It is me against Boris! Never a boring game. I am white.
Also below are 35 games (starting with game 23) I won against chess.com level 10, which is much harder than Sparkchess.com's Boris.
1. Boris 10-27-18 (won middle game)
6. A draw against Claire :(
Sacrifice fell short but able to draw - though down a N.
(Playing "whatever" opennings today.)
Boris just played terrible in this game.
Forgot openning above but able to win in B vs N end game.
White wins the exchange and the game.
Boris managed a perpetual check draw. Not a good opening.
Boris trades queens on move 24 but lets me have passed pawns.
I trick Boris into openning up his king's file to win a pawn.
Guardian News: Magnus missed the win today
So many errors! Black could have won.
Boris thought a long time then sacrificed its Q (???) without any mate threat.
Strategy: take N to get doubled Ps on black's KR file, move 2 N there to drive Q away, then launch attack with Rs, Q, and N. Having developed P chain on Q side to slow counterattack. They must have improved the Boris program tonight as it thinking a few seconds when in difficulty. Harder to win tonight. At one point it flashed an red AI error message, so seems they were tweaking it.
GM Alex Yermolinsky (Uncle Yermo) brings you the post-game analysis!
Guru is the highest level of SparkChess.com
If you look at the checkmate threat it missed in its analysis you will understand that engines have trouble with dark matter.
Guru crashed my Chromebook twice trying to escape the loss but Chrome recovered the game each time.
Working now Trying to defeat Chess.com level 10. Have a game now where I opened up K side with N sacrifice and also trapped a Black B to make the game even, but impossible to find any advantage for attacks. All attacks just decrease score. A lot of energy and not much fun just to get an even position. Level 10 is a very hard nut to crack as it uses Stockfish engine. But will try again tonight. 11-14-18.
Tonight I will try level 10 again.
Managed to have an advanced N, so now it is a 2 N vs. 2 B game. Suspect it will trade off a B for my N at some point. May reinforce advanced N with the other one. Then B would have no way to get rid of it. What is best move for white now?
Tried to duplicate this but Chess.com did not play the same moves.
Best to play the London openning against Chess.com 10. Gives most even game for the human player.
White was up a rook but black draws by repetition. Notice that it cares little for king protection.
No mistakes, perfectly symmetrical finish.
I rejected a Q trade. I had tiny advantage but with 2 Qs black easily drew with perpetual check.
It is difficult to actually win against chess.com 10 since it does not make blunders or tactical mistakes. It still makes strategic mistakes for tactical advantage, such as trading a R for a B, in which case you may be able to win!
Went back and changed a move in game 22 to get a winning advantage. In the end, black was reduced to a King. I love to take all the pieces.
Caruana played R d1 to prepare for Nc4. Game ended in draw. Caruana is not risking enough to win.
A drawn game. But will try again to win.
Wow, 132 moves but victory at last! The key to avoiding a draw was to exchange the light square bishops in the end game. If black is allowed to keep the B pair, it is impossible to win because of perpetual check and the checkmate threat by black. Final positioning to queen the pawn probably could have been more efficient, but I figured it out in the end. A win is a win after all.
On move 28, black can't take the bishop because of a discovered check. Cute. On move 34 black is forced to trade his rook for a bishop. Always dangerous to send your rooks downstream. On move 45, I trade Qs to use my endgame rook advantage. On move 67, I advance a pawn and sacrifice another to open up the board. On move 76. I trade white bishops to get rid of black's B-pair threat. The decisive step to preventing the draw.
Main lesson form game 25 is: do not let your opponent keep a B-pair. You cannot defeat it even if you have a R to B advantage. Try to work the game down to your R and a pawn against the computer's B . Then you can win.
Why not just let chess.com level 10 play itself and see who wins? Have tried this, but so far, it always results in a draw. Stockfish will never show you a win! This is probably because it uses the same logic for both sides. This is also the reason why you cannot improve much by just playing against yourself. You have to play other people, other programs. You have to figure out the asymetircal winning moves for yourself! Sort of like Rubik's cube, only harder.
