Scrabble GM vs. AI -- the Rematch! Game #50

Поделиться
HTML-код
  • Опубликовано: 16 ноя 2024

Комментарии • 23

  • @almightyhydra
    @almightyhydra 3 дня назад +14

    Just the casual bingo, into 7 random tiles that make a bingo, into 7 more random tiles that make another bingo, into Z bomb

  • @100beep5
    @100beep5 3 дня назад +15

    15:45 You could go for either BEDSPREAD or BEDSHEET

    • @Fexghadi
      @Fexghadi 3 дня назад

      Damn, you beat me to bedsheet :P didn't know the other one though

    • @Goryus
      @Goryus 3 дня назад +2

      GODSPEED is also good everywhere except...oh

    • @mackmeller
      @mackmeller  3 дня назад +3

      Ooh nice spot, missed that! Could totally see (B)URY for 10 here trying to draw EH for BEDSHEET, it might get blocked but seems like as good a shot as anything and in the unlikely event I do connect it scores almost 150 which is probably enough to win

  • @JohnDoe-ti2np
    @JohnDoe-ti2np 3 дня назад +3

    23:52 Older chess players will remember the interzonal tournaments of the latter half of the 20th century, which were part of the world championship cycle.

  • @cheerio662
    @cheerio662 3 дня назад +10

    As a casual viewer, the post game analysis on both wins and losses makes me feel like these games often just come down to how good both sides draw

    • @bruceyuchuanyu
      @bruceyuchuanyu 3 дня назад +2

      Well that's a sign for high level Scrabble
      Although end game blunder still happens

    • @axcertypo
      @axcertypo 3 дня назад +7

      That's unfortunate if that's your takeaway, but it makes sense. Mack doesn't do a deep analysis of his games, so he only really looks for plays that he missed, or misjudged the leave value of. As one of the best players in the world, he believes in his ability to disambiguate the pros and cons of each candidate play he sees. However, there are very few games in this series where it would be a safe bet to say Mack didn't have a chance. Since many decisions seem to resemble each other in terms of their value, a lot of Mack's choices are stylistic, especially those that favor playing more tiles in search of blanks and S's. If another top player decided to deeply analyze these games, there would probably be at least one decision per game where Mack could have (and maybe should have) chosen a different route. Because Scrabble as a game is not solved and is incredibly complex, any questionable decision has a good chance of being wrong, and who knows what would have happened if another play had been made. So as top players, it's important to be confident in our ability to find the best plays, and not every game deserves to be deeply analyzed, and it often takes an outside influence to show us where we likely erred in our games, if it's not already plainly obvious from a static analysis like the one Mack does after each game. The most important thing is to never be fully satisfied with our level of play, even if we would have probably lost despite playing slightly better than we did in any given game.

    • @sjoerdglaser2794
      @sjoerdglaser2794 3 дня назад +1

      ​@axcertypo as a novice, that was a very interesting look into high level scrabble. Thank you for writing this.
      This is also one of the reasons I love the internet. Before that, it was near impossible to find such direct takes from experts that didn't go through done kind of media filter, like an interview.

    • @cheerio662
      @cheerio662 3 дня назад

      @@axcertypo I agree with you, and I don't think that my perception is correct because I don't understand high level scrabble and I assume there are things I am missing- I was just pointing out how it appeared to me on the surface as someone who really doesn't even play scrabble

    • @quinnbartlett7233
      @quinnbartlett7233 3 дня назад +1

      @axcertypo
      But it is the truth, at the end of the day if you have 2 equally skilled players the only real factor is how the tiles are drawn.
      The more similar the skill level between the players the more the outcome is just luck, it's exactly the same in poker.
      And hey, there's an amount of luck that would let a complete novice beat Nigel Richards, astronomically unlikely, but it's possible!

  • @AmaranthRBY
    @AmaranthRBY 3 дня назад +2

    FENCIBLE was such a gut punch, that should not have been a real bingo line most of the time
    Maybe rather than TRAY you're supposed to try to stall a few turns and try to use the threat of outbingoing through the AGAR line long term? Like if you hit AUSTERE immediately, like you said in the postgame analysis, you still lose, but if you burn a few more tiles and threaten AUSTERE later that might work out? It seems reasonably difficult to block conclusively, short of the bot having the A itself. Very far shots in any case

    • @mackmeller
      @mackmeller  3 дня назад

      Possibly -- I just feel like I need to score more with my bingo than the 60-70 pointers I'll get with AGAR given the size of my deficit and that BestBot isn't going to carelessly empty the bag with HMPWY on its rack. Of course only scoring 7 with TRAY isn't great either, though, so I'm by no means saying it's definitely best. Really unpromising position either way

  • @bro91wn
    @bro91wn 3 дня назад

    Halfway through!
    I liked Y(EN) at the TR(A)Y turn to hope the bot has no vowels while keeping the B column open, but by that point it didn't matter much anyway.

    • @mackmeller
      @mackmeller  3 дня назад

      YEN isn't a bad idea at all! Anything is going to be a real long shot like you said, but that's what you've got to do when you're in these kind of desperate positions. If the bot does have no vowels or only Us as vowels (unlikely but not impossible) then it can't block and I could definitely hit a big bingo there next turn

  • @asdfasdf4924
    @asdfasdf4924 3 дня назад

    Even if the bot triple-tripled after FOLIATES you can't start assuming that playing the bingo was the wrong play

  • @JakeGayer
    @JakeGayer 3 дня назад

    “Oafiest 🫤”