not allowing the check(like Be6) would instantly lose the game for black(which is a deception for human eye) so Alphazero plays Bd7 to allow c5+ consolidating his pawn structure for a winning end game.
Unlikely cause even half way through the game the og alpha will have the better position it would be more interesting to swap stock fish and alpha half way through
I just wanted to tell you Agadmator that after months of watching your channel, despite never really being into Chess my whole life (I am 31), I went played my first few games on Lichess! I lost a classic game, then lost a Rapid then I won a 5+3 and did it because he didnt protect his 'back rank weakness' you are always talking about! He should watch more Agad! Its a great feeling. Pls never stop. You are possibly the best channel on YT. You are amazing. Ty.
I’ve played chess as a hobby, very sporadically, for some time now (probably ~500 games). I’m not very good, but I’ve always found your videos to really be excellent. This was the first video of yours where I was actually able to keep up with the strategic hypotheticals (“captures, captures, captures...”) and not get lost due to being out of my depth. After a lot of listening, the discussions you have are starting to make a lot more sense to me from a tactical standpoint. Just wanted to say that, as well as thank you for your analyses! You really do a great job!
I recreated this game on Lichess and looked at Stockfish 11 suggestions and evaluations. After 7:16 pawn push the evaluation was +0.8 and it was suggesting Q>f7 for black after which the evaluation was still 0.8. But instead I played the Alpha move with Knight to g5 and initially it shot up evaluation high in favour of white but slowly as it processed more it came down and came to -0.7. This is one of those instances which make me love AI engines.
Stockfish lost this more than A0 won. Game was over at 6:38 with Be2 according to current SF15. Super cool reviewing old engine games with the newest models
Instead of Rook capturing Bishop on H3, what if Bishop on F1 capture Bishop H3 ? , Bishop then to E5 would make Rook open check, and prevent the King move to G8. Else if Qeen captures the Bishop on H3, then the Rook capture Bishop on H3 and Check. When the Black King moves, White Qeen can escape the threat from F3 Hourse and if moved to H7 could give a Check.
If Bxh3 - Nxc2 Rc2 - Bd4 Bf2 - Bxf2 Rxf2 - now u get Kg7 to unpin queen as next move from white will be to move light square bishop, now black has a queen and 2 pawns (pawn on f2 will be lost i am not sure) for a bishop and a knight and the knight is still on bad spot. Now my 1st thought was if Bh3 - Nxc2 to move bishop away, you cant move to f5 e6 d7 cuz you lose a piece after black plays Bxd4 and pushes f2 ending up in worst situation. So it thought playing Bf1 could be played but this is actually just the same. Black can just play Bf1 - Bxd4+ Bf2 - Bxf2+ you block with bishop Either move now loses a piece. Black can also play Bf1 - f2+ Bxf2 - Qxh2+ Kxh2 - Rxf2+ here black is just a full rook up Conclusion: Bh3 is a blunder
Just a little explanation about the differences between Alpha and Stockfish: Stockfish operates ‘traditionally’ and checks every move and subsequent moves for viability. Alpha uses an entirely different type of algorithm which ‘thinks’ more similarly to a human because it omits moves without actually checking them for viability first. It uses a deep learning neural network which takes all the chess board data in, then a bunch of magic maths is done, then the output is the move Alpha wants to take. The ‘magic maths’ is progressively honed and refined during Alpha’s allowed training time before games.
@Rosy Afternoon actually, even with his database Stockfish still cant do much against Alpha, If you study deep learning's basics you will understand that Alpha actually just gets better simply by playing, and this fact will grants Alpha a absolute superiority over time..
@Rosy Afternoon or maybe one day we will be able to create a a perfect playing machine, just like what we already have right now in checkers, a machine that can only draw or win in a game because It already calculated the best move for EVERY single position that can occur in a game
12:20 The material balance is not one rook for a bishop and a knight, it is one rook AND three pawns for a bishop and a knight. So it's Black who's actually got a big material advantage. Agadmator tends to ignore the pawns when taking stock of the material. Same story at 14:20, "OK, you've got a bishop and a knight for my rook..." Well, for my rook and two pawns, one of them a passed e-pawn...
