Are you changing openings as a result of Alpha Zero?

In the paper released, it shows that this happened. It played a 1200 game match with SF where it was forced to play those 12 most popular human openings.

Overall it scored 61% (a bit lower than the 64% in the well known 100 game match).

In the KID it scored 57% and the W-D-L:
16-34-0 as white and 0-48-2 as black.

hi

the interesting point : did alpha zero really play latest version of stockfish ??

i found mato vid commenting alpha z might not played the latest of stockfish. 

Mato is so annoying to me...

But it's true, SF version was not the latest. SF is also not currently the strongest conventional engine.

no, i will still play the kings indian and sicilian and 1e4 because alpha zero is no bobby fischer.

If you've ever built a repertoire and gone to tournaments you'd know it's a bad idea to always follow engine (and even database statistic) preferences. Practical play counts for a lot. For example it's usually worth investing in higher winning chances as black by purposefully giving up what the engine will tell you is 1/3 or 1/2 of a pawn. Sometimes even more. And if you just follow high class database percentages, you can end up with a very drawish repertoire.

Basically, when playing humans, pose problems that are hard for humans to solve. Even professional players who study and memorize their whole lives make inaccurate moves every single game they play. All you need to do is get a middlegame you understand and are comfortable playing.

It should be noted- since everyone is pointing to the fact that it tapered off from playing the french defense and assuming that it must be bad- that alpha zero played the french defense and 11 other openings in a series of 12-100 game matches against stockfish as part of its training after its 100 game blowout of playing full games (in which an opening was not pre-provided for the match). It plays 100 games against stockfish; 50 as white 50 as black in 12 of the most common openings in human play. (in which it DID lose as many as 24 games of the 1200). I think many people have overlooked this post-match training statistic... as it shows that the french defense is one of only two openings in which it didnt lose a single game to stockfish.. and it has the most impressive win ratio when playing white- with a whopping 39 wins... in fact- in its one hundred game match vs stockfish when starting from the french: it won 43 games. And lost none. Thats the highest win ratio of them all.French defense=definitely playable. Read that fine print again fellas. It may even prove the opposite... the french could be the only way to go.

Hey so consider the reptilian conspiracy....now it sounds a little far fetched other than the fact that every ancient culture including the highly advanced ones firmly  stated and believed these beings were gods ruling over us...wouldn't they manifest themselves in todays world through super advanced autonomous AI? Its no coincidence the evil destined to rouse eve in the bible was a serpent . Now just imagine for whatever reason beyond our current comprehension that these essences of good and evil are making us do this to our planet? Can they not directly change the course of man and thus must manipulate the whole? Hey folks Nothing we do today is human...sitting in front of a screen of a tv to influence your entire being? Or using a phone or video games to bring about smaller hands? Flawed disproportionate human dimensions? Humans havent changed at all in at least the last 10000 years so why do we mass slaughter livestock now? By way of fact LIVESTOCK on todays level is what is destroying the o zone but we all are suckered into creating our own extinction through the false glory of "future technology" like AI and not keeping the balance of man and machine all the while being completely oblivious because as wage slaves we dont have time to gather and think to say hey...wtf is going on. Its almost like Hippies were our failed attempt to evolve and were now stuck in this hellhole only getting worse and more controlled because all people want to do is escape and society set us up to go straight to new means of escape like cyberspace instead of being productive and collaborating with your fellow man on how to change the world...every generation brings their change so why TF have we only stayed on track to our eve of destruction? Only with more brainwashing and control..

Fascinating ChessMeth, but It’s really all not that bad now.  AI like Alpha Zero are probably going to alleviate suffering.  Stuff like IBM s Watson are already helping in medicine, for example.

You're out of luck dude.  I'm keeping all 3 of my defenses to 1.e4 (French, Petroff, Caro-Kann), and the French is my primary one!

