played and rating. Then you make a scatterplot and fit it with an asymptotic regression.

TheJobavaSicillian

If that's too confusing put my paragraph into chatgpt and ask it how to do what you just said.

TheJobavaSicillian

Long story short there will be a high correlation from like 0-5,000 games played, then the correlation will become very weak, as other factors like natural talent, other forms of practice, tactics, openings studied, etc.

TheJobavaSicillian

If i had to put a blanket median on it probably 1000 games 1200. 5000 games 1600. 15000 games 2000. 50000 games 2300+. But very high rated players(2200+) will likely be OTB players, or already strong from other accounts. So the noise on my data collection suggestion will be huge in the 2300-3400 range. Probably not worth even collecting. I'd also throw out outliers. Like a 2000+ that's played less than 1k games, etc.

ChessMasteryOfficial

Some players focus on playing a large number of games without dedicating sufficient time to study and improvement. The quality of learning and reflecting on games is crucial for progress.

cR1NN
ddave2150 wrote:

How much does the amount of games correlate to rating?

It doesnt

Antonin1957

So many people here are so incredibly focused on ratings, I wonder if they actually care about chess. Study chess principles, study great games from the past to better understand what made great players great. There are no shortcuts. To improve, you must put in the work. And you must love chess for its own sake.

Antonin1957

There is no "leveling up" in chess. You can't just grind out a few thousand games and automatically "get good."

Antonin1957

If you play through the games of the Fischer- Spassky match and say truthfully that you understand why Fischer won and where Spassky went wrong...that is worth more than the fact that you have blazed through 5000 or 9000 games of blitz or bullet.

Cold_W1nter

I don't think that caring about rating is necessarily a bad thing, nor indicates a mindset of looking for shortcuts. Instead I think it can be a healthy way of setting goals, which encourages that behavior you listed above.

ddave2150
Antonin1957 wrote:

There is no "leveling up" in chess. You can't just grind out a few thousand games and automatically "get good."

Yes you can... I know plenty of people have haven't done any studying and just shot up to 2000 plus. Some people are naturally talented.

Chess147

If there was an option to hide player rating I would use it because it's become a distraction. I found myself basing decisions on my opponents rating and have decided that's not the way to learn and I should play every move based on the board not the player. I'm not going to talk myself out of a move based on nothing more than because my opponent is a few hundred points higher than I am and equally not take liberties if their rating is a few hundred points lower. I'm learning to be patient, clinical and final.

Ratings are important to an extent because they are a by-product of performance and if there's a steady increase you know your game is going in the right direction but it will always be fluctuating as you win and lose. I'm taking every game I play seriously and thinking about every move with the long term plan that the more daily games I play = more positions = improved memory muscle. The more I play the more my board vision improves provided I spend time thinking about the move and why I need to do that particular move. Blindly making moves without any thought is a complete waste of time but being patient from the opening and learning from what works and what gets you checkmated by move 10 is the way to turn a game in to a positive learning experience regardless of who won.

Antonin1957

@ddave - I stand by what I said. If you think you or "plenty of people" can shoot up to " 2000 plus" without studying, all I can do is laugh. That is an unprovable troll comment, and because I don't believe in feeding trolls, I won't respond any further.

ddave2150
Antonin1957 wrote:

@ddave - I stand by what I said. If you think you or "plenty of people" can shoot up to " 2000 plus" without studying, all I can do is laugh. That is an unprovable troll comment, and because I don't believe in feeding trolls, I won't respond any further.

????? Im not trolling at all... There are so many players that became good without studying and only playing online games tf? Im sure many people on this forum know people who have.

Just because you are 800 and haven't gotten good even with studying doesn't mean it invalidates my point lil bro.

putshort
Mine doesn’t correlate as I joined a bunch of tournaments pairing opponents all at the same rating as me at a time when my rating was very much lower than you could imagine if you were also forced to start at a 400 rating because the site didn’t explain that it would do that that is way things go oviously
Chess147

I think chess is a prime candidate for a game where the best way to improve is to play, lose and repeat. Embrace the experience and knowledge that very soon you will know all the basics and be able to get to a middlegame with strong players. From there you're on your own. Becoming a very good chess player doesn't take intelligence or wisdom which is why people of all ages and intellectual abilities can become titled players. I estimate that when I've finished all of my current daily games it will be roughly 50/50 win loss rate and the extra wins I get from time outs will be offset by games I throw by blundering. I'm not starting new games and wish I hadn't started playing tournaments because all the timeout wins has messed up my stats. I assume it's because some tournaments start months after being created and some of the players are not available so all of their games timeout.

TheJobavaSicillian

I've "coached" a few guys that just do training games with me. Some of them were around 18-20 when we started doing training games. Couple of them have hit 2000 just playing training games against me, doing puzzles, and playing rapid/analyzing. That's really all you need to get to a high level if you have the talent.

Jklenear

very good question I think of the times like Morphy or better yet Philidor (invented a opening,endgame and style of play!) were their primary source of improving was simply to play tough opposition as often as they could then there's those like Morphy who more or less invented an entire style/set-of-guidelines still used by masters today and he's estimated to be about as strong as a IM in his peak. This can maybe be answered best by looking at the history of computers.