Those rook maneuvers to win that a-pawn were very interesting and informative! I'm looking forward to the Capablanca series, he was one of the greats of the game. His end game knowledge was unsurpassed.
Sorry Antonio. 14:40 u need the bishop to attack the base of the pawnchain, knight from d7 to c6, blocks the pawn path and defend the c8 square for the bishop to attack the pawn
At 12:33, after alpha moved his King to G7, why didn't dogfish play Bishop to G4? Seems like alpha would be unable to stop dogfish from capturing the passed pawn on F3. Surely I must be missing something simple, but I'm a novice, so please go easy on me. 😁
I was so bored asking myself where is agad why no video today so i was watching your old videos to have fun and here is notification for new one thanks
It would be interesting to see alpha zero down under(starts without knight) and see how good it is under pressure. This would certainly exhibit zeros defensive capabilities
One thing that saddens me is Chess no longer remains a sport were in TOP grandmasters are busy creating beautiful approaches to various lines but rather a sport where which grandmaster can play the most number of best recommended engine moves regularly and repeatedly?
You need more Bronstein! That guy was a hair's breadth from the WCC! If you covered matches that might be cool. But you're doing great. Love your channel.
I think it was way more than 13 moves what DeepMind calculated, probably from the time it played Ng5, it knew that one of the end game would be R vs NnB
A tiny suggestion: Could you please say when the engines are out of the book they are set up with? For people unfamiliar with the openings, this helps a lot. (Also, some openings are longer and have more theory than others).
I would be curious how many times the engines would do different moves than the openings they had to play from. Like, at move one we know they would probably choose something else, but how about at move two, move three, etc. In other words, how bad are us mere humans (but also how much there is to learn at every stage of a game).
When the end game “starts” why not Knight to B6 and then bishop to C8 to attack the base of the pawn chain? The rook would be attacking the bishop but the knight would be defending it from B6. I’m sure there’s something I’m missing I just want to know what
As a human being i would guess the reason was the protection/threats of the diagonals. Dark diagonal had the opponents black square bishop and queen nearby, while no black pawns blocking that diagonal. The light diagonal had 2 black pawns blocking any possible attack, while Stockfishs light square bishop wasn't anywhere threatening. But who knows what an AI like A0 considered.
I could imagine that Alpha Zero wanted to add protection to the g7 pawn with a discovery on the light square diagonal even though there was no immediate thread expect maybe knight h4
I can't believe I was able to find the winning move, that too for an AlphaZero game!! I mean it seem logical to save the Queen by blocking the black rook's path, since the black Queen was creating so much tension against Stockfish, and the only move available was Bh3!! I mean, I obviously couldn't predict beyond 2-3 moves, but still.. I PLAYED LIKE ALPHAZERO!!!!!
To quote a Stockfish developer: "The match results by themselves are not particularly meaningful because of the rather strange choice of time controls and Stockfish parameter settings: The games were played at a fixed time of 1 minute/move, which means that Stockfish has no use of its time management heuristics (lot of effort has been put into making Stockfish identify critical points in the game and decide when to spend some extra time on a move; at a fixed time per move, the strength will suffer significantly). The version of Stockfish used is one year old, was playing with far more search threads than has ever received any significant amount of testing, and had way too small hash tables for the number of threads. I believe the percentage of draws would have been much higher in a match with more normal conditions."
Isn't this quote referring to the first match? I'm pretty sure both Alpha and Stockfish got much more time for each move in the new, expanded match that this game is from.
