Subject: Re: Rodent games on Pi Fri Jan 15, 2021 4:32 am
Rodent IV v0.32 Tal vs. Saitek RS ChessMaster 2
One more......Final Rod4 personality, Tal. A quick win, but Rodent4 move 12. .. Nxf2? looks very suspect....it only works if White later blunders with 14. Kf1???? Game Over (Kf1 should maintain a +2 White lead). So, RSCM obliged and blundered with 14. Kf1....mate at move 28.
[pgn] [Event "30 sec/mv"] [Black "Rodent IV v0.32 Tal Brain"] [Date "Jan 15, 2021"] [Result "0-1"] [Round "Another strange middlegame"] [Site "Rodent= NO book"] [White "RS ChessMaster #2"]
Subject: Re: Rodent games on Pi Sat Jan 16, 2021 4:14 am
Rodent IV v0.32 1400 vs. Saitek Turbo Advanced Trainer
Wanted to exercise an old TAT and pitted it against Rod4 @1400. TAT is a GK2000 clone somewhere in the mid 1900 Elo range on Aktivschach. The first thing that happened is Rod4 1400 has no book, so it calculated a kind of Scandinavian and spent a lot of time moving the Queen around early (definitely old school tabletop chess). This kept TAT in the game for 15 moves. But, TAT 15. Ne4? begins a downward slide that by move 30 (or maybe move 40) is Game Over.....this is Game Over to Rod4 @1400 Elo! However, at level 1400, Rod4 simply has been dumbed down so much that it can not win ....a winning B+3Ps position. After 52. .. Kh4?? the game is drawn. But, I went to move 75 and then pulled the plug. [pgn] [Event "30 sec/mv"] [Black "Rodent IV v0.32 1400 Brain"] [Date "Jan 15, 2021"] [Result "1/2-1/2"] [Round "TAT draws 1400 level"] [Site "Rodent= NO book"] [White "Saitek Turbo Adv Trainer"]
Subject: Re: Rodent games on Pi Sat Jan 16, 2021 10:33 pm
Piman wrote:
Rodent IV v0.32 Spitfire vs. GK2000
Spitfire is problematic (replay below game):
With mate in 5, Rod4 Spitfire locks up the Pi Rodent4 Spitfire should have actually LOST to old tabletop GK2000
A QGD: Semi-Tarrasch and after Spitfire 14. .. f5?, SF evaluator actually gives White a +3.2 lead....which is certainly enough for a draw and maybe even a win for White. The GK2000 is mid 1900's in ELO and eventually loses to the superior Rodent4 endgame (GK blunder 32. Na3?? will lose a Knight and the game), however a stronger tabletop (after 14. .. f5?) would have at least drawn.
So, something about the Spitfire personality makes it remarkably weaker than the other Rodent4 personalities....Spitfire is not like the others. Also, this Pi lock up, now convinces me that Rodent4 is NOT FF on the Pi. This game shows clearly that Rodent4, on occasion, still locks up the Pi, when "mate in x" is valid. And this time, I assured that the "fully charged", Green LED was lit, on the DGT eBoard, from the jump.
[pgn] [Event "30 sec/mv"] [Black "Rodent IV v0.32 Spitfire Brain"] [Date "Jan 14, 2021"] [Result "*"] [Round "Spitfire locks up Pi"] [Site "Rodent= NO book"] [White "GK2000"]
Although black has a completely won position, the rules of computer chess tournaments are that if a computer cannot generate a move, or generates an illegal move, it forfeits the game, so the score should show a win for white.
Good rule for life: if you cannot play the game well, then maybe you can be the last man standing!
Piman likes this post
Piman
Posts : 114 Join date : 2020-11-23
Subject: Re: Rodent games on Pi Sat Jan 16, 2021 11:51 pm
TheSelfImprover wrote:
Although black has a completely won position, the rules of computer chess tournaments are that if a computer cannot generate a move, or generates an illegal move, it forfeits the game, so the score should show a win for white.
