Milhouse takes Quiz #10 – Double Cross

Checker expert Jim Loy has a number of quizzes on his website, including the following one that I found as part of my earlier post on the Double Cross opening:

Quiz #10 – Double Cross.

Here are the moves that milhouse chose with a hard time limit of 30 seconds per move, along with the resulting points awarded…

5-9 (+5),16-20 (+5), 8-11 (+5), 11-16 (+5), 5-9 (+5), 5×14 (+3), 7-11 (+2), 6-9 (+5), 7-10 (+10), 1-6 (+10), 9-13 (+10), 3-7 (+15), 13-17 (+10), 6×13 (+5)

The moves marked in bold are moves which differed from Jim’s main line.   Milhouse scores an impressive 95 out of a hundred.  The only move which causes it to not score a perfect 100 is the position where Milhouse chose to play 7-11 instead of the preferred 4-8 line.

Let’s look at the position…


Red to move, Jim suggested that 4-8 is the right move, Milhouse played 7-11.   Who is right?
Red to move, Jim suggested that 4-8 is the right move, Milhouse played 7-11. Who is right?

I’ve just begun to toy around with analyzing this, but both Milhouse and Cake seem to not like the response that is given in the line. Milhouse isn’t smart enough to see the line as won when it tried 7-11, but after 4-8 22-18, it sees itself as up a full man in only a 17 ply search.

I’ve sent an email to Jim Loy for more expert analysis. We’ll see what he has to say about the matter.

But still, 95/100 is pretty good. 🙂

Dusting off my checkers program Milhouse…

I noticed that Martin Fierz released a new version of his Checkerboard program, so I thought I’d set it sparring against my own program, milhouse. The Cake engine it ships with walks all over my program, but it managed this win against Simple Checkers.

[Event ""]
[Date ""]
[Black "Milhouse"]
[White "Simple Checkers"]
[Result "1-0"]
1. 9-14 23-18 2. 14x23 27x18 3. 5-9 32-27 4. 12-16 27-23 5. 16-20 23-19 6. 20x27 31x24 7.
10-14 26-23 8. 7-10 24-20 9. 8-12 28-24 10. 1-5 22-17 11. 9-13 18x9 12. 5x14 25-22 13. 6-9
23-18 14. 14x23 22-18 15. 13x22 21-17 16. 23-26 30x23 17. 3-8 17-14 18. 10x17 19-15 19.
2-7 23-19 20. 17-21 18-14 21. 9x18 15-10 22. 7x14 19-15 23. 22-25 29x22 24. 18x25 15-10
25. 25-30 10-6 26. 14-18 6-1 27. 18-23 1-6 28. 23-27 6-10 29. 27-32 10-14 30. 12-16 14-18
31. 32-28 18-14 32. 28x19 14-10 33. 30-26 10-7 34. 19-15 7-10 35. 15x6 *

Addendum: I did some quick analysis using both Cake and Milhouse this morning, trying to locate the bad moves that each engine made. I haven’t stared at it too deeply, but it appeared that Simple Checkers walked into a poor opening which I suspect that if I looked hard, I could find in one of my checkers references. Cake finds the first 5 moves in its book, but then Simple Checkers plays 5. … 23-19, which is out of book. Still, upon a 31 ply search, Cake scores the position at -126, which is over a full man down, which would be tough to overcome.


 White to move, but is already behind. Simple Checkers has walked into a poor opening.   Cake and Milhouse both agree that this position is weak, and White is already almost a full man down.
White to move, but is already behind. Simple Checkers has walked into a poor opening. Cake and Milhouse both agree that this position is weak, and White is already almost a full man down.

Addendum: As I suspected, the opening is a known weak one for White, and it begins right back at the second move of the game, although the opening is still part of the standard 3 move opening set, so some possibilities remain. The opening is called the Double Cross, and is considered weak for White (in particular, 1. … 23-18 is not considered a proper response.) Jim Loy has a nifty page full of openings with some insightful commentary, and shows a Tinsley/Fortman game where this opening was featured, with an interesting cook played by Tinsley.

Addendum: Jim Loy has some interesting analysis of the Double Cross opening here.