A good exercise is to take a drawn computer game and turn it into a win. Where does the chess master intervene?
To create a win, you have to take a risk. You have to create a slightly asymmetrical position, which may even give you a slight negative score temporarily.
If you get a large enough intitial advantage, it will usually grow if you avoid errors from then on. So you will see your positive score increase. If you are losing, the negative score will grow, so just stop playing -- it is a lost game.
Two different chess engines are more likely to result in one side losing, but those games are often very strange looking. They are not necessarily great games. Watch and decide:
Looks like this will be a drawn game and match.
An interesting game, when Magnus Carlsen was only 13 years old.
Stockfish avoids my opening trap with 6. ... Nec6.
No way this line can be anything but a draw.
Traded rooks and used split pawns to win the end game with a King attack. Stockfish can't count.
Magnus tries for a win with black but result is another draw. A good game.
Played against the Sicilian Defense: Kan, Knight Variation, Wing Attack, 6.Bd3 Qb6 7.Nb3
Keys to winning this game: exchange dark square bishops (can never win endgame if black has B-pair), 18. Be7, using knights and pawns to close the game restricting black to 1/3 of board, opening up black's defense with triple threat on black knight with 52. Qc1 forcing black to give up exchange on 52. ... Rb6, capturing passed pawn on 65 Qb4, getting a protected knight entrenched on 67. Nf6, so it was a forced mate at the end, white winning! When you replay it, it looks easy, but it was difficult to figure out how to open up the Q side. Playing a high level Stockfish is very educational.
Could not get a win in this game, try as I might. Black's dark square bishop just stifles my endgame. Stockfish finally trades B's when it realizes it is one step ahead in the pawn race. And so poof, two Kings left on the board.
Revisted the French Defense to get a win for white. 12. c3 was needed to hold on to the center pawn. 23. fc3 takes with the pawn because all you need is 2 pawn on the king side when black has a double pawn. 2 can stop 3. Also I wanted to avoid R getting pinned. 30. ed4 and the K's pawn moves all the way to the Q side! 35. g4, always advance your pawns as far as you can in the end game. 46. Kb4 and the passed pawn is ready to push. 52. Kd5, now head for the king side to take all black's remaining pawns and win convincingly! Easy peasy. Don't take Stockfish's advice in the end game. It will more often than not give away your pawns and then you can only get a draw. You can consider it's recommendation but think for yourself!
Played several variations of this, all ending in a draw. White basically has a weak opening and so needs to exchange everything to get a draw. Magnus needs to come up with a much stronger opening next time.
World champions play for draws and it is considered great. A least I was playing for a win here. Maybe this game could be turned into a win with another try. Often you can try again with same strategy and get a win.
I pushed all the pawns to open up the board. King's Indian game. Got bishop ahead and used King, Bishop, and Rook against a Rook to push a pawn to victory. Black's strategy was to collect pawns and had a massive pawn advantage, but they all fell in the end. Stockfish made a mistake in end game here as alternative endings lead to a draw.
Interesting to see how to checkmate with Rook and Knight against a Bishop. Uses mate threat to capture the bishop.
An interesting game. 6. d5 is important to grab a lot of space. 8, Bg5 is a cute maneuver to tempt black to open up K-side pawns, never mind eventual capture of B on g3. Always take with h pawn. 36 ... Rf8 not clever considering good position of other white rook to take pawns. Major flaw for black is his K way of on the side, away from the pawn action! So eventually white promotes.
Not a defeat but a static position after 85 moves.
Black wins, but not so easily. There were 170 moves to checkmate. Magnus' offer of a draw was thus justified. Even though slightly ahead, Magnus had no clear path to winning. I took some of Stockfish suggestions for black, but if I took all as it would have been a draw.
Stockfish always draws when playing itself in a near even position, less than a 1.5 point advantage. If the advantage is pure material, with a 3 point advantage, such as queen vs 2 bishops, Stockfish may still draw playing itself. Most people rate a N and B as both 3. But if you have a bishop pair, that pair is probably worth 7.5, just a little less than a queen, 9. In the end game it is nearly impossible to win instead of draw if your opponent has a B-pair. By the same token, a N pair is probably worth 6.5 as they can protect each other, while a B and N is worth 6.
