# Chess Q: 3 pawns vs. Knight **URL:** \ **Category:** The Game Room\ **Tags:** chess\ **Created:** [December 24, 2023, 8:11pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091 "2023-12-24T20:11:44Z")\ **Posts on this page:** 20\ **Page:** 1
**Author:** ![Pavelb1](https://avatars.discourse-cdn.com/v4/letter/p/439d5e/32.png) [@Pavelb1](https://boards.straightdope.com/u/Pavelb1)\ **Post date:** [December 24, 2023, 8:11pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/1 "2023-12-24T20:11:44Z")
King and three pawns vs King and Knight The three pawns are on their starting squares and connected. Their king is behind them. The other king and knight are centralized but in front of them basically in the middle of the board. I just wanted to make sure whoever answers knows that theres no weird advantage going on. So my Q is: At high levels, do you all just call it a draw right there? Or play a bit? Analysis says its a draw, but I don’t know if high level players know how to make that happen. ---
**Author:** ![Maserschmidt](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/maserschmidt/32/18829_2.png) [@Maserschmidt](https://boards.straightdope.com/u/Maserschmidt)\ **Post date:** [December 24, 2023, 9:51pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/2 "2023-12-24T21:51:24Z")
I think it would be really unusual for the GM with three pawns to agree to a draw without at least testing the waters first. I’ve seen GMs play on in a lot of theoretical draws, though not the most simple/obvious ones. Witness the pain of Danya! [![](https://img.youtube.com/vi/GsFBidLyIag/maxresdefault.jpg "Knight vs 3 Pawns") ](https://www.youtube.com/watch?v=GsFBidLyIag) ---
**Author:** ![Maserschmidt](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/maserschmidt/32/18829_2.png) [@Maserschmidt](https://boards.straightdope.com/u/Maserschmidt)\ **Post date:** [December 24, 2023, 9:55pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/3 "2023-12-24T21:55:18Z")
Though the pawns weren’t on starting squares. ---
**Author:** ![Saint\_Cad](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/saint_cad/32/18907_2.png) [@Saint\_Cad](https://boards.straightdope.com/u/Saint_Cad)\ **Post date:** [December 25, 2023, 2:39pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/4 "2023-12-25T14:39:13Z")
I thought a knight and king couldn’t ever produce mate. If that’s the case what do you have to lose going for the win with pawns? ---
**Author:** ![Snarky\_Kong](https://avatars.discourse-cdn.com/v4/letter/s/a183cd/32.png) [@Snarky\_Kong](https://boards.straightdope.com/u/Snarky_Kong)\ **Post date:** [December 25, 2023, 4:23pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/5 "2023-12-25T16:23:56Z")
I don’t think this is technically true with other pieces on the board. Pretty sure firouzja lost a game because he times out and it was theoretically possible for his opponent to mate him if practically impossible. ---
**Author:** ![Maserschmidt](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/maserschmidt/32/18829_2.png) [@Maserschmidt](https://boards.straightdope.com/u/Maserschmidt)\ **Post date:** [December 25, 2023, 4:57pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/6 "2023-12-25T16:57:46Z")
I mean, hypothetically possible, but highly unlikely. Highly, highly unlikely. [![Imgur](https://i.imgur.com/Bnr7Jeq.jpeg "imgur.com") ](https://imgur.com/Bnr7Jeq) ---
**Author:** ![Saint\_Cad](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/saint_cad/32/18907_2.png) [@Saint\_Cad](https://boards.straightdope.com/u/Saint_Cad)\ **Post date:** [December 25, 2023, 5:01pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/7 "2023-12-25T17:01:19Z")
Yeah, I should have remembered smothered mates are a thing. Except in your example the white king is in check. ---
**Author:** ![Snarky\_Kong](https://avatars.discourse-cdn.com/v4/letter/s/a183cd/32.png) [@Snarky\_Kong](https://boards.straightdope.com/u/Snarky_Kong)\ **Post date:** [December 25, 2023, 5:14pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/8 "2023-12-25T17:14:35Z")
Not if the board is from blacks perspective. ---
**Author:** ![Pavelb1](https://avatars.discourse-cdn.com/v4/letter/p/439d5e/32.png) [@Pavelb1](https://boards.straightdope.com/u/Pavelb1)\ **Post date:** [December 25, 2023, 6:42pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/9 "2023-12-25T18:42:55Z")
Yeah, I was just trying to measure how ‘automatic’ this is at high levels. If Danya messes it up…then its probably not very automatic. In the game I played that made me think to ask…after I analyzed it, you have to send your knight off into the hinterlands to swing it back to get to the right spot. Something that would never occur to me. ---
**Author:** ![glee](https://avatars.discourse-cdn.com/v4/letter/g/b5a626/32.png) [@glee](https://boards.straightdope.com/u/glee)\ **Post date:** [December 25, 2023, 9:00pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/10 "2023-12-25T21:00:03Z")
As an experienced player (FIDE Master), I would be very confident of holding the draw, even against a grandmaster. The key thing would be that I know all positions where K+P v K is a draw. (Basically the attacking King needs to be in front of its pawn - and even then it can depend on whose move it is.) So I would be looking to either blockade or sacrifice my knight for two pawns, transposing into a drawn ending. ---