@MTRredux because most people don't really understand AI/ML and are just jumping on the bandwagon to say "oh is using AI it must be the best", when in reality it may not be true.. i guess those who supports stockfish understands this
As you can see by the linked article, your statement of "fixed 1 minute/move" is wrong. It is stated, that both engines had 3 hours per game plus an additional 15s per move in the presented series (where the video is part of). Furthermore, you can see the impact of opening books and the impact of playing versus the latest available stockfish version. Those arguments are also brought up quite often, and it's interesting how stockfish is only able to win any games at all when using an opening book, cheating for the first ~10 moves. Kind of feels like an engine cheating with an engine :D www.chess.com/news/view/updated-alphazero-crushes-stockfish-in-new-1-000-game-match
The latest paper released by Deepmind is full of bizarre Stockfish moves. It makes multiple blunders in single games. You can run Stockfish for yourself. Watch GMChesspuzzler videos.
i think stockfish 9 will be same skill lvl like A0....so will theese games be even a little higher lvl? nice videos best chess channel by far despite i dont know any ^^ but also most enjoyable channel on YT. nice job
Man whenever i see human chess nowadays i cant help but think about how AlphaZero would crush any human easily. Just makes me sad, that machines can be that much better..
"Better" doesn't have to mean more enjoyable. There is still and always will be great fun to be had playing and watching chess, so don't get too down. Who knows what the chess world has in-store for us...
Alpha played against humans, and they drew the games. There's an article somewhere, but the general rundown was that Alpha doesn't necessarily know that it's stronger than its opponents and that holds it back from definitively winning games because humans can trade down pieces until a draw is forced.
@@rickrick5041 They are fun af I have to tell. I'm "teaching" (I'm about 1700 ELO) my gf chess (she is about 1200 now) and at faster time formats I always end up cracking out loud.
Agadmator, first of all I wanna point out, how much I admire your channel. Watching it daily since two years. Your analysis is always of high quality. I would even more enjoy , if you could manage to check the sound before you start the video...unfortunately it casts a damning light on you,that doesn't suit you. The rest is just great stuff, keep going!!!
Hello, i am chess beginner and i was wondering if someone can recommend THE book to learn about chess openings, and a book that teaches about basic concepts...
you say alpha saw 19 moves in the future. this game is on time, isnt it? what would happen if he has double or tripple time? i mean ha alpha the ability to see the perfect game from beginning and the only thing is that time make a wall against him? I looks like the real opponent for alpha is the clock. With black he has much more trouple than with white - i mean his attack end up like this (Rook vs Knight & Bishop). Btw, where do you get the informations about these game, about the 19-moving calculate? Greeting a follower
Me thinking: Queen b3 has a nice check after pawn c5.
Agadmator: White lost that game.
not allowing the check(like Be6) would instantly lose the game for black(which is a deception for human eye) so Alphazero plays Bd7 to allow c5+ consolidating his pawn structure for a winning end game.
😂😂😂😂
Amartuvshin Altaisaikhan mM
And this is why alpha zero is the base for an unbeatable computer chess engine. If alpha zero isn’t unbeatable by humans already
dont worry. u found the move of an engine many times smarter than any human(except for u)
The ability of Alpha Zero to sacrifice pieces to open up positions and ultimately crush its opponent is just marvellous !
Reminiscent of young Kasparov.
RK831 you mean Tal
@@iargueta1220 Tal 🔥
Unlikely cause even half way through the game the og alpha will have the better position it would be more interesting to swap stock fish and alpha half way through
@@hihi-ol5um Interesting
If you understand this game you are an excellent robot.
You can't identify all the picutres with traffic lights then.
Bullshit this is against SF 8, SF 10+ regulary crushes Leela when google deep mind doesn't cheat
I meant you'll fail captcha test where google tells you to select images with street lights
@1.3B • Views Deepmind: doesn't make any illegal moves in chess.
Guy that fights windmills: Cheater!
I'm a Pirate!
I just wanted to tell you Agadmator that after months of watching your channel, despite never really being into Chess my whole life (I am 31), I went played my first few games on Lichess! I lost a classic game, then lost a Rapid then I won a 5+3 and did it because he didnt protect his 'back rank weakness' you are always talking about!