In this WCC game 12, Fabiano was running out of time, only 8 minutes for 9 moves, and Magnus was clearly thinking on both players time. Magnus is a much faster thinker and is confident of winning the tie break. Perhaps Magnus did not want to see Fabiano blunder because of time. Magnus is a nice guy and they are not enemies. Some people are rooting for Fabiano because he is an American. I have no such bias. I do not consider country of origin, only playing ability and courteous behavior .
Ready for another draw? https://www.chess.com/computer-chess-championship
Arriving at same pieces left over now as my game 35 above. 4:58pm EST 11-27-18
Now it looks drawn. Ethereal did not know how to win the ending. 5:15 pm
Buth stockfish trying for a win. But can it take three connected pawns with a rook?
Still a draw. 5:24 pm. Draw by repetition called 5:26 pm.
One, Two, Three...Magnus is still the champ!
So drawing the twelves games was a rational choice, as Magnus knew he could easily win the rapids. Better not to risk too much, and you do need to take risks to win at chess when your opponent is very good. Winning is everything after all if you are world champion. You do not really want to risk your title just to please the critics.
2012 CBS video: World Chess Champion Magnus Carlsen stays number one
Computer misjudges that pawn advantage will allow it to give u a piece. on 24. ... Rc4. However, even if the King takes the bishop, white still wins as below.
So this is a pretty solid won game. A good opening to play against a strong opponent if you are looking for more than a draw. Also, don't count your pawns as remaining until down to rooks.
I have been playing chess.com tactics lately, which is a lot of fun. Very tricky. Counts as incorrect if your checkmate solution is not exactly what they want. So have to figure out what they want to not get tricked.
Queen's Pawn Opening: Mason Attack. Ending is completely equal with the two kings fending off pawn attacks on opposite sides of the board. Shows how king can defend against two advancing pawns.
Hey! This stuff is not easy. But it is a lot of fun. Started with Queen Pawn Mason Attack but that was just the beginning.
Winning the end game with an extra pawn. Keep your extra pawn. The moves chess.com suggests for you seem not quite as good as what it plays against you. It gives up your pawns too easily for example. It also often suggests you retreat. The winning moves you have to figure out for yourself. Key here was pushing the "h" pawn all the way down for a mate threat with the queen. When the computer plays itself, it is always a draw. Also, did not castle as Q is usually traded. Castling just makes your K a nice target. Computers often don't castle.
In this end game the knight proved superior to the bishop. I used a straightforward strategy here of putting all my pawns on a dark squares and using the K and N to destroy all the black pawns. A beautiful ending!
A more solid opening but long end game. Black made a mistake on move 43. ... g5 otherwise would have been a draw,
This game shows the futility of trying to win if your opponent has a B-pair in the end game. Your King is immobilized.
No win found here. You can always draw by taking computer suggestions. But some games can only draw, there being no win possible. To win you do have to find at least a pawn advantage with a good position. The computer will not do this for you of course.
A very interesting game. Old Indian Defense. Don't trade automatically if you want to win! After winning the pawn on 41. Qxc5, white just wants to trade queens. As much fun as recent GM games.
Kasparov resigned after 46. a6
I let Chess.com 10 finish the game to see if Gary really would have lost. He would have. Was a lost games several moves earlier. Gary was too passive in this game and kept repositioning his knights. Once white got into the backfield, it was all over.
Analysis said I made 2 inaccuracies and 1 blunder! But I easily won. Very easy game in fact. Computer says it made no mistakes, but then why did it lose? I would say losing a pawn is a mistake. Stockfish does not care enough about its pawns. So go after them! I think 28. .. c5, pushing its pawn 2 spaces only to trade it was a mistake for black. It should have kept more pawns on the dark squares as it had a dark square bishop. White then ate the pawns on the light squares with its light square bishop.
Kasparov resigned after 19 moves. This shows the continuation to checkmate with Stockfish 9 playing both sides. Obviously Gary was trying to be too fancy here.
