|
People keep saying Kyunghoon solved monorail but I've been looking at his suggested opening move and I'm reasonably certain that given perfect play, its a player 2 win.
|
# ¿ Aug 6, 2015 21:12 |
|
|
# ¿ Mar 29, 2024 08:11 |
|
qbert posted:Yep, I think whoever puts down the 4th tile is guaranteed a win. I'm not 100% sure your system works, or at least, I can follow your moves and come out with a player 1 win if you aren't careful about where the tiles are placed. If player 1 leaves you with 3 in an L shape for you final turn you're screwed. There is more than one possible winning line, as there are multiple configurations for the 2x3 area in the bottom left, everything else is set in stone though. Assuming all the other neccessary pieces are in place, that 2x3 rectangle is a win for whoever's turn it is, so the game becomes "who can make sure they aren't the last to complete the neccessary pieces", which player 2 can ensure by playing the horizontal line [straight, straight, upbend] at 1,3 (0,0 being the left square of the station). I thought maybe player 1 could get a win by ignoring the neccessary pieces and trying to fudge something with that 2x3 area before the end but whatever move they make player 2 has a winning strategy.
|
# ¿ Aug 6, 2015 22:00 |
|
I've been looking at monorail some more and I'm now 100% convinced that Kyunghoon/Junseok's opening move is a loss. This is player 2's winning move, S being station, 1s being first move, 2s second move, puncuation showing the bends etc: code:
More generally, any shape on the board can be examined in isolation and determined whether the first or second player to move in them has a forced win. Whichever player splits the board up into n pairs of equivalent first player win shapes wins, second player win shapes can be safely ignored by both players.
|
# ¿ Aug 9, 2015 21:51 |
|
Your first example has an equivalent counter its just not an identical move, by spliting the right hand side into 2 ones which are 1 valued (first player win), the counter becomes creating a 0 valued (second player win) section on the left. Second example, to a degree, you're correct, I missed that move completely and it unbalances the two sides so they are not functionally equivalent. However if player 2 plays this in response: code:
Possible responses all lead to P2 win: P1: 3 tiles in the 3 section -> P2: 1 tile in the 2 section P1: 2 tiles in the 3 section -> P2: 2 tiles in the 2 section P1: 1 tile in the 3 section -> P2: 1 tile in the 1 section P1: 2 tiles in the 2 section -> P2: 2 tiles in the 3 section P1: 1 tile in the 2 section -> P2: 3 tiles in the 3 section P1: 1 tile in the 1 section -> P2: 1 tile in the 3 section edit: I think in my brevity I explained badly, mirrored moves don't have to be the same, or even leave the same result, they just have to keep the game board in the status quo of having pairs of equivalent 1 valued sections and an arbitrary number of 0 valued sections. Blue Star Error fucked around with this message at 23:48 on Aug 9, 2015 |
# ¿ Aug 9, 2015 23:39 |
|
Fast Luck posted:You can always lose these death matches. Gyul Hap is the closest after Twelve Janggi to one you can take with pure skill but Quattro? That's so much luck, and double-sided poker will have plenty of luck too. I'd rate gyul hap as 'fairer' fwiw than twelve janggi which has a second move advantage. If they fix the boards so that given perfect play each player would have a chance to call gyul 5 times it would probably be the only completely fair deathmatch.
|
# ¿ Aug 19, 2015 14:24 |
|
2nd player has an advantage in 12 janggi but there is very little chance someone has memorised the entire perfect play game tree. I don't think its so large as to be completely impossible to memorise after all people memorise large chunks of chess openings, but to do so you'd need to look through every branch of the tree and no-one's going to be doing that manually, and as far as I'm aware none of the remaining players are programmers.
|
# ¿ Sep 2, 2015 14:34 |
|
It probably wouldn't have made in difference to the end result but they really need to get rid of that stupid draw rule with their betting deathmatches
|
# ¿ Sep 2, 2015 18:33 |
|
If there's another season I'd like to see Hyunmin and Kyunghoon's guests as players, Hyunmin's guy has been in two episodes and won twice, and Kyunghoon's just seemed quite fun.
