I was laughing so hard at what transpired at the end there. I feel like against any human, Mack would (correctly) infer that his opponent is just desperately hoping that he doesn't have an S or a blank to punish that move so they can reap the benefits themselves on their next turn. But, because it is BestBot, Mack immediately assumes this is some supergenius play where the bot wins by letting him have a 90 point play 🤣
JOGGLER is fine. I've made plays like this before. Happens more than you'd think, and I've seen Will and others make a play like this before, too. Scrabble preendgames are fun because all sorts of plays become optimal that seem nutty at first.
Ooh! I ran into that same bug when I made my Anagrams website. When certain events happened simultaneously, tiles would occasionally get 'eaten' just like the J. It was definitely a client-side bug-- information not getting updated onscreen like it should. Glad to see that even better programmers grapple with this issue.
I once took a workshop on joggling. The teacher, Matt Feldman, had set a world record by running 100 meters in 13.3 seconds while keeping 5 balls in the air. Very disappointed to see that that's not what "joggler" means.
My thoughts on the JOGGLER turn: If you hit that spot hard with a V, C or M, the bot has chances to play outbingos that leave you enough tiles on rack to win anyway. (V)ENTURIs, (M)UTINeER, (C)eEINTURE, (C)INCTURe, etc... I'm curious if the bot might calculate that opponents will want to get rid of the V more than most other letters, making the V a likelier floater to rely on. If the bot somehow gets the blank, it also has RE(F)UTING at the top. The Bot's POV shows 2 S's, 1 E and 1 blank in the bag, so there might be more outbingos when an unsuspecting player cashes in on the VIMS spot. Playing fewer tiles from the bot's POV also leaves more tiles on your rack when it outbingos. Of course why these gambly lines were chosen over other options, who knows.
Yeah well put -- the more you look at my options there, the more the logic behind JOGGLER "makes sense." I put "makes sense" in quotes since it makes sense in that I understand the idea behind the logic, but it doesn't fully make sense that it would be better than other more "normal" options. It certainly might be, but as humans we lack the ability (or would seem to) to conclusively determine which approach is better.
The bot just broke Mack hahaha. For the INBY turn, the bot is just simming, which people always assume yields the best results but often it really doesn't.
If it's a word, then it's playable! It's assumed that your opponent is mature and knows that such things aren't personal. Almost all obscenities and their derivations are currently legal: I once had FUCKWIT played against me and was very amused. A recent update to the word list did take out a lot of the slurs that were still legal. F*A*G and F*A*G*O*T were left in because they have innocuous meanings (a cigarette and a bundle of sticks, respectively), and also the dictionary considers only the two-G variant of F*A*G*G*O*T to be offensive (even if that isn't true in practice). *asterisks in case of automatic content filters
From what I checked, joggler is the second best move. The best move is what I first saw, which is Turnip. Turnip apparently wins around 48%, and Joggler wins around 45%. I can understand Turnip. Your rack is terrible, so emptying the bag to give your opponent only one chance to bingo is probably your best chance. Fishing right now just doesn't seem to accomplish anything. I could never play Joggler.
@@DU1994 You can trust sims to get close enough for human purposes. It's not going to be perfect, but the point is just to roughly approximate the best line, for entertainment reasons.
@@anewfuture I've had cases where pre-endgame sim (esp. with 1-2 in the bag) confuses 0% win for 100% win and vice versa. I really doubt TURNI(P) and (JOGGLE)R are close. Since TURNI(P) empties the bag we can calculate its winning % using macondo's peg module - in reality TURNI(P) wins only 34,24%. The results are far off from the results you've got from simming. A thorough analysis would be needed to establish which play is better. My intuition tells me (JOGGLE)R is better than TURNI(P), but I wouldn't trust it more than the sim. And there's also other plays like (D)UR(O) to consider.
words with secondary definitions were kept in, after consideration. Sadly FAGOT is used to designate a bundle of sticks, and still comes up in certain publications as such.
@@richardhudson4649 lots of people agree with that. There's also the idea that words tending to trigger groups of people, such as RAPE or GANGBANG or MURDER or LYNCH, shouldn't be in because they evoke such negative emotions. Many words cause a lot of damage, but not necessarily because they are inherently damaging.
Bestbot: “I can’t win this any other way, let’s try to overload the flesh-being’s circuits.”
flesh-being? I think we all know what Bestbot's pet name for Mack is. 0:21
I was laughing so hard at what transpired at the end there. I feel like against any human, Mack would (correctly) infer that his opponent is just desperately hoping that he doesn't have an S or a blank to punish that move so they can reap the benefits themselves on their next turn. But, because it is BestBot, Mack immediately assumes this is some supergenius play where the bot wins by letting him have a 90 point play 🤣
Haha sounds about right!
Your face when that R came down was amazing
JOGGLER is fine. I've made plays like this before. Happens more than you'd think, and I've seen Will and others make a play like this before, too. Scrabble preendgames are fun because all sorts of plays become optimal that seem nutty at first.
How nice of the bot to show you some mercy 😂
Ooh! I ran into that same bug when I made my Anagrams website. When certain events happened simultaneously, tiles would occasionally get 'eaten' just like the J. It was definitely a client-side bug-- information not getting updated onscreen like it should. Glad to see that even better programmers grapple with this issue.