**Author:** ![Chronos](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/chronos/32/134_2.png) [@Chronos](https://boards.straightdope.com/u/Chronos)\ **Post date:** [December 26, 2023, 1:50pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/11 "2023-12-26T13:50:57Z")
> [@glee](#): > > As an experienced player (FIDE Master), I would be very confident of holding the draw, even against a grandmaster. You specifically would, but would you be confident of any given Master-rank player doing the same? Even at the same overall level of skill, not everyone has the same set of skills, and there are even some grandmasters whose endgame skills are lacking. ---
**Author:** ![glee](https://avatars.discourse-cdn.com/v4/letter/g/b5a626/32.png) [@glee](https://boards.straightdope.com/u/glee)\ **Post date:** [December 26, 2023, 5:40pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/12 "2023-12-26T17:40:28Z")
> [@Chronos](#): > > You specifically would, but would you be confident of any given Master-rank player doing the same? Even at the same overall level of skill, not everyone has the same set of skills, and there are even some grandmasters whose endgame skills are lacking. Yes, I would expect all such strong players to know their K+P endings. The interesting example above (post 2) was a time scramble, so I could see the games continuing under that circumstance. ---
**Author:** ![Pavelb1](https://avatars.discourse-cdn.com/v4/letter/p/439d5e/32.png) [@Pavelb1](https://boards.straightdope.com/u/Pavelb1)\ **Post date:** [December 27, 2023, 7:49pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/13 "2023-12-27T19:49:03Z")
> **[Chess: can0n vs DaleKentSams](https://www.chess.com/game/daily/598757173)** > > can0n (1097) vs DaleKentSams (1384). can0n won by resignation in 45 moves over 2 days. Click to review the game, move by move. So after blacks move on move 40, you’d expect at very high levels with plenty of time, you’d offer a draw as white? ---
**Author:** ![Snarky\_Kong](https://avatars.discourse-cdn.com/v4/letter/s/a183cd/32.png) [@Snarky\_Kong](https://boards.straightdope.com/u/Snarky_Kong)\ **Post date:** [December 27, 2023, 8:10pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/14 "2023-12-27T20:10:40Z")
Stockfish shows white as winning in that scenario ---
**Author:** ![Maserschmidt](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/maserschmidt/32/18829_2.png) [@Maserschmidt](https://boards.straightdope.com/u/Maserschmidt)\ **Post date:** [December 27, 2023, 8:45pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/15 "2023-12-27T20:45:25Z")
When I run Stockfish I get +0.23 after black’s best move, which isn’t even a quarter pawn. ---
**Author:** ![Pavelb1](https://avatars.discourse-cdn.com/v4/letter/p/439d5e/32.png) [@Pavelb1](https://boards.straightdope.com/u/Pavelb1)\ **Post date:** [December 27, 2023, 8:58pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/16 "2023-12-27T20:58:12Z")
> **[Analysis Board - 365Chess.com](https://www.365chess.com/analysis_board.php)** > > Review and analyse chess positions and games on an interactive chess board. Analyze your positions and games online with a powerful chess engine. This site i use has it at 0.00 Note, the board is not set up, I’m just showing the site. ---
**Author:** ![glee](https://avatars.discourse-cdn.com/v4/letter/g/b5a626/32.png) [@glee](https://boards.straightdope.com/u/glee)\ **Post date:** [December 27, 2023, 9:55pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/17 "2023-12-27T21:55:32Z")
Once you get to a small number of pieces (e.g. 6), switch to endgame tablebases for 100% accuracy. 😀 After 40. … Nxe4, it’s a drawn position. [KPPPvKN – Syzygy endgame tablebases (syzygy-tables.info)](https://syzygy-tables.info/?fen=8/8/5nk1/7P/6P1/6K1/5P2/8_b_-_-_0_1) You can easily use the tablebase for analysis - I reckon that spending an hour or two doing that would give you a clear idea how to play the position for either side. ---
**Author:** ![Chronos](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/chronos/32/134_2.png) [@Chronos](https://boards.straightdope.com/u/Chronos)\ **Post date:** [December 28, 2023, 2:36pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/18 "2023-12-28T14:36:45Z")
I’m a little surprised that Stockfish apparently doesn’t use endgame tablebases. ---
**Author:** ![Petek](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/petek/32/1590_2.png) [@Petek](https://boards.straightdope.com/u/Petek)\ **Post date:** [December 28, 2023, 6:16pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/19 "2023-12-28T18:16:50Z")
Chessbase sells a USB stick that contains all five-piece endings and most six-piece ones: [Endgame Turbo 5 USB flash drive (chessbase.com)](https://shop.chessbase.com/en/products/endgame_turbo_5) Stockfish will consult this database when appropriate, and will display the results. Fairly expensive (€179.90), though. ---
**Author:** ![Petek](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/petek/32/1590_2.png) [@Petek](https://boards.straightdope.com/u/Petek)\ **Post date:** [December 28, 2023, 7:00pm UTC](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091/20 "2023-12-28T19:00:19Z")
Upon viewing the Readme on this page: [GitHub - syzygy1/tb](https://github.com/syzygy1/tb) it seems that one can configure Stockfish to use all six-piece endings for free. The skills required to do so are beyond me, though, [Next page](https://boards.straightdope.com/t/chess-q-3-pawns-vs-knight/995091.md?page=2)