Chess Apps for Endgame Practice

Capablanca's rule still holds: study endgames first. A century later, almost no club player does. Every plateau between 1500 and 1900 has the same three causes — and one of them is reliably endgame technique. Up an exchange and a pawn in a rook ending and drawing it is the most common 50 ELO leak in the database.

ChessDir is run by a 1700 FIDE / 2100 Lichess player. Below is a tested set of apps for endgame training, ordered by how much rating per minute they return. None of them is exciting. All of them work.

Why endgames need a different app

Tactics and opening apps don't cover endgames well, for two structural reasons.

1. Endgame puzzles aren't tactics puzzles. Tactical motifs reward calculation depth. Endgames reward technique — the ability to convert a known winning structure (KP vs K, R+P vs R, opposition, Lucena, Philidor). General puzzle apps mix endgame positions in randomly, which trains nothing systematic.
2. Endgames are memory more than tactics. Lucena is a recipe; the Vancura defense is a recipe. Apps without spaced repetition or themed drill cycles let those recipes fade. Most players know Lucena exists, far fewer can execute it under time pressure.

The picks

For free, structured drilling — Lichess Endgame Practice

Lichess ships an underrated endgame practice mode under Learn → Practice. Themed drills cover the canonical positions: K+Q vs K, K+R vs K, opposition, Lucena and Philidor in rook endings, knight and bishop checkmate, same-color and opposite-color bishop endings. Each drill plays the defending side at perfect tablebase strength — there is no "easy mode" sandbagging the practice.

Free forever. The right place to start, regardless of rating. A 1500 player who can't reliably win K+R vs K under 50 moves needs to spend an hour here before any further endgame app.

For calibrated drilling and weakness tracking — Chess Tempo Endgames

Chess Tempo's endgame trainer is the most rigorous on the market. Positions are drawn from a curated database, calibrated to the solver's level, and statistics break performance down by ending type — "rook endings 78%, minor piece endings 41%." That's the diagnostic that turns endgame practice into a directed lab.

Free tier covers a daily quota; the Standard tier (~$4/month) unlocks full statistics and unlimited problems. For any player at 1500+, this is the highest-leverage paid endgame tool.

For systematic theory — Chessable endgame courses

Chessable hosts the canonical endgame courses in MoveTrainer format. Two stand out: 100 Endgames You Must Know by de la Villa, and Silman's Complete Endgame Course. Both pair annotated theoretical positions with spaced repetition — exactly the memory architecture endgames need.

These courses convert "positions seen in a book once" into permanent recall. After working through 100 Endgames at the appropriate rating tier, the most common endings stop being decisions and start being procedures. Course prices run $30-60 one-time; Chessable Pro at $74.99/year unlocks both courses plus the rest of the catalog.

For low-rated endgame walkthroughs — Dr Wolf

Dr Wolfincludes basic endgame positions with narration: "Bring the king to the center first, the rook stays cut off behind." Useful for players in the 800-1300 range learning the core mates (K+Q, K+R, K+two bishops, K+B+N) and basic king-and-pawn principles.

Above 1300, Dr Wolf's endgame content thins out — graduate to Lichess Practice and Chess Tempo Endgames at that point.

A practical endgame routine

Realistic for a club player with limited training time:

• Daily — 5 minutes, Chess Tempo Endgames. Three calibrated positions before tactics. Misses go into a review pile.
• Weekly — 20 minutes, Chessable endgame queue. One MoveTrainer session through de la Villa or Silman. Adding 10-15 new positions per week converts the canonical theory in 4-6 months.
• Monthly — one Lichess Practice marathon. Run through every theme drill in one hour. The drills that crack first are the ones to bring back into weekly rotation.

Total: ~60 minutes per week. Less than tactics, less than openings, but with a higher per-minute return at the club level because endgame skill is so unevenly distributed.

The endgames most worth drilling first

Order matters. The list below is roughly the priority order for a 1400-1800 player.

1. King and pawn vs king (opposition, key squares). The base case. Without this, every other endgame collapses.
2. Lucena and Philidor positions in rook endings. Rook endings are the most common ending in OTB chess. Lucena (winning the R+P vs R) and Philidor (drawing the same down a pawn) decide most of them.
3. Rook endings: cutting off the king, active rook principles.The technique behind "rook endings are drawn" — and the technique that wins them anyway.
4. Same-color bishop endings with passed pawns. Often misjudged by club players who treat them as drawn.
5. Opposite-color bishop endings. Drawing tendency well known; the conversion technique when ahead two pawns is not.
6. Knight endings. Fewer occur in real games but the technique is genuinely tricky and worth the small allocation.

FAQ

Endgame books or endgame apps?Apps for daily reps; a book for the underlying ideas. Dvoretsky's Endgame Manual is the reference; pair it with Chessable's 100 Endgames for the spaced-repetition layer.

How early to start serious endgame work? Around 1300-1400. Below that, basic mates and tactics return more rating per hour. Above that, endgames become the highest-leverage neglected area.

Are tablebase endings practical training?Yes for 7-piece-or-fewer theoretical positions. Beyond that, tablebase moves often look superhuman and don't teach human-playable principles. Stick to the canonical positions in the apps above.

Which apps cover Capablanca's endgame approach? Chessable's Silman course is the closest match in app form — the book argued for studying endings by rating tier (which positions matter at 1200, then 1400, etc.) and the course preserves that structure.

Free vs paid endgame stack? Free stack: Lichess Practice + 100 problems/day Chess Tempo. Covers most of the territory at the cost of the diagnostic statistics. Paid stack adds Chessable theory and Chess Tempo full statistics, which is where serious gains compound.