Hello,
I am absolutely amazed that a chess playing "Turk" sits in my basement. It does not require a midget and many of the programs are stronger than the any human chess player, ever.....and....it's in my basement.
But it is not just the "computer ". There are 3 basic components:
1. Program (some call it an engine) 2. computer (call it a brain or microprocessor) 3. The compile for that language (necessary for the processor to interpret program instructions)
On the Pi, as you've seen, there are occasional problems (very, very small percentage). One of the 3 above components (or a combination) have failed when there is a problem. I find the occasional problem to be a very small price to pay to have "magic" in my basement.
I am appreciative to all the the programmers and the Pi compiler (he knows who he is ) who have, and continue, to make this happen. I do not have the technical know-how or even the desire anymore to debug the reason for Pi lock up. I just deal with it.
As to tournaments, I know of at least one (TCEC) that awards wins either by evaluation or endgame tablebase....so in the above Rodent4 game, Black would be given a point not long after move 32.
TheSelfImprover
Posts : 3095 Join date : 2020-11-18
Subject: Re: Rodent games on Pi Sun Jan 17, 2021 12:20 am
Piman wrote:
As to tournaments, I know of at least one (TCEC) that awards wins either by evaluation or endgame tablebase....so in the above Rodent4 game, Black would be given a point not long after move 32.
It was a quip really, not a serious complaint. However, at the 2000 world computer chess championship at the Alexandra arena in London, I operated Crafty, and in one game my opponent's computer made an illegal move. The referee awarded the game to Crafty. I also sat opposite Ed and played against Rebel in that tournament (Rebel won - grrr! ).
Mclane and Piman like this post
Piman
Posts : 114 Join date : 2020-11-23
Subject: Re: Rodent games on Pi Sun Jan 17, 2021 3:14 am
TheSelfImprover wrote:
also sat opposite Ed and played against Rebel in that tournament (Rebel won - grrr! ).
Its been a long time ago, but I do believe Rebel Decade was the first program for PC (heavens to Betsy...a "non dedicated" ) that I ever used. Excellent program (where's the thumbs up emoji?)
Piman
Posts : 114 Join date : 2020-11-23
Subject: Re: Rodent games on Pi Mon Jan 18, 2021 12:26 am
New Personalities! (Scally)
Rodent4 Topalov vs. Excalibur GM @29.5 MHz
A surprising game: sub 2000 Elo, Excalibur GM goes toe to toe with "Topalov" for over 40 moves and even had +2 leads at move 19 (but couldn't calculate 19. Rxc7!) and move 36 (but couldn't calculate 36. Ne1!). Topalov is a very "daring" personality...sacrificing a Knight for 2 Pawns and a much stronger pawn formation...early in the middlegame. At move 39, SF eval still gives White a +1 advantage, BUT Black has those passed and connected Queenside pawns...and those two will eventually doom White. As stated, GM29.5 calculated well for over 40 moves, but then the Excalibur Achilles Heel was exposed: the endgame. SF says "45. Qe2"....when GM29 calculated 45. Qc1??, SF says "Game Over"...Then, 52. Bd4? is mate in 10.
And...Rod4 Topalov is FF to mate on the Pi. [pgn] [Event "30 sec/mv"] [Black "Rodent IV v0.32 Topalov Brain"] [Date "Jan 17, 2021"] [Result "0-1"] [Round "Topalov wins in endgame"] [Site "Rodent= NO book"] [White "GM 29.5 MHz"]
Question for the readers: What methodology gets the pgn to display evaluation in the move list (I've seen this on a recent post, elsewhere...please don't tell me that evals must be entered manually).
