You are probably practising the wrong way.
Not because you are lazy. A study of 44 213 online chess players found that people spent about 90% of their practice time on the activity linked to the least improvement per hour. The activities linked to the most improvement got only a sliver of their time.
The lesson reaches far beyond chess.
90% of practice time went to the least efficient habit
Daniel Southwick of the University of Pennsylvania, with colleagues at Brigham Young University and the U.S. Army Research Institute, followed Chess.com players who signed up in March 2024. The study was published in Psychological Science in June 2026.
For six months, the site logged what every player did and how their rating changed. No memory. No self-reports. Objective, time-stamped records.
Here is where the time went:
- Playing rated games: about 90%
- Reviewing finished games with the site’s analysis tool: under 5%
- Puzzles, short tactical problems: about 4%
- Video lessons on openings, endgames and strategy: about 1%
Players spent almost all their time simply doing the thing.
An hour of review was linked to as much gain as five hours of play
The researchers had predicted in advance that structured practice would beat playing. It did, by a wide margin.
Rating points gained per hour, after accounting for starting skill:
- Playing: about 0.8
- Puzzles: about 0.7, no better than playing
- Game reviews: about 4.4, roughly 5.6 times more than playing
- Lessons: about 5.2, roughly 6.6 times more
Taken together, the structured activities were linked to 3.6 times more improvement per hour than games.
The most useful hours were the ones players spent least.
Reviews helped at every level, and the best players used them most
The advantage was biggest for beginners. Among advanced players, only game reviews still clearly beat playing.
The strongest players reviewed more. They spent about 14% of their time going over games. Complete beginners spent under 4%.
Your own mistakes are the teacher you keep ignoring
A finished game gives you experience. It tells you very little about what went wrong, and it gives you no chance to try again.
A review does both.
It points to your specific mistakes. It shows you a better move. It lets you replay the position until you find it yourself. Those are the core ingredients of what psychologists call deliberate practice: a clear goal, immediate feedback and repeated work on your own weak spots.
If you lose a game and click “new game”, you are practising playing.
If you replay the position where you blundered, you are practising not blundering.
Puzzles should work the same way, but they didn’t here. The authors suggest that speed-based puzzle modes may lead people to solve problems they already understand, rather than work on weaknesses. They note this explanation is untested.
How to use this in anything you practise
The study only looked at chess. The advice below is a reasonable inference from it and from wider research on deliberate practice, not something it tested in other skills.
- Review your own performance. Record yourself playing music or speaking, reread your code or writing, or replay your games.
- Hunt for specific weaknesses. Look for the moments that went wrong, not just the overall result.
- Get feedback. Use a teacher, a knowledgeable friend or a good analysis tool.
- Redo the hard part. Repeat the difficult passage or position until you get it right.
- Keep doing the activity too. Playing still helped. The point is balance, not replacing practice.
What the study can’t tell you
- It’s observational. Players chose how to practise, so it can’t prove that switching would bring the same gains. Motivated learners may both review more and improve more.
- Time was estimated from activity counts and average session lengths.
- Practice off the site wasn’t tracked, such as club games, books or coaching.
- Only players who stayed active for six months were included.
- Most were beginners and amateurs. Results may differ for elite players.
- Chess may not generalise to every skill.
Stop starting the next game
Doing something a lot isn’t the same as getting better at it.
The next time you finish a game, a rehearsal or a draft, don’t start another one straight away.
Look at what went wrong first.