Ultimate Scrabble battle: Grandmaster vs. AI! Game #2
Vložit
- čas přidán 26. 09. 2023
- In this series, I'll be playing the AI-powered BestBot in 100 games of Scrabble! Who will prevail? Watch the videos to find out -- and hopefully pick up a few tips too!
BestBot is the upcoming ultimate Scrabble AI from Woogles.io, to be launched in 2024. For questions, please email woogles@woogles.io.
Want personalized help taking your game to the next level or a fun gift for a friend? Check out www.mackmeller.com/lessons! - Hry
I'm starting to think it's called BestBot because it always gets the best draws...
Probably focused more on bingo leaves while hasty usually just went by what scores the most
@@king812345678hasty does consider bingo leaves, it’s sorted by best equity play, not best score
@@king812345678 yea but it scored like 4 bingos in a quick succession which means it had 7 tiles turned over and just drew really well hahaha.
Okay, I *know* Cesar uses a true randomizer for tile draws with this bot, but with these first couple games, and after the 10 or so games I've played, it seems like it draws better than Hasty o_o and bingoes a lot more
Going on the record that I saw pomade. This channel and a bunch of sporadic practice are truly making me see words more quickly and reliably I think!!
That's awesome!
BestBot's secret is not some cutting-edge AI algorithm, but rather it peeks when picking tiles out of the bag...
My prediction: 35-65 -500 but I'm hoping it doesn't come true!
I guess Bestbot expected that VAT spot to remain open because it always assumes the highest score+rack play for the opponent
Honestly given that you didn’t make any serious mistakes in the first 2 games, the first 2 games can be chalked up to incredibly lucky draws from the bot.
I still think you win 53-47 with a leave of -100. Go humanity!!!
Fingers crossed... appreciate the optimism!
LuckyBot is on fire!
I'd like to make my prediction after 10 games😂😂
Haha all predictions must be in before game 4 is up, so use the limited sample size you'll have at your own risk! :)
I'm very proud of myself for seeing POMADE where you didn't :D . First time that happened watching one of your videos. It's actually the top move too!
one thing (r)AM does is it sets up DEF at 5G which is 34 points and still keeps Q options for the subsequent turn. Maybe the bot had a moment of clarity and realized GRIZES is getting removed from the NWL dictionary soon, quickly remembering on its next turn that it is indeed just a bot and literally doesn't care.
and I guess BEF(I)T draws into QAT again a decent amount, QUAD and QADI N11 a lot, as well as blocking JO(S)H or JU(S)T and has better timing, but I still don't think that it's correct. Complicated pre-endgame so it's probably just cutting itself off too early when simulating
I'll predict a 65:35 win for bestbot with a spread of +1500 for bestbot
2:36 you were missing your old pet dog there
hahaha I'm sure that won't be the last time!
-4 this time :)
On second thought, I’m changing my guess to 69-31 win for bestbot with a spread of 4000 (in bestbot’s favour)
Prediction:
Mack: 51 wins
Bestbot: 49 wins
Spread: +200
After dozens and dozens of videos and about two boatloads of false alarms, it finally happened: I spotted a word that Mack missed :-) Woot!
I'm going to predict a slaughter in favour of the Bot: 22-78, -6000
It's tempting to just make my final prediction and say my first thought of 25-75 is my final prediction but I could definitely see myself changing my mind one way or the other so I'll wait until game 5 to make my final prediction. It's even possible that Mack loses all 5 games handily and then I'd have to make an even direr prediction... But yeah I think it's going to be an uphill battle to even win 25% of the time in a 100 game series against this thing. I played 2 games against it myself, go watch on my channel for those games. In the second game I played, I actually was close to beating it but it made a crazy play at the end which kindof impressed me on its endgame strategy.
Just a reminder that predictions are only open through game 3 though!
@@mackmeller ok I thought you said game 5. I'll make mine on game 3 in that case.
My prediction is 43-57 -313. I think you've gotten some pretty unlucky draws over the last two games and have been playing super well, but I still think BestBot will take it by a little
RAM is probably blocking bingos front hooking ICKY and going downwards which could maybe give you comeback chances if you draw two back to back. Even after RAM you might have ways to get back to back bingos through the floaters in ALLEGERS or AIRLINE but they look much more difficult to fit on the board.
BEFIT though, I got nothing. It'll be interesting to see how it does in closer endgames and if it still does these oddball plays, or if they're just things that come up when it calculates 99.2% vs 99.5% win odds or something like that
Rough draws. I'll predict 70 - 30 +1000 in favor of BestBot
I will flip your record against HastyBot and say your record against BestBot will be 40-60 with a spread of -1200. I still have the highest regard for you as a player of the human variety.
Seems to be a very hard challenge. Sadly I think the bot will win the tournament.
Prediction. Bestbot to win 83-17 with spread +11000
I predict you will win 65-35 with a +350 point spread
10:23 IT PLAYED MY NAME?
My guess is 28-72 with bestbot winning and people are guessing far far too optimistic with the spread, i'm going to guess -6233
And good luck!
@@quinnbartlett7233 thanks!
Sorry if stupid question but what do you mean by spread?
Total score difference - spread is the number of points you win or lose by. For instance if you win two games by 50 points each but lose one game by 200 points, your record would be 2-1 (in your favor), but your spread would be -100 (100 in favor of your opponent).
Definitely not a stupid question -- below answer is exactly correct
Prediction: 50.5-49.5 -480 in Mack’s favor
49.5 + 51.5 = 101???
@@appleberry27_ my math degree really paying off here
O O O Ozempic
I'm sorry but I think Bestbot will get the better of you and the final score will be 46-54 with a spread of -1732
51-49 -100
48-52 +2
40-60 -5000
37-63 games -2473 spread. Sorry for being so pessimistic
No worries! I didn't exactly produce a performance worthy of being anything *but* pessimistic these first two games hahaha
looks like I'll be pretty spot on with my spread, but far off with my game prediction
oooo, unfortunate
69 wins for the ai
Bot wins 36-64 -895
54-46 +369
39-61, -2444
35-65 -1054
41-59 -776