He should watch more Agad! Its a great feeling. Pls never stop. You are possibly the best channel on YT. You are amazing. Ty.
6:10 if that a5 pawn could think: oh no i'm under attack, but i have nothing to worry, that bishop would never get me because i don't worth enough
The pawn was defended and even if it wasnt you would just pin yourself it would be a blunder
Stockfish should have kept playing, then resign on move 69.
Ah yes an intellectual
Engines can't resign
Roses are red
Violets are blue
left a good agadmator's video
to watch another good one too
I’ve played chess as a hobby, very sporadically, for some time now (probably ~500 games). I’m not very good, but I’ve always found your videos to really be excellent.
This was the first video of yours where I was actually able to keep up with the strategic hypotheticals (“captures, captures, captures...”) and not get lost due to being out of my depth. After a lot of listening, the discussions you have are starting to make a lot more sense to me from a tactical standpoint.
Just wanted to say that, as well as thank you for your analyses! You really do a great job!
I'm in the same boat. If you're on chess.com, add me. Let's play a game. I'm only ELO 700ish. Search up user: PvtGrass.
I’m glad Alpha also likes the Dutch defense
Congratulations. You're an excellent returner of books.
I recreated this game on Lichess and looked at Stockfish 11 suggestions and evaluations. After 7:16 pawn push the evaluation was +0.8 and it was suggesting Q>f7 for black after which the evaluation was still 0.8. But instead I played the Alpha move with Knight to g5 and initially it shot up evaluation high in favour of white but slowly as it processed more it came down and came to -0.7. This is one of those instances which make me love AI engines.
Stockfish lost this more than A0 won. Game was over at 6:38 with Be2 according to current SF15. Super cool reviewing old engine games with the newest models
Instead of Rook capturing Bishop on H3, what if Bishop on F1 capture Bishop H3 ?
,
Bishop then to E5 would make Rook open check, and prevent the King move to G8.
Else if Qeen captures the Bishop on H3, then the Rook capture Bishop on H3 and Check.
When the Black King moves, White Qeen can escape the threat from F3 Hourse and if moved to H7 could give a Check.
I keep asking my self the same
If Bxh3 - Nxc2
Rc2 - Bd4
Bf2 - Bxf2
Rxf2 - now u get Kg7 to unpin queen as next move from white will be to move light square bishop, now black has a queen and 2 pawns (pawn on f2 will be lost i am not sure) for a bishop and a knight and the knight is still on bad spot.
Now my 1st thought was if Bh3 - Nxc2 to move bishop away, you cant move to f5 e6 d7 cuz you lose a piece after black plays Bxd4 and pushes f2 ending up in worst situation. So it thought playing Bf1 could be played but this is actually just the same. Black can just play
Bf1 - Bxd4+
Bf2 - Bxf2+ you block with bishop
Either move now loses a piece.
Black can also play
Bf1 - f2+
Bxf2 - Qxh2+
Kxh2 - Rxf2+ here black is just a full rook up
Conclusion: Bh3 is a blunder
Just a little explanation about the differences between Alpha and Stockfish:
Stockfish operates ‘traditionally’ and checks every move and subsequent moves for viability. Alpha uses an entirely different type of algorithm which ‘thinks’ more similarly to a human because it omits moves without actually checking them for viability first.
It uses a deep learning neural network which takes all the chess board data in, then a bunch of magic maths is done, then the output is the move Alpha wants to take. The ‘magic maths’ is progressively honed and refined during Alpha’s allowed training time before games.
@Rosy Afternoon actually, even with his database Stockfish still cant do much against Alpha, If you study deep learning's basics you will understand that Alpha actually just gets better simply by playing, and this fact will grants Alpha a absolute superiority over time..