Sc@lly likes this post
Piman
Posts : 114 Join date : 2020-11-23
Subject: Re: Rodent games on Pi Mon Jan 18, 2021 1:22 am
Rod4 Tarrasch vs Exc. GM24 MHz
I believe the quickest computer chess game that I have played: 18 moves. GM gets caught with an embarrassing smothered mate:
Subject: Re: Rodent games on Pi Mon Jan 18, 2021 4:41 am
OK...so I'm sick in the head: one more tonight:
Rod4 Swapper vs Fidelity 12B Turbo 8 MHz overclock
An oldie: Fidelity 12B, but with a 4x overclock upgrade board (Elo estimate~1947) Cheeky 12B Turbo selects a King's Gambit ....Swapper turns it into a KGA: Fischer. Swapper dismantles the White position early and SF eval says 12B move 15. Na4? is Game Over. Fine attack by Swapper. Both computers: FF to mate.
Subject: Re: Rodent games on Pi Tue Jan 19, 2021 2:12 am
Rod4 Steinitz vs Exc. GM27 MHz (overclocked)
Although the game took 50 moves, Rod4 Personality Steinitz had a win by move 25. An Alapin Sicilian (2. c3), because ALL Excalibur computers are programmed to play 2. c3....in response to a 1. e4 c5.... every.... single.... time; shoddy, boring, third-rate Sicilian book.
The only other comment concerns Rod4 Steinitz, which at mate in 2...locked up the Pi. Steinitz found mate at about move 46...and then began moving instantly. Then, suddenly, at move 50, Steinitz didn't move instantly......the Pi clock started counting down, like Steinitz was "thinking". I immediately knew that the Pi would lock. It did. No biggie...just documenting what happened.
[pgn] [Event "30 sec/mv"] [Black "Rodent IV v0.32 Steinitz Brain"] [Date "Jan 17, 2021"] [Result "*"] [Round "Steinitz wins, but locks Pi"] [Site "Rodent= NO book"] [White "GM 27 MHz"]
Subject: Re: Rodent games on Pi Wed Jan 20, 2021 12:36 am
Rod4 Spassky vs Exc. GM28 MHz (overclocked)
Spassky plays slow and steady, but is helped along by GM28 with a Queenside castle right into Rodent4's wheelhouse. By move 21, GM was already dead. Rod4 Spassky marched patiently to mate...no Tal here (although 36. .. RxN! was enterprising). Rod4 personality Spassky was FF to mate on the Pi.
[pgn] [Event "30 sec/mv"] [Black "Rodent IV v0.32 Spassky Brain"] [Date "Jan 19, 2021"] [Result "0-1"] [Round "Spassky, slow and steady, wins"] [Site "Rodent= NO book"] [White "GM 28 MHz"]
Subject: Re: Rodent games on Pi Thu Jan 21, 2021 2:32 am
Rod4 Reti vs Excalibur GrandMaster
A new one for me : QGD Alatortsev (3. .. Be7) and both computers calculate that line for 9 full moves. GM, always greedy, then calculates 10. Qb3 and grabs the pawn, next move 11. QxP. Apparently, personality Reti is the "cautious" type and for 18 moves Excalibur Grandmaster is able to keep up. But, 19. Rd1? (SF says safety move Qa4 or maybe counter with Ne4) and GM is starting down the "slippy" slope. Then, it gets downright slippery after 22. Nc3??, in fact, Game Over. 24. Qf5? is mate in 7.
Rod4 Reti is FF to mate on the Pi....cautious, but strikes when the iron gets hot.
[pgn] [Event "30 sec/mv"] [Black "Rodent IV v0.32 Reti Brain"] [Date "Jan 20, 2021"] [Result "0-1"] [Round "Reti pounces on ExGM mistakes"] [Site "Rodent= NO book"] [White "Exc. GM stock"]
Last edited by Piman on Thu Jan 21, 2021 2:06 pm; edited 1 time in total
Sc@lly likes this post
TheSelfImprover
Posts : 3095 Join date : 2020-11-18
Subject: Re: Rodent games on Pi Thu Jan 21, 2021 9:33 am
Piman wrote:
Rod4 Reti vs Excalibur GrandMaster
A new one for me : QGD Alatortsev (3. .. Be7) and both computers calculate that line for 9 full moves. GM, always greedy, then calculates 10. Qb3 and grabs the pawn, next move 11. QxP. Apparently, personality Reti is the "cautious" type and for 18 moves Excalibur Grandmaster is able to keep up. But, 19. Rd1? (SF says safety move Qa4 or maybe counter with Ne4) and GM is starting down the "slippy" slope. Then, it gets downright slippery after 22. Nc3??, in fact, Game Over. 24. Qf5? is mate in 7.