INACCURACY (+1.70) A better move was 7... Bd6
11. Bf4 b5
INACCURACY (+2.24) A better move was 11... Nh7
So Kasparov should have won if he had played a little bit more accurately. Players have obviously improved since 1997 due to practice with computer chess engines.
If you take the program suggestions, you can only draw as it only shows you good moves, not the winning ones.
Black trades Queen for 2 Rooks. White trades Queen for Rook, but ends up with 2 Queens to checkmate. Interesting game.
Gruenfeld Defense: Russian, Smyslov, Main Line, 9...Nb6 10.Rd1 Nc6 11.d5 Ne5
1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.Nf3 Bg7 5.Qb3 dxc4 6.Qxc4 O-O 7.e4 Bg4 8.Be3 Nfd7 9.Qb3 Nb6 10.Rd1 Nc6 11.d5 Ne5
White wins 47%, Black 22.9%
This is a pretty easy opening to play. A lot of flexibility. Following the AlphaZero idea of pushing pawns and sacrificing a night to push pawns forward.
52. Alt Ending game 51
Preemptive B g6 causes black to advance pawns weakening king side and launching futile attack with Q that backfires, losing pawns.
An interesting opening. The queen comes out.
English Opening: King's English Variation.
English Opening Symmetrical Variation
Automatically simplifying can lose you the game. Complex queen trade here.
King's Indian Defense: Petrosian Variation
Old Indian Defense
Indian Game: East Indian, Przepiorka Variation, 3...Bg7 4.Bg2 O-O 5.b3 d6 6.Bb2
King's Indian Defense
1.d4 Nf6 2.c4 g6
Tarrasch Defense: Symmetrical Variation
1.d4 d5 2.c4 e6 3.Nc3 c5 4.e3 Nf6 5.Nf3 Nc6
Black's attack failed, then just trading off pieces.
Queen's Pawn Opening: Horwitz Defense, 1.d4 e6 2.Bf4
Won the ending with equal pieces. A simple stratagem that Stockfish could not fathom.
Stockfish gives up rook exchange for protected passed pawn but then trades off its bishop and gets lost game. You can't defend against rook with just a knight!
16. f4 and 17. f5 were key moves to pursue the advantage
When Black loses the analysis always finds lots of errors.
Benko Gambit Declined
Queens'Pawn Opening: English Defense 3. e4 Bb7 4. Bd3
Video: Nice lc0 vs Stockfish game
This looks like fun: Simulation: Alexandra Botez
These are mostly games I played against the computer using premium.sparkchess.com and chess.com
Boris plays very good tactical chess. You can win however with the right spatial strategy, even without a sacrifice. Boris' end game is not quite as good as a human chess expert. Chess is an art. Chess games can be very pretty when they are well fought by both sides. Stockfish has a very good endgame but you will only draw if you blindly take it's advice.
Sparkchess says Boris is rated at 1700. Sure, this may explain why it is not so hard, but I just enjoy winning without stress! Boris is not so easy either. You cannot make any mistakes.
If you consider that even GMs make lots of mistakes, you have to admit that humans are overrated! If you win when your opponent makes a mistake, does this make you a better chess player? Not at all. So better to play against a computer that does not blunder, at least very often. I find that playing Boris improves my game and Boris is not annoying (unlike many opponents online). The good thing is that when you blunder and lose, your number does not go down, so you risk nothing! I think the whole rating thing makes chess much less fun and distorts people's view of themselves, and others.
I am very interested in the psychology of chess and its players. Chess to me needs a philosophical justification and some major rules of etiquette. Chess should not need arbiters. I see playing chess with people as a way of making friends, not enemies. Who needs more enemies?
I beat the Play Magnus app age 12 with black, which has rating of 2250, but that experience was too stressful and the small screen and harsh colors hurt my eyes, not enjoyable at all. Only crazies would want to play chess on a phone. I looked at the Play Magnus website. They are looking for a UX engineer. You would have to move to Oslo. Coldsville! Hope they find one soon and port the app it to the chromebook!!!