@Rosy Afternoon or maybe one day we will be able to create a a perfect playing machine, just like what we already have right now in checkers, a machine that can only draw or win in a game because It already calculated the best move for EVERY single position that can occur in a game
I love you infinitely Agad thanks for the A0 games best part of my day is your videos
12:20 The material balance is not one rook for a bishop and a knight, it is one rook AND three pawns for a bishop and a knight. So it's Black who's actually got a big material advantage. Agadmator tends to ignore the pawns when taking stock of the material. Same story at 14:20, "OK, you've got a bishop and a knight for my rook..." Well, for my rook and two pawns, one of them a passed e-pawn...
Played like AlphaZero without even pausing the video. I truly feel like an excellent subscriber.
Those rook maneuvers to win that a-pawn were very interesting and informative! I'm looking forward to the Capablanca series, he was one of the greats of the game. His end game knowledge was unsurpassed.
At 9:57 after black pawn moves to f2 why not just Bxf2 followed by exf2, Qxf2, Rxf2, Rxh5 with check, Kg8, Kxf2?
Sorry Antonio. 14:40 u need the bishop to attack the base of the pawnchain, knight from d7 to c6, blocks the pawn path and defend the c8 square for the bishop to attack the pawn
I love this series, but it would be nice to know when the TCEC opening moves are over and the engines themselves start playing their moves.
I know you had that Tal picture for ages but it is so creepy XD
Classic Tony Phrases: "although" "he will allow it" "He's not interested" "that's unfortunate"
How will Black continue attack at 10:27
afla not only calculates like machine but also thinks like a human... a total beast.
brilliant exposition of the significance of the alphazero games
At 12:30 I like bishop to G4 much better than the knight move. Impossible to defend that pass pawn for black.
Looking forward to the Capablanca series! I ordered the book by Edward Winter
Question please: What opening(s) would Alpha Zero play if given its choice? Thanks.
Queen's Gambit and English Opening.
Almost always it started with d4.
Alpha sees how to evaluate pawns more. He sees in endgame and possibilities for promotion.
What about pawn f2 check?? 11:20
Is pawn to f2+ possible at 11:34?
12:57 wasnt Bg4 by white winning the pawn on f3??
you know you are one of my favorite channels right?
God I'm so proud of myself for finding that Bishop h3. That shit made my day.
please make a video Alpha (handicap) vs Stockfish 8
handicap = without 1 pawn or knight from beginning.
Was such a game really played by Deepmind?
I don't know. But it should be played, because Alpha is too strong for Stockfish
The program is not available for public yet.
Stevie I see you in the comments of "Secular Talk"
So Google is raping Stockfish without his consent ?
it's about who won and how enjoyable the games are.
Really nice series! I was wondering if anyone knows why they use an opening book though?
Fabulous analysis...love your videos...thanks,
At 12:33, after alpha moved his King to G7, why didn't dogfish play Bishop to G4? Seems like alpha would be unable to stop dogfish from capturing the passed pawn on F3. Surely I must be missing something simple, but I'm a novice, so please go easy on me. 😁
@14:31 seems Stockfish could play Nb6 followed by Bc8 to pick up AlphaZero’s pawns
I was so bored asking myself where is agad why no video today so i was watching your old videos to have fun and here is notification for new one thanks
Ticking the "I'm not a robot" square proves you're a robot: 🗑️🗑️🗑️
Understanding the game between two chess engine proves you're a robot: 👑👑
It would be interesting to see alpha zero down under(starts without knight) and see how good it is under pressure. This would certainly exhibit zeros defensive capabilities
2:28 Now we're officially out of book.
If you ever wondered what it would be like to play against God...well, now you know.
I have some ideas as to how to defeat AlphaZero. How can I get a copy to try? I think I've identified a weakness by watching how it plays.
what software do you use to show these games?
these games are from set starting positions,... so what was the move where the engines start?
Can you flip the board if you want to annotate black pieces?
Yaay im realy excited about the capablanca series.
Now I’m starting to feel bad about Stockfish, please show a game where alpha lost
One thing that saddens me is Chess no longer remains a sport were in TOP grandmasters are busy creating beautiful approaches to various lines but rather a sport where which grandmaster can play the most number of best recommended engine moves regularly and repeatedly?