Rod4 Reti is FF to mate on the Pi....cautious, but strikes when the iron gets hot.
[pgn] [Event "30 sec/mv"] [Black "Rodent IV v0.32 Reti Brain"] [Date "Jan 20, 2021"] [Result "0-1"] [Round "Reti pounces on ExGM mistakes"] [Site "Rodent= NO book"] [White "Exc. GM stock"]
There was a typo in the PGN: I'm just correcting it so that I can see the game.
Piman likes this post
Piman
Posts : 114 Join date : 2020-11-23
Subject: Re: Rodent games on Pi Thu Jan 21, 2021 2:10 pm
TheSelfImprover wrote:
There was a typo in the PGN: I'm just correcting it so that I can see the game.
Thanks. I missed that...I just hit "send" after cut and paste. don't know how those letters got there in the header
Rod4 will have a couple more easy tabletops (they need the buttons exercised)
Then CM9000 will get its shot: Rodent4 personalities will have a tough time then
Piman
Posts : 114 Join date : 2020-11-23
Subject: Re: Rodent games on Pi Thu Jan 21, 2021 11:06 pm
Rod4 Petrosian vs Excalibur Ivan 12 MHz version
First and foremost: A humble apology to Excalibur/ R. Nelson for the false statement that Excaliburs only play the Alapin Sicilian. My brain is starting to go ...and fast. No excuses: GM 24 played a Rossolimo previously and below Ivan plays the Sozin.
"Cautious and Defensive" personality Petrosian, slowly builds up a lead against Ivan, finally winning a pawn at move 22. Ivan blunder 29. Be3?? is Game Over. Rod4 personality Petrosian is FF to mate on the Pi.
[pgn] [Event "30 sec/mv"] [Black "Rodent IV v0.32 Petrosian Brain"] [Date "Jan 21, 2021"] [Result "0-1"] [Round "Petrosian slogs to a win"] [Site "Rodent= NO book"] [White "Exc. Ivan Decal 12 MHz"]
Subject: Re: Rodent games on Pi Fri Jan 22, 2021 4:45 am
Rod4 Nimzowitsch vs Excalibur Ivan, original 10 MHz version.
A Semi-Slav and old Ivan did very well for nearly 20 moves. But, 20. Be7? misses crisp rejoinder 20. .. Qc6! which protects the Bishop, and looks very dangerous on the diagonal for the White King. By move 28, Ivan is "panicking" and tries an exchange loss to relieve pressure 28. RxB.....but SF says that's: Game Over. Excalibur tries cheeky 43. Rd1, because 43. .. PxR=Q is a draw. Rod4 respectfully declines with 43. .. PxR=R.
Rod4 Nimzowitsch is FF to mate on the Pi....albeit, a somewhat conservative Nimzowitsch.
[pgn] [Event "30 sec/mv"] [Black "Rodent IV v0.32 Nimzowitsch Brain"] [Date "Jan 21, 2021"] [Result "0-1"] [Round "Nimzo eventually wears down Ivan"] [Site "Rodent= NO book"] [White "Exc. Ivan Old 10 MHz"]
Subject: Re: Rodent games on Pi Sat Jan 23, 2021 12:49 am
Rod4 Morphy vs Excalibur Igor
My last tabletop opponent...and Rodent4 is gonna be in real trouble vs CM9000 (next) if this Morphy game is any indication.
