Jerome Gambit: Against the Titans (Part 2)

[continued from the previous post]

perrypawnpusher - Chess Titans

Casual Game, 2019

Black's move seems to be the most natural in the world, but, after the game, Stockfish 10 preferred 13...Kf7. Can you see why? The reason is covered in the next note.

14.Qg5+ Ke8 15.Qxd8+ 

I made this move quickly, seeing an opportunity to simplify things by getting the Queens off of the board, leading to a position where I would be a pawn up.

Had I looked closer - or been less anxious - I would have found the alternative, 15.Qxg7, which would scoop up another pawn, and leads, after 15...Qf6 16.Qxf6 Nxf6, to an exchange of Queens, anyhow.

Playing against computers makes me nervous. (Against human beings, too.)

15...Kxd8 16.Rxh2

16...Be6 

Instead, 16...Nf6 17. d3 Ng4 18. Rh4 Nf6 19. Be3 Bg4 20. Bd4 Bd1 21. Na3 Be2 22. Kf2 Bxd3 23. cxd3 c5 24.Bxf6+ gxf6 25. Nc4 Ke7 26. Ne3 Kf7 27. Rah1 Kg6 28. Rh6+ Kg7 29. Nf5+ Kg8 30.Rxf6 d5 31. Nh6+ Kg7 32. Rf7+ Kg6 33. e5 Rhg8 34. Rf6+ Kg7 35. Nf5+ Kh8 36.Rxh7+ Kxh7 37. Rh6 checkmate, was Born Loser - NN, 2005.

17.Nc3 g6 18.d4 h5 

The rest of the game, starting with this move, has a bit of an odd tinge to it. Some of Black's moves are hard to fathom. Readers may also be frustrated by my sudden intense caution.

19.d5 Bf7 20.e5 dxe5 21.fxe5 

The central "Jerome pawns" give White a winning advantage.

21...c6 22.d6 

It was better to advance the other center pawn.

22...Bd5+ 23.Nxd5 cxd5 24.Be3 

Safe development, but 24.e6 was stronger.

24...Kc8 25.Rd1 a5 26.Rxd5 Nh6 27.Bxh6 Rxh6 28.Kf3 a4 29.Re2 h4 30.gxh4 Rxh4 

31.e6 Rh3+  32.Kg4 Rh4+ 

Black throws away the Rook, in order to push even a more dire position from appearing on its analysis "horizon". I have seen examples of computers tossing piece after piece, just to make an even-more-unfavorable future position "disappear", only to have it "reappear" after the opponent makes a capture. In the end, more material is lost than in the "dreaded" situation it is trying to avoid. (The similarity to human clinical depression is notable.)  

33.Kxh4 a3 34.d7+ Kc7 35.e7 Rh8+ 36.Kg4 Rh4+ 

See the previous note.

37.Kxh4 g5+ 38.Kxg5 b6 39.d8=Q+ Kb7 40.e8=Q axb2 41.Re7+ Ka6 42.Qa4 checkmate 

Jerome Gambit: Against the Titans (Part 1)

Sitting at the computer the other day, I suddenly decided to challenge one of the programs to game of chess. I played the Jerome Gambit, of course. I've done this before - see "Jerome Gambit: Artificial Piffle" and "Contempt?!" 

Quickly we raced toward one of the more deadly refutations, a trip down memory lane for me that was lined with past escapes - and non-escapes. Suddenly, my opponent came up with a novelty that undid all of its hard work. The rest of the game was a combination of me timidly moving forward, while the computer seemed to lose much sense of things. Come, see. 

perrypawnpusher - Chess Titans
casual game, 2019

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ 

4...Kxf7 5. Nxe5+ Nxe5 6. Qh5+ Ke6 

7.Qf5+ Kd6 8.f4 Qh4+ 

The first example that I have found of this move appeared in the American Chess Magazine of June 1899, in a satirical piece titled "A CHESS SCRAP" by "R.F.", which presented a supposed new-fangled thing - a chess game played over the telephone. The move is not a joke, however.

Interestingly enough, when I reported on last year's game against Chess Titans, I noted

I was pretty sure that my opponent was not seriously booked up on the Jerome Gambit, and that the killer move 8...Qh4+!? was not going to show up now.

Back then, it played the insufficient 8...g6. This time, it appeared to be better prepared.

9.g3 Nf3+ 

Black can quietly remove his Queen to f6, but the text is more aggressive.

10.Kf1

I have also played 10.Kd1: perrypawnpusher - electrahan, blitz, FICS, 2009 (1-0, 35); perrypawnpusher - Banassi, blitz, FICS, 2009 (1-0, 28); perrypawnpusher - fortytwooz, blitz, FICS, 2010 (1-0, 11); perrypawnpusher - javistas, blitz, FICS, 2010 (1-0, 11); and perrypawnpusher - thinan, blitz, FICS, 2010 (1-0, 11).

The text is best, but, it still leaves Black better.

10...Nxh2+

As I have suggested, "Black can get away with this, but it is not best."