You need more Bronstein! That guy was a hair's breadth from the WCC! If you covered matches that might be cool. But you're doing great. Love your channel.
You enjoy your Friday too bro. Thanks for sharing.
I think it was way more than 13 moves what DeepMind calculated, probably from the time it played Ng5, it knew that one of the end game would be R vs NnB
Edward Winter, didn't he compose 'Frankenstein' in the 1970s?
A tiny suggestion: Could you please say when the engines are out of the book they are set up with? For people unfamiliar with the openings, this helps a lot. (Also, some openings are longer and have more theory than others).
To compensate for the fact that Stockfish was an old version, AlphaZero learned chess only for 4 hours.
4 hours in the google super computer, is not the same as 4 hours in your laptop, probably alpha has played the more matches of history
#suggestion Can you clarify where the book moves end and the engine moves begin?
I hope “Let me just return these books here” is the new “sorry about that”
what is the term when you position a bishop in front of the castled king?
Isn't there any way to program a language into these engines so they can tell us what they're thinking with each move?
Someone already did it, its called ChessCoach, but its pretty basic.
I would be curious how many times the engines would do different moves than the openings they had to play from. Like, at move one we know they would probably choose something else, but how about at move two, move three, etc. In other words, how bad are us mere humans (but also how much there is to learn at every stage of a game).
“Completely Winning” merch please
At 1 million subs agad should do more alpha zero. Just about 750k rn and growing so fast.
Such complexity. ♥️😍
At 12.46 Why stock fish did not play knight to g4 attacking f3 pawn?
10:25 why not Qh2 and sacrifice the Queen for rook and escaping the threat???
When the end game “starts” why not Knight to B6 and then bishop to C8 to attack the base of the pawn chain? The rook would be attacking the bishop but the knight would be defending it from B6. I’m sure there’s something I’m missing I just want to know what
Hi sir. Is there a match bewtween Alpha0 vs Komodo? I just wanna know...
Because
Komodo
It is both amazing and horrifying how two computer programs are able to assess and formulate plans for those positions that were never played before..
If only I could predict where I'd be 13 days ahead based on the move I make today.
what is your twitch streaming schedule?
I hope we can see some compositions on this channel. I love a good chess problem.
As a gentleman from A0 HQ said is a Tal+ karpov,how you deal with that ?
Who are you?
@@justinfoley7136 i m an excelent sub from Greece,a terrible player getting ready to work 😁
@@gidiaprovata let's play on chess.com
How can I buy those books agadmator? I don’t have any bookshop nearby who sells chess books.
At 12:47 why not bishop to g4
Can you show a game where Stockfish wins?
+
I want to see more alpha zero haha. I don't understand it but I like it.
Deja vu... I bought Winter's Capablanca book during a stroll yesterday.
Please show us a game that stockfish won! We need to know defence our selves from AI
My god, that was impressive.
10:02
After Qxh2, hg works too
I was always mesmerized by the beauty in chess but alphazero took it to another level. Just got an alphazero tattoo
Stage 2 of the Computer Chess Championship started today & Stockfish
is leading. Stockfish won Stage 1
www.chess.com/computer-chess-championship
4:20 why Alpha Zero doesn't go for h8 when it goes there later
As a human being i would guess the reason was the protection/threats of the diagonals.
Dark diagonal had the opponents black square bishop and queen nearby, while no black pawns blocking that diagonal.
The light diagonal had 2 black pawns blocking any possible attack, while Stockfishs light square bishop wasn't anywhere threatening.
But who knows what an AI like A0 considered.
I could imagine that Alpha Zero wanted to add protection to the g7 pawn with a discovery on the light square diagonal even though there was no immediate thread expect maybe knight h4
I can't believe I was able to find the winning move, that too for an AlphaZero game!!
I mean it seem logical to save the Queen by blocking the black rook's path, since the black Queen was creating so much tension against Stockfish, and the only move available was Bh3!!