A Nimzo-Indian and by move 16...dusty, old 1800 Igor has a +2 eval lead . By move 30, Igor was still in it, but couldn't figure out how to defend the Queenside with 30. Rb3. After 30. Bb2?, things went downhill fast for Igor; another goofy Bishop move 36. Ba3?? is Game Over. Morphy doesn't mate until move 60....not a strong performance. As stated, next up is CM9000 to take a look at the remaining Rodent4 GM emulations (Marshall, backwards down to Alekhine).
Another FF performance to mate on the Pi by Rodent4 .
[pgn] [Event "30 sec/mv"] [Black "Rodent IV v0.32 Morphy Brain"] [Date "Jan 22, 2021"] [Result "0-1"] [Round "Morphy wins, but not too swift"] [Site "Rodent= NO book"] [White "Exc. Igor "]
Subject: Re: Rodent games on Pi Sun Jan 24, 2021 1:50 am
Rod4 Marshall vs CM9000 GM
Rodent4 is not in Kansas anymore: Tabletops are gone and CM9000 is "in the house". Below, a Najdorf Sicilian and "Marshall" is already in trouble after 17. NxN? (SF says Bd4). SF also says 23. Rce1?? is Game Over, but CM misses the strongest continuation 23. .. a5! However, 31. Qh4? is the end as a series of trades leaves the White position hopeless. Both the pinned Rook AND the trapped Knight will be lost; mate at move 55. CM9000 has weaker personalities and I will start to dumb down the strength and eventually get to an equilibrium point. That being said, the CM9000 GM win, though not perfect, was outstanding.
Subject: Re: Rodent games on Pi Sun Jan 24, 2021 5:28 am
Laskerfest 2021 Rodent4 personality Lasker vs CM9000 personality Lasker
And....Lasker wins! A terrific battle...a semi-Slav and Rod4 Lasker sacs the "a" pawn early.... I thought CM9000 would punish that insolence... BUT, incredibly Rod4 never surrendered the initiative! In fact, after CM Lasker 30. .. Nf8?, SF eval says Game Over! However, Rod4 can't get up the nerve for 31. Bxh6! And then, the next 35 moves was basically CM consistently wiggling out of certain mate as Rod4 couldn't find the killer move. Incredibly, at move 67, CM still has a draw with the mind boggling 67. .. a2!! (In a very difficult predicament with only 30 seconds/move)... but, calculates the "Hail, Mary" move 67. .. Qe6? That's Game Over. An absolutely outstanding game; the best game of all of the Rodent4 games that I have posted.
[pgn] [Event "30 sec/mv"] [Black "CM9000 Lasker"] [Date "Jan 23, 2021"] [Result "*"] [Round "Rod4 Lasker wins, but locks up"] [Site "Rodent= NO book"] [White "Rodent IV v0.32 Lasker Brain"]
One final note: Rodent4 Lasker locked up the Pi @ mate in 3....a minor issue, in light of the great game.
TheSelfImprover and Sc@lly like this post
Piman
Posts : 114 Join date : 2020-11-23
Subject: Re: Rodent games on Pi Mon Jan 25, 2021 12:08 am
Larseny
Rod4 Larsen vs CM9000 Larsen
When you see this type of matchup, your gonna get a weirdo opening, right? Right: A Wade Defense.
Another great Personality emulation game: wacky early as CM Larsen gives up Q for R+N, and a small disadvantage. But, by move 25 CM has virtually equalized....looked like it would be a draw. CM does not correctly calculate the danger of its back rank at move 30, pawn grabbing with Rxa4?? instead of the prudent g5. The CM Kingside seemed to look OK, but Rodent4 Larsen steadily dismantled it with pawns then Q+R sorties. CM 34. .. Nd7? was the end.
A very entertaining game. Rodent4, to my surprise, has taken a 2-0 lead in "head to head" personality match ups vs CM9000 (although I could swear that I played a "same" personality game in which CM won previously ).