I actually met 10...Qh3+ (likely a mouse slip) twice, perrypawnpusher - rjbriggs, blitz, FICS, 2011(1-0, 10) and perrypawnpusher - inansoy, blitz, FICS, 2014 (1-0, 51).

I have had trouble in the past with 10...Qh6 -  perrypawnpusher - Temmo, Jerome Gambit thematic, ChessWorld.net, 2008 (0-1, 43) and 10...Qf6 -  perrypawnpusher - james042665, blitz, FICS, 2008, (0-1, 18); and perrypawnpusher - HarlemKnight, blitz, FICS, 2014}(0-1, 24).

I was familiar with the text move, having found it in the game Born Loser - NN, 2005 (1-0, 37) in White's unique book, Unorthodox Chess (2005).

11.Kg2

11...Qd8 

An odd move. I have faced a number of alternatives:

11...Qg4 (probably best) in perrypawnpusher - jgknight, blitz, FICS, 2011 (1-0, 68);

11...Nf6 12.gxh4 in perrypawnpusher - mikelars, blitz, FICS 2011 (1-0, 12);

11...Qe7 12.Qd5 checkmate, perrypawnpusher - intssed, blitz, FICS, 2010 (1-0, 12);

11...Qh6 in perrypawnpusher - JTIV, blitz, FICS, 2010 (1-0, 32); 

11...Qf6 in perrypawnpusher - IlToscano, Chess.com, 2016 (1-0, 18); and

11...Ne7 in perrypawnpusher - constipatedguru, blitz, FICS, 2017 (1-0, 20)

12.Qd5+ 

This move is good and easy to find. After the game, however, Stockfish 10 recommended that I ignore both the Bishop on c5 and the Knight on h2, at least at first, and play 12.Nc3!? The point is that neither Black piece can move to safety effectively - and a couple of possible Bishop moves would lead to checkmate in 2 or 3 moves.

The computer recommends 12...Nf6, which falls right away to the fork 13.e5+. The alternative, 12...Qe8, simply allows 13.Qd5+, and the Bishop will fall next move, and the Knight soon thereafter. White is better, in any event, but the computer is always looking for a "better better."

12...Ke7 13.Qxc5+ d6 

[to be continued]

Jerome Gambit: Artificial Piffle

Okay, I admit it, I was bored. I certainly was not up for a complex, multi-layered intellectual exercise. So, seeing a chessplaying program on my computer, I thought I would while away some time with it - and the Jerome Gambit.

It turns out that I did this once before, 7 years ago - see "Contempt?!" The game does not seem to be in The Database, however. Odd.

I think Microsoft created the Chess Titans program to introduce people to chess. In a world where the world champion would have little chance of defeating top silicon challengers, it turns out that my opponent the other day was not so strong. 

perrypawnpusher - Chess Titans
casual game, 2018

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+ 

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ke6 7.Qf5+
Kd6 8.f4 

I was pretty sure that my opponent was not seriously booked up on the Jerome Gambit, and that the killer move 8...Qh4+!? was not going to show up now.

For the record, 8.Nc3 was played by my son against the program years ago, in jon - Chess Titans, 2011 (1-0, 14). That game does not appear in The Database, either. (I will remedy.)

8...g6

I suppose that this was part of the computer's programming, to keep it "weak" - occasionally producing stunners like this.

9.Qxe5+ Kc6 10.Qxh8 

Taking the Rook is best. I got a great game soon after 10.Qd5+, however, frittering it all away and barely surviving to draw in perrypawnpusher - spontex, 8 8 blitz, FICS, 2009 (1/2-1/2, 59). 

10...Qf8 

I also faced 10...d6, leading to an interesting battle, in perrypawnpusher - vermifugo, blitz, FICS, 2014 (1-0, 38)

11.d4 Bxd4 12.Qxd4 h6 

Okay, if it can be said that there was any kind of thread running through this game, it is clear that by now, CT has lost it.

13.Qd5+ Kb6 14.Be3+ c5 15.b4 Qxf4 

I suppose if Chess Titans were my client, I would diagnose clinical depression. Still, there was nothing for Black to be excited about in the computer-recommended 15...Kc7 16.Nc3 Qg7 17.Qxc5+ Kd8 18.Bd4 Qe7 19.Qa5+ Ke8 20.O-O-O. 

16.Qxc5+ Ka6 17.Qa5 checkmate

Contempt?!

If two chess players are equally matched and have fought hard in a game against each other, a draw might be a reasonable outcome.

If opponents in a game differ greatly in their strengths, the weaker player might be quite satisfied to split the point, while the stronger player might be unwilling to do so, except as a last resort.

Chess-playing programs have to take this into account: given that most of their opponents will be weaker than them, how ready should they be to accept a draw? It would be the height of absurdity if a human could open a game, for example, with 1.e4 and an offer of a draw – and the computer, seeing itself as worse off (even slightly, Black's fate) would agree to cease hostilities...

