AlphaGo vs. AlphaGo with Michael Redmond 9p: Game 39

Sdílet
Vložit
  • čas přidán 30. 04. 2020
  • Michael Redmond 9p, hosted by the AGA E-Journal’s Chris Garlock, reviews the 39th game of the amazing AlphaGo vs. AlphaGo selfplay games. The 50-game series was published by Deepmind after AlphaGo's victory over world champion Ke Jie 9p in May 2017.
    Chris and Michael chat about Redmond's new CZcams channel -- / @michaelredmondsgotv -- and catch up on how Japanese pros are dealing with the COVID19 shutdown.
    "In this game, you're going to be surprised a few times, when AlphaGo plays away from a position," says Redmond, "plus there's the usual stuff that it does with ladders. AlphaGo has special ways that it deals with ladders, which is something that's caught on with human players."
    Check out the sgf file -- with variations/comments -- here: www.usgo.org/news/2020/05/alp...
    Produced by Stephen Hu, Allen Moy and Chris Garlock
    Thumbnail image of Rock 'em Sock 'em Robots by Lorie Shaull - Own work, CC BY-SA 4.0, commons.wikimedia.org/w/index...
  • Hry

Komentáře • 13

  • @jacklyfgardon9254
    @jacklyfgardon9254 Před 4 lety +1

    This is still nice after 39 episode to see you both enjoying go. Thanks you for the hard work done.

  • @SamuelGrguric
    @SamuelGrguric Před 4 lety +3

    game starts around 9:00

  • @coastside
    @coastside Před 4 lety +3

    At 19:04 you say you didn't research whether black pushing at C16 and allowing white to enclose the corner at P3 would be something the computers would indicate was advantageous to white. You were right. I checked with Leela and indeed the move at C16 is considered to be a loss for white, with a drop of 5% in the win rate. That's a considerable drop which I would say means the computer thinks that would be a "mistake" for black. So answering whites pressing moves in the lower right was bigger than moving in the upper left.

    • @ConsciousBreaks
      @ConsciousBreaks Před 4 lety

      Just looking at the position, it does look better for White to me. White has completely enclosed a Black stone and has really good influence on both sides, while White's top left stone still has some potential in the future.
      Like Michael said though, it's probably not decisive even for top pros since humans can make huge mistakes.

  • @ignotasanimum
    @ignotasanimum Před 4 lety

    Michael's channel: czcams.com/channels/RJyagla1B5cxIfR4i2LdgA.html

  • @harrybaldock6010
    @harrybaldock6010 Před 4 lety

    Thanks guys, great video! Here's a link to Michael's youtube channel: czcams.com/channels/RJyagla1B5cxIfR4i2LdgA.html. I'm enjoying it a lot!

  • @cedricholle9709
    @cedricholle9709 Před 4 lety +1

    I looked up your idea n10 (move 143, ~1:07:00) with leela and after a while it finds a way to counter it: l11 o9 and now p9 seems to be much better. in this variation white can still connect at o8. so after l10 k10 m11 the move j10 is sente against the 3 stones. In your variation black could give up 3 stones and capture whites stones at n8 if white plays j10

  • @Tolux303
    @Tolux303 Před 4 lety +1

    So what is Michaels youtube channel? Link in description please

    • @ConsciousBreaks
      @ConsciousBreaks Před 4 lety +1

      Michael Redmond's Go TV:
      czcams.com/channels/RJyagla1B5cxIfR4i2LdgA.html

  • @timefororbit
    @timefororbit Před 4 lety +1

    Domo arigato, Mr AlphaGo

  • @ImTheReal
    @ImTheReal Před 4 lety

    Look out my comment about kgf file - with variations/comments - here:
    Produced by Nuno, Absolut and Omoshiroi