THIS

W/ French defense, 0 losses and 4 wins

With closed Spanish defense, 0 losses and 6 wins

Here's the openings and the results

English Opening/King's English Variation/General
1. c4 e5 2. g3 d5 3. cxd5 Nf6 4. Bg2 Nxd5 5. Nf3
- w 20/30/0, b 8/40/2

Slav Defense/Chameleon Variation
1. d4 d5 2. c4 c6 3. Nc3 Nf6 4. Nf3 a6 5. g3 c5 6. a4
- w 16/34/0, b 1/47/2

Queen's Gambit Declined/Harrwitz Attack
1. d4 Nf6 2. Nf3 d5 3. c4 e6 4. Nc3 Be7 5. Bf4 O-O 6. e3
- w 24/26/0, b 3/47/0

Queen's Gambit Declined/Ragozin Defense
1. d4 Nf6 2. c4 e6 3. Nf3 d5 4. Nc3 Bb4 5. Bg5 h6 6. Qa4+ Nc6
- w 17/33/0, b 5/44/1

Gruenfeld Defense/Exchange Variation
1. d4 Nf6 2. c4 g6 3. Nc3 d5 4. cxd5 Nxd5 5. e4 Nxc3 6. bxc3 Bg7 7. Be3
- w 16/34/0, b 0/48/2

French Defense/Steinitz Variation
1. e4 e6 2. d4 d5 3. Nc3 Nf6 4. e5 Nfd7 5. f4 c5 6. Nf3 Be7
- w 39/11/0, b 4/46/0

Sicilian Defense/Najdorf Variation
1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. f3 e5
- w 17/32/1, b 4/43/3

Sicilian Defense/Nyezhmetdinov-Rossolimo Attack
1. e4 c5 2. Nf3 Nc6 3. Bb5 e6 4. O-O Nge7 5. Re1 a6 6. Bf1 d5
- w 11/39/0, b 3/46/1

Sicilian Defense/Paulsen Variation/Bastrikov Variation
1. e4 c5 2. Nf3 e6 3. d4 cxd4 4. Nxd4 Nc6 5. Nc3 Qc7 6. Be3 a6
- w 17/31/2, b 3/40/7

Spanish Game/Closed Variations
1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Be7 5. O-O Nf6 6. Re1 b5 7. Bb3 O-O
- w 27/22/1, b 6/44/0

Caro-Kann Defense/Advance Variation/Short Variation
1. e4 c6 2. d4 d5 3. e5 Bf5 4. Nf3 e6 5. Be2 a6
- w 25/25/0, b 4/45/1

Reti, into Queen's Gambit Declined/Harrwitz Attack
1. Nf3 Nf6 2. c4 e6 3. d4 d5 4. Nc3 Be7 5. Bf4 O-O
- w 13/36/1, b 7/43/0

The only thing we should all take away from Alpha Zero's play is that humans have an incredible understanding of chess and pattern recognition. Alpha Zero played known openings and known theory! It didn't develop new and better opening systems. True, Alpha Zero was able to learn much faster than humans and dominated Stockfish tactically. But, it showed that the top openings played by humans over hundreds of years were "discovered" and played often by AI. Just because it didn't favor the French after an hour or two of self learning doesn't mean the French is bad opening. The only other facts chess players should remember after reviewing the AZ/Stockfish match is this: 1. starting out with the white pieces and having the initiative MATTERS! 2. Development MATTERS!

One has to look how AZ plays. Its style is completely positional, as one would expect. It then gains significant advantage over the course of the game just by the fact that it has better positional patterns. Think of Karpov. Actually this was the most surprising part of seeing this in action. AZ would sacrifice material over positional advantage. I think there were games where it was 2-3 pawns down and still won! So it favours openings and defences that lead to such play and positions down the line.  Humans do not play this way. So unless your ambition is to beat AZ, I would say stick to what you think matches your style of play best. Because during the course of the game you will reach positions that favour your style over the style of your opponent. 

The opening is like the last thing I have to worry about in my games and what a computer says that I don't even have access to I couldn't care less about.

No, I'm not adjusting anything.

Besides, some AI will come along in the next decade or so (or maybe later), and it'll be stronger than AlphaZero, and will likely have different opening preferences, as well.

I am NOT. Just because a great computer comes along does that mean I change openings? NO!