Win Percentage Question

Hi everyone.

Am I correct in my thinking that, as a beginner, having a win percentage that is significantly higher than 50-ish% is basically impossible?

I’ve for some reason become fixated on my win percentage, but then it kinda dawned on me that if the win percentage was going up constantly, it would mean I’m more than likely not playing opponents the same skill as me.

Is it correct to understand that the ranking system makes sure you are playing similarly skilled opponents, and therefore hovering around 50% is perfectly normal?

For what it’s worth, I’m 48/48/4 for percentages, with usually only a few games separating my wins from losses, but I am seeing a steady climb in my ranking as I’ve gotten better at “seeing the entire board,” so I feel like I’m getting better, but I’ve just  become fixated on that percentage.

Should I worry about it at all?

Thanks
Jeff

Good thought and yes you are on the right track about winning percentages

If you are playing players around your rating/ability all the time, then you should theoretically remain close to 50% winning percentage for half the times you lose and half the times you win (draws equalize out). However, some people will have a winning percentage of something like 70% and this isn't necessarily because they are chess hustlers When you increase in rating, then you are generally a stronger chess player and there are obviously fewer players of said ratings the higher up you get (it is lonely at the top right? xD). For this reason, if you are super high rated, then you might win more often simply because you encounter people rated close to you much less often. This is common for 2000+ ratings for sure, but most will never reach that level and experience this problem. 

If you use the chess.com pairing system for whoever it pairs you against, then usually someone in the entire world will be close to your rating and so expect the win percentage to always be around 50% give or take maybe 10% for variance. 

I wouldn't worry about winning percentage too much at all though. I would only make note of it it you significantly win more often with the White pieces or the Black pieces. This is itself isn't too uncommon (usually people do slightly better with White because they play their chosen opening but as Black they don't know what the opponent will try), but if you are really far off, then it might be something to take note of. 55% to 45%? Nothing to worry or think much about. 70% to 30%? Not a big deal, but you probably are much higher rated than others you play often or REALLY love one specific opening you know very well. Something like winning with Black 90% of the time would be uncommon that that might mean your repertoire for the White pieces needs to change!

In summary, it is a neat little statistic, but it usually doesn't mean much at all. It will almost always be close to 50% regardless of rating (assuming you play players near your rating often).

Also, if you want a quick estimate of your win probability versus an opponent - you can mess around with an elo calculator like this one:

https://www.3dkingdoms.com/chess/elo.htm 

Just type in the ratings and hit calculate if you are curious of the matchup

Few things to keep in mind though:

-It is only an estimate - not the actual percentage. 

-It is only as accurate as the information. You opponent might be 1500 rated, but if you are playing well and they are having an off day with lack of sleep, then obviously they might play more like a 1400 player but the calculator won't factor these things into account.

-this estimate also doesn't include opening repertoire changes or similar factors that might make the more experienced player a better chess player than calculated. Perhaps their favorite opening you know nothing about! That is an edge to them for sure. Conversely, your chosen opening might refute their dubious opening choice if they choose to play something unsound.

* there are probably more reasons to not take estimates too seriously, but I also just wanted to comment that this is an elo calculator and chess.com uses glicko ratings - not elo. However, they are virtually identical, so I use this anyway. I find the main differences in glicko and elo to be the k-factor for the calculation process (which applies more for newer accounts anyway), so for estimation purposes, elo and glicko and basically the same thing

If a person's win percentage is higher than 50%, that means the person is not playing against stronger opponents.  The way to improve is to keep player stronger opponents. 

Having a high win percentage only means that the person is beating up on weaklings and one's ego is too fragile to handle losses. 

Win % is a measure of opponent's strength, not your strength. If you want to progress, then your win % should be below 50%, otherwise you play too weak opponents.

I have just under 3000 games played and my win percentage is 48%, so not 50% but for me it's still quite good, but yeah I am worried it'll get worse once I get much better! lol

Well, I have 50/13/37 - 50% win 13% draw and 37% lost, but I am gaining each week more and more elo (for this week right now 40 elo) + I am getting better (by puzzles, courses and analysing) so you are not right.