I mean, I obviously couldn't predict beyond 2-3 moves, but still.. I PLAYED LIKE ALPHAZERO!!!!!
You made one move you aren’t good
As one of the subscribers who asked for an alpha with black game ... Thanks !!!
To quote a Stockfish developer:
"The match results by themselves are not particularly meaningful because of the rather strange choice of time controls and Stockfish parameter settings: The games were played at a fixed time of 1 minute/move, which means that Stockfish has no use of its time management heuristics (lot of effort has been put into making Stockfish identify critical points in the game and decide when to spend some extra time on a move; at a fixed time per move, the strength will suffer significantly). The version of Stockfish used is one year old, was playing with far more search threads than has ever received any significant amount of testing, and had way too small hash tables for the number of threads. I believe the percentage of draws would have been much higher in a match with more normal conditions."
Isn't this quote referring to the first match? I'm pretty sure both Alpha and Stockfish got much more time for each move in the new, expanded match that this game is from.
@@DaCarnival Yeah, old quote.
@MTRredux because most people don't really understand AI/ML and are just jumping on the bandwagon to say "oh is using AI it must be the best", when in reality it may not be true.. i guess those who supports stockfish understands this
As you can see by the linked article, your statement of "fixed 1 minute/move" is wrong. It is stated, that both engines had 3 hours per game plus an additional 15s per move in the presented series (where the video is part of).
Furthermore, you can see the impact of opening books and the impact of playing versus the latest available stockfish version. Those arguments are also brought up quite often, and it's interesting how stockfish is only able to win any games at all when using an opening book, cheating for the first ~10 moves. Kind of feels like an engine cheating with an engine :D
www.chess.com/news/view/updated-alphazero-crushes-stockfish-in-new-1-000-game-match
The latest paper released by Deepmind is full of bizarre Stockfish moves. It makes multiple blunders in single games. You can run Stockfish for yourself. Watch GMChesspuzzler videos.
i think stockfish 9 will be same skill lvl like A0....so will theese games be even a little higher lvl? nice videos best chess channel by far despite i dont know any ^^ but also most enjoyable channel on YT. nice job
Man whenever i see human chess nowadays i cant help but think about how AlphaZero would crush any human easily.
Just makes me sad, that machines can be that much better..
"Better" doesn't have to mean more enjoyable. There is still and always will be great fun to be had playing and watching chess, so don't get too down. Who knows what the chess world has in-store for us...
Since 1997 the machines are better ask Gasparov.
Alpha played against humans, and they drew the games. There's an article somewhere, but the general rundown was that Alpha doesn't necessarily know that it's stronger than its opponents and that holds it back from definitively winning games because humans can trade down pieces until a draw is forced.
@@brianfrommars So let's watch 1200 players' games
@@rickrick5041 They are fun af I have to tell. I'm "teaching" (I'm about 1700 ELO) my gf chess (she is about 1200 now) and at faster time formats I always end up cracking out loud.
What are the must read chess books ??
We want to see MORE ALPHA!!!!
Thanks for references
6:20 what just did happen!?
Agadmator, first of all I wanna point out, how much I admire your channel. Watching it daily since two years. Your analysis is always of high quality. I would even more enjoy , if you could manage to check the sound before you start the video...unfortunately it casts a damning light on you,that doesn't suit you. The rest is just great stuff, keep going!!!
Please show us from winning side player
Hello, i am chess beginner and i was wondering if someone can recommend THE book to learn about chess openings, and a book that teaches about basic concepts...
you say alpha saw 19 moves in the future. this game is on time, isnt it? what would happen if he has double or tripple time? i mean ha alpha the ability to see the perfect game from beginning and the only thing is that time make a wall against him? I looks like the real opponent for alpha is the clock. With black he has much more trouple than with white - i mean his attack end up like this (Rook vs Knight & Bishop). Btw, where do you get the informations about these game, about the 19-moving calculate? Greeting a follower