Subject: Re: Rodent games on Pi Mon Jan 25, 2021 8:48 pm
Korchnoi
Rod4 Korchnoi vs CM9000 Korchnoi
A quick draw in a wild French: MacCutcheon. CM9000 Korchnoi sacs 2 pawns for speedy development and exposure of the White King. In the end, Rod4 Korchnoi refuses to put the Queen on g6 to continue the battle....first draw, in this series.
[pgn] [Event "30 sec/mv"] [Black "CM9000 Korchnoi"] [Date "Jan 25, 2021"] [Result "1/2-1/2"] [Round "We have a draw"] [Site "Rodent= NO book"] [White "Rodent IV v0.32 Korchnoi Brain"]
Subject: Re: Rodent games on Pi Mon Jan 25, 2021 9:19 pm
Piman wrote:
Korchnoi
Rod4 Korchnoi vs CM9000 Korchnoi
A quick draw in a wild French: MacCutcheon. CM9000 Korchnoi sacs 2 pawns for speedy development and exposure of the White King. In the end, Rod4 Korchnoi refuses to put the Queen on g6 to continue the battle....first draw, in this series.
[pgn] [Event "30 sec/mv"] [Black "CM9000 Korchnoi"] [Date "Jan 25, 2021"] [Result "1/2-1/2"] [Round "We have a draw"] [Site "Rodent= NO book"] [White "Rodent IV v0.32 Korchnoi Brain"]
Very surprising (to me) that both machines took the draw there: both must have felt that making a different move would put them at a serious disadvantage.
Piman likes this post
Piman
Posts : 114 Join date : 2020-11-23
Subject: Re: Rodent games on Pi Mon Jan 25, 2021 9:48 pm
TheSelfImprover wrote:
Very surprising (to me) that both machines took the draw there: both must have felt that making a different move would put them at a serious disadvantage.
i do believe Rodent4 "saw" something bad in the future with Qg6.
I seem to recall another wild opening on the Queenside where a computer calculates Qxb7....and then can't get the Queen out anymore: Draw. If I find that one, I'll post it
Piman
Posts : 114 Join date : 2020-11-23
Subject: Re: Rodent games on Pi Tue Jan 26, 2021 1:32 am
Kasparov
CM9000 Kashdan vs Rod4 Kasparov
Wanted to play a full Kasparov game...but CM9000 has no personality Kasparov (can't find the reason on the internet...seems strange? Maybe somebody knows?), so strong Isaac Kashdan on CM9000, subbed. A draw for 37 moves, as Kashdan went toe to toe with Kasparov: Kasparov is up a pawn...... but, it is doubled and Kashdan has an advanced "a" pawn as compensation.. But SF eval says Kashdan Kingside foray 38. f4? is a mistake (SF says Bb4). In 10 moves, that "a" pawn is gone and Rodent4 Kasparov has a totally winning R+3P vs R+P position. Good win for Rodent4 on the Pi.
[pgn] [Event "30 sec/mv"] [Black "Rodent IV v0.32 Kasparov Brain"] [Date "Jan 25, 2021"] [Result "0-1"] [Round "Kasparov outplays Kashdan in endgame"] [Site "Rodent= NO book"] [White "CM9000 Kashdan"]
Home 
). This kept TAT in the game for 15 moves. But, TAT 15. Ne4? begins a downward slide that by move 30 (or maybe move 40) is Game Over.....this is Game Over to Rod4 @1400 Elo! However, at level 1400, Rod4 simply has been dumbed down so much that it can not win ....a winning B+3Ps position. After 52. .. Kh4?? the game is drawn. But, I went to move 75 and then pulled the plug.
) who have, and continue, to make this happen. I do not have the technical know-how or even the desire anymore to debug the reason for Pi lock up. I just deal with it.
) that I ever used. Excellent program (where's the thumbs up emoji?)
).
....Swapper turns it into a KGA: Fischer. Swapper dismantles the White position early and SF eval says 12B move 15. Na4? is Game Over. Fine attack by Swapper.