1) Expected score: the heart of Elo
Elo assumes each rating difference implies a probability. If you’re rated higher, you’re expected to score more points. If you beat someone much higher rated, that’s “surprising,” so you gain more points.
2) K-factor: why some ratings move faster
The K-factor controls how quickly ratings react to new results. New players often have higher K, so ratings adjust faster. Established players have lower K, so ratings feel sticky.
3) Why your rating feels stuck even when you improve
Improvement often shows up as fewer disasters before it shows up as a big rating jump. If you reduce blunders, your results stabilize. Then you can convert small advantages more consistently, which is when Elo climbs.
4) Ratings are noisy. Track the right thing.
One week of rating movement means almost nothing. Track patterns: opening losses, time trouble, endgame technique, tactical blindness. Fix one bottleneck at a time.
Practical next step
Analyze 10 recent games, tag every decisive mistake, and count categories (opening, tactics, endgame, time). That histogram is your real training plan.
Try it: Game Analyzer