Programming a level of resistance to accepting draws is called setting its "contempt" level. Set it high enough, and the computer will play on, down a Rook or a Queen. That's contempt!

Recently I found myself away from home, staying in a hotel. While checking my email on the available computer, I noticed that it had the Chess Titans program. Time, I thought, for a quick Jerome Gambit game!

I've posted a Chess Titans game in the past (see "Artificial Ignorance" Parts 1 and 2) and the silicon beast appeared to be the perfect foil for such an offbeat gambit.

Kennedy - Chess Titans
casual game, 2011

1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Nc3 Bc5 5.Bxf7+

The Italian Four Knights Jerome Gambit.

5...Kxf7 6.Nxe5 Nxe5 7.d4 Bb4

See "Casually" for my most recent game against this line, or  "Coffee Break" for a fuller treatment. As the titles suggest, Black's defense is not the most strenuous.

8.dxe5 Bxc3+ 9.bxc3 Nxe4 10.Qd5+

10...Kg6 11.Qxe4+ Kf7

White has recovered his sacrificed material, but his pawn advantage is almost meaningless. He has to use the open lines against Black's King.

12.Qd5+

Hoping to drive the King back.

12...Kg6 13.Qe4+

At this point I was not interested in forcing a draw by a three-fold repetition of position; I was simply curious as to how the computer would see things.

13...Kh5

Wow, no draw now!

That's serious contempt.

14.Qf5+ Kh4 15.g3 checkmate

Artificial Ignorance (Part 2)

The Wikipedia entry for the computer program Chess Titans (see "Artificial Ignorance Part 1"), which comes with the Windows 7 operating system, includes this dead-pan bit of information

Chess Titans has been criticized for being less sophisticated than other chess video games. However, while its algorithms may not be as sophisticated, it is still a good game for beginners or those wanting a refresher.

Chess Titans has also been criticized for having bugs that allow the computer to cheat. However, this criticism is generally unfounded and often comes from beginners who are unfamiliar with more artisan moves such as castling and en passant.

My son was pleased that the program was willing to answer his 1.e4 with 1..e5 (I haven't shown him the "Sicilian Jerome Gambit" yet), and he quickly went about offering the computer "Jerome Gambit odds".

Jon - Chess Titans
casual game, 2011

1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+

4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ke6

Book or calculation? The computer chooses a fighting line.

7.Qf5+ Kd6 8.Nc3

Checking The Database I see that this line scored 23% in 45 games. This game gives it a nudge upward.

8...Bxf2+

This "theoretical novelty", however, restores White's chances to a-little-better-than-even. It is hard to see where the move came from, as the psychology of the "Anti-Bill Wall Gambit" and its relatives is largely a human phenomenon.

Certainly 8...Qf6, instead, was the way to go for Black.

9.Qxf2 g6

More strangeness, and now White is winning, as he quickly shows.

10.Qd4+ Ke6 11.0-0

Castling his King into safety, developing his King Rook and following the advice restrain the King, then checkmate him. It is a shame to have to point out that 11.Qd5+ first, and after 11...Kf6 then 12.0-0+ was stronger, e.g. 12...Kg7 13.Qxe5+ Nf6 and White's build-up of attack on the pinned Knight will be decisive.

11...d5

One step too far. 

12.exd5+ Ke7

Opting for a quick end.

Instead, 12...Kd6 13.Nb5+ Ke7 14.Qxe5+ Kd7 15.Qe6 checkmate lasts a bit longer.

"Best" is defending with 12...Kd7 but after 13.Qxe5 White is threatening both material gain and checkmate, and Black's survival (however long he lasts) will be a tortured one. 

13.Qxe5+ Be6 14.Qxe6 checkmate

 (Hey, Boss, look out!)

Artificial Ignorance (Part 1)

How do you get a chess-playing computer to play poorly?

In the case of programs like Talking LCD Chess – see "Jerome Gambit and the Perfesser (Part I)", Part II, Part III and Part IV – you can limit how deeply it searches for each move. It sees less, it misses more.

In the case of programs like Chess Challenger 7 – see "A Jerome Gambit 'Challenger' " –  you can limit the amount of time it spends on choosing each move. Again, playing strength has a lot to do with how far the computer "sees".

Or, when you design a program, you can have it play a relatively decent game, but every once-in-a-while have it choose the 4th or 5th or worse move choice. That's the blunder-as-a-ticking-time-bomb model: with strong players, large mistakes are rare; but with weak players things go *boom* quite regularly.

I was thinking about this "problem" (most of the time programmers are trying to make their chess engines stronger and smarter) today while discussing the Chess Titans program (which is included in the Windows 7 operating system) with my son, Jon.

The youngest of the "Kennedy Kids", home on vacation from his work in Haiti, has been spending more time on chess lately. Of course, he wanted me to show him the details of the Jerome Gambit, and of course I spent a lot of time doing so.

He wants to return to The Haitian Project, play his boss at chess, and beat him with the Jerome Gambit...