I read a Wikipedia article about the popularity of chess handicaps in the 18th and 19th centuries. Many master players would offer odds to a weaker player in order to convince them to wager money on a game of chess. I was curious how MadChess would fare in such conditions, so I created an opening book of typical chess odds, and ran a series of matches against elite engines.
MadChess had “move” odds in all the games. That is, MadChess played white. Two move odds means MadChess played white and made two moves to open the game. The time control was 2 minutes / game + 1 second / move.
I was surprised how well my engine performed, considering it’s a 2200 Elo engine competing with 2800+ Elo engines. I was confident it would win games with queen or two rook odds, but was surprised to see it draw and win games with only pawn and a move odds. Also, I was surprised to find Stockfish the weakest competition, while a relatively old elite engine, Sjeng, was the strongest.
| Odds | Stockfish | Rybka | Junior | Sjeng | Shredder |
|---|---|---|---|---|---|
| Two Moves (e4) | 0 | 0 | 0 | 0 | 0 |
| Two Moves (d4) | 0 | 0 | 0 | 0 | 0 |
| Two Moves (c4) | 0 | 0 | 0 | 0 | 0 |
| Pawn Odds (f7) | 0 | 0 | 0 | 0 | 0 |
| Pawn and Move (e4, f7) | 0 | 1 | 0 | 0 | 0 |
| Pawn and Move (d4, f7) | 0 | 0 | 0 | 0 | 0 |
| Pawn and Move (c4, f7) | ½ | 0 | 0 | 0 | 0 |
| Knight Odds (b8) | 0 | 0 | 0 | 0 | 1 |
| Rook Odds (a8) | 1 | 0 | 1 | 0 | 1 |
| Rook and Pawn Odds (a8, f7) | 1 | 0 | 1 | 1 | 1 |
| Two Minors Odds (b8, f8) | 1 | 1 | 1 | 1 | 1 |
| Two Minors Odds (b8, g8) | 1 | 1 | 1 | 1 | 1 |
| Two Minors Odds (c8, f8) | 1 | 1 | 1 | 0 | 1 |
| Two Minors Odds (c8, g8) | 1 | 1 | 1 | 1 | 1 |
| Rook and Knight Odds (a8, b8) | 1 | 1 | 1 | 1 | ½ |
| Queen Odds | 1 | 1 | 1 | 1 | 1 |
| Two Rooks Odds | 1 | 1 | 1 | 1 | 1 |
| MadChess Score | 9.5 / 17 | 8 / 17 | 9 / 17 | 7 / 17 | 9.5 / 17 |
I work as a Software Architect at an insurance company. We're building a modern tech stack in the cloud using C#, Azure, ASP.NET Core, Angular, TypeScript, etc.