So far, reviewing these games, I notice that what wins is not very subtle chess, but hard hitting aggressive tactics. Whatever side the Boris puts his K on, that is where you have to attack -- with overwhelming force. In game 16, Boris keeps attacking with his queen until he gives it up. You have to use all your pieces in the attack. Concentrate fire power where the K is. Look for the mating net. Game 19 is a good example of an overpowering attack on weakened K side.
Use caution before trading a B for a N. Knights are better in the opening but bishops are more useful in the end game, especially if you coordinate a pair of them. Destroy your opponent's B-pair or you will never win the endgame! Today I watched a GM willingly give up a B for a N and he wound up with a lucky draw, being a pawn down. He also did not focus his attack where the K was.
You have to go after the K! Don't be timid. Don't just look for the draw. Just waiting for a mistake to win is not a good way. This is why it is important to practice beating an engine, because it does not make tactical mistakes.
When you play a computer engine, it is like playing Bobby Fischer in 1972. Every move counts. You cannot sit back and relax although your game will be better if you do not get nervous. Stay calm and concentrate.
Position your Ns to neutrlaize those of your opponent. Trade them off if opponent gets into your territory. Tempt your opponent to trade a B for a N. Do this by placing your Ns as forward and center as possible. The B is better in the end game. Very important to keep Boris' Ns out of your territory in the opening!
Just managed a win against chess.com level 10 using the London system. See game 23. Watch the following video to learn some tricks for this sytem.
This is good for beginners through chess.com level 10.
Do not change your move just because the computer rates your position as slightly negative. If you play by just watching score, you will always lose.
How strong is chess.com, which uses stockfish? One person calculates ratings in link below. Says level 10 is 3000+ It is certainly hard to beat, so I would believe this! This only means that you can iprove your game by learning how to beat your stockfish friend: chess.com level 10. Don't be afraid (???)
Do ratings mean anything? After all, when you play a computer you can take back moves, but then so does the computer in a sense. It sees many moves ahead, many more than a human possibly can. So it is only fair to be able to take back moves when playng a computer. Just being able to take back moves will not help you win of course. You will just get frustrated tring to play over your skill level. Even grandmasters make mistakes, which casues them to lose games and lower their ratings.
I do not play any rated games because I do not believe in assigining people numbers. I prefer to play in Washington Square park, or NYC Chinatown if playing XiangQi. I would rather play someone I know than someone on the internet, especially fast games. Speed chess just teaches you all the wrong things. If both players are good, they will play fast enough anyway. When teaching chidren chess, having a clock helps them to focus, but I ignore the time on the clock.
You can probably see some XiangQi style in Chinese GMs, such as Ding Liren, such as avoiding trading -- to play for the win. Often, it is not advantageous to accept a trade. Otherwise, it would probably not be offered.
My purpose in playing chess is purely for mental improvement. Chess will improve your concentration and thinking ability in general, ie, intelligence, and can do so for any age level. It is also fun to know people through the game, unless they try to badger or insult you of course.
In addition to playing chess engines, you should watch the youtube chess videos and maybe read a good chess book or two. There are books for XiangQi as well, written in Chinese of course, but you can follow the game notation fairly easily. Visit your local Chinatown bookstore. I still teach children Chinese Chess, and sometimes International Chess, at our Chinese School.
Here is a link for learning and improving your chess:
Have been playing "tactics" at chess.com. Good exercises but don't like that they take off points when you miss one. Makes my heart jump, so I stopped doing that. Maybe I will try them with a real board instead of just in my head (77% pass rate of the 96 puzzles attempted) to make sure I don't miss any at all. Don't like point penalties. Of course, doing them in your head is a better exercise.
Below is a picture of a young Ray Robson playing a conservative opening of Chinese Chess against Eric Spurgeon at our Chinese School, maybe 1999. Eric usually lost to Ray, who I said played chess like a "little computer." Perfect concentration as well as high intelligence. Ray loved to play chess. His dad started taking him to chess clubs and tournaments and he became a GM at a young age and plays in international chess tournaments today.
Game 66 Benoni Defense