38:37 omg thanks for fixing your mic!
Was sweating when your clock was ticking away while you sat there baffled 😂
I once took a workshop on joggling. The teacher, Matt Feldman, had set a world record by running 100 meters in 13.3 seconds while keeping 5 balls in the air. Very disappointed to see that that's not what "joggler" means.
Dang that's like faster than most people could run while NOT juggling, wild
My thoughts on the JOGGLER turn: If you hit that spot hard with a V, C or M, the bot has chances to play outbingos that leave you enough tiles on rack to win anyway. (V)ENTURIs, (M)UTINeER, (C)eEINTURE, (C)INCTURe, etc... I'm curious if the bot might calculate that opponents will want to get rid of the V more than most other letters, making the V a likelier floater to rely on. If the bot somehow gets the blank, it also has RE(F)UTING at the top. The Bot's POV shows 2 S's, 1 E and 1 blank in the bag, so there might be more outbingos when an unsuspecting player cashes in on the VIMS spot.
Playing fewer tiles from the bot's POV also leaves more tiles on your rack when it outbingos.
Of course why these gambly lines were chosen over other options, who knows.
Yeah well put -- the more you look at my options there, the more the logic behind JOGGLER "makes sense." I put "makes sense" in quotes since it makes sense in that I understand the idea behind the logic, but it doesn't fully make sense that it would be better than other more "normal" options. It certainly might be, but as humans we lack the ability (or would seem to) to conclusively determine which approach is better.
JOGGLER the typa play my grandma would make no cap bruh
It'd be interesting to see a Quackle sim of the JOGGLER position, to see if it's as stumped as you
Maybe the bot was like "for content let's play (JOGGLE)R "
The bot just broke Mack hahaha. For the INBY turn, the bot is just simming, which people always assume yields the best results but often it really doesn't.
on a streak now. clearly bestbot’s angle was to surprise you so much that you would forget to play and flag. almost worked too.
😮
Best bot trying to make you lose by flagging out because of what a weird move that was?
Haha I don't think it's quite capable of that!
BestBot simply ran out of juice and returned to HastyBot for that last play. Outstanding
Hahaha
Most entertaining game I’ve seen on this channel, btw in scrabble tournaments what is the etiquette for playing non PC words?
If it's a word, then it's playable! It's assumed that your opponent is mature and knows that such things aren't personal. Almost all obscenities and their derivations are currently legal: I once had FUCKWIT played against me and was very amused. A recent update to the word list did take out a lot of the slurs that were still legal. F*A*G and F*A*G*O*T were left in because they have innocuous meanings (a cigarette and a bundle of sticks, respectively), and also the dictionary considers only the two-G variant of F*A*G*G*O*T to be offensive (even if that isn't true in practice).
*asterisks in case of automatic content filters
Seems like getting up to a ~30% win% would be solid. I wasn't here for the last series, where did things end up?
42-58
I've wondered before if the Bot's logic is to try and win games on time by simply playing fast. In this case, fast and confusing!
Bot starting out offensive with this one 🔥 🔥🔥
Nah, it's just bundling some sticks :)
after "joggler" i gave mack a 0% shot of winning. thought the out bingo was assuredly coming down and the game was over.
Hahaha me too, I thought I was definitely about to get swindled
HI MACK❤
Hi!
From what I checked, joggler is the second best move. The best move is what I first saw, which is Turnip. Turnip apparently wins around 48%, and Joggler wins around 45%.
I can understand Turnip. Your rack is terrible, so emptying the bag to give your opponent only one chance to bingo is probably your best chance. Fishing right now just doesn't seem to accomplish anything.
I could never play Joggler.
You can't trust sim's percentages in these pre-endgames. Sim assumes Mack will play the top equity move, which will often not be the case.
@@DU1994 You can trust sims to get close enough for human purposes. It's not going to be perfect, but the point is just to roughly approximate the best line, for entertainment reasons.
@@anewfuture I've had cases where pre-endgame sim (esp. with 1-2 in the bag) confuses 0% win for 100% win and vice versa. I really doubt TURNI(P) and (JOGGLE)R are close. Since TURNI(P) empties the bag we can calculate its winning % using macondo's peg module - in reality TURNI(P) wins only 34,24%. The results are far off from the results you've got from simming. A thorough analysis would be needed to establish which play is better. My intuition tells me (JOGGLE)R is better than TURNI(P), but I wouldn't trust it more than the sim. And there's also other plays like (D)UR(O) to consider.
0:21 The bot not hiding it's feelings toward Mack today.
I thought slurs were removed from the word list a few years back.
not a slur
This one has an alternate meaning that’s still used in other contexts
words with secondary definitions were kept in, after consideration. Sadly FAGOT is used to designate a bundle of sticks, and still comes up in certain publications as such.
@@axcertypo This one feels so offensive that even with secondary definitions it should probably been removed.
@@richardhudson4649 lots of people agree with that. There's also the idea that words tending to trigger groups of people, such as RAPE or GANGBANG or MURDER or LYNCH, shouldn't be in because they evoke such negative emotions. Many words cause a lot of damage, but not necessarily because they are inherently damaging.
The first play is wild straight fire🔥🔥🔥🔥🔥