|
# ¿ Sep 2, 2015 22:06 |
|
Doctor Reynolds posted:It was like that all game, not just at the end. It was certainly edited to give the impression that it was a late game tactic but he could well have been doing it for longer. You are right though that Kyunghoon has had more than his fair share of luck in the deathmatches, but he's been exceptionally good at exploiting the advantages that chance gives him though so its not like its undeserved.
|
# ¿ Sep 2, 2015 22:38 |
|
Zythrst posted:No way was that intentional, because the one player couldn't see the shoe at all. You can't have an 8/10 because its even/odd on each side, and that bet wouldn't be a certainty because there's a 20% chance your opponent has a back 10, it would only be a certainly if your opponent bet on the 3 first, which would be a odd move, considering they'd already know that their 3 loses to your front card. Even if you have the best possible card (a 9/10 or 10/9) there's always a 20% chance your opponent can pull out a draw, which in a double sided bet is a win for your opponent. The lowest return of a won double sided bet is winning 13 chips, while a losing one loses you 3 chips, so you only need 1 in 5 to come off to be in profit. Which means if both your cards are greater than their front card, and your Betting games are not my forte so that might be slightly off but its along the right lines. Blue Star Error fucked around with this message at 16:06 on Sep 7, 2015 |
# ¿ Sep 7, 2015 16:00 |
|
That was great Dongmin is incredible loving hell, just no weaknesses at all. Generally I think I'd be pretty good at genius deathmatches but mystery sign what the gently caress. I didn't get a single one.
|
# ¿ Sep 15, 2015 13:40 |
|
MisterZimbu posted:I answered exactly one of them. lol I did exactly the same Power of Pecota posted:I mean the alternative would be what, bloating up the upload with the end credits looped 20 times? They could film and show all three games, and say the winner only gets 100% of the prize if they win all 3 matches, otherwise the loser gets 10% or something. At least then we'd get to see all 3 games they've prepared
|
# ¿ Sep 16, 2015 02:01 |
|
If anyone wants to try out 12 janggi I knocked up an AI you can play against here. You can play as 2nd player by changing the player=0 in the url to player=1 and similarly raise or lower the difficulty. For various reasons the AI isn't that great but its an alright challenge at difficulty 7 or 8, past that I think it actually gets worse because I set a 30 second limit for it to generate moves so you aren't waiting minutes for it to evaluate a billion moves.
|
# ¿ Sep 28, 2015 01:52 |
|
Kangra posted:Very cool. Elite posted:The AI seems to randomly give up a lot of the time for me (I've had more failed games than completed games now). Usually it comes up with a move within about 5 seconds, but sometimes it randomly does nothing even when leaving it for minutes (so significantly longer than the full 30 seconds calculation time). This even happens for the opening moves of the game, where it can't possibly be in a lost position. Not sure what's causing that. Thanks for the feedback. Looks like I didn't actually upload the version with the 30 second limit oops. I think there's something weird going on with my hosting, its lovely shared hosting so its possible I'm coming up against some CPU limit. When I run it on my PC it never fails to generate a move, even if it knows its going to lose, but you're both right it does appear to be locking up on the live server. I think I'll have to rewrite the AI to make it more streamlined.
|
# ¿ Sep 28, 2015 12:14 |
|
The Genius done on ITV will be loving awful. It needed to be picked up by BBC4 or 2 to work.
|
# ¿ Oct 6, 2015 18:07 |
|
NowonSA posted:Yeah, I feel the same way as pokeyman. Need to check out the Tittycast later. ITV's target viewers are, to put it bluntly, morons. They are by far the worst British channel the rights could have gone to. They will either do a faithful adaption which will utterly bomb or more likely, dumb it down and make it poo poo. I can't think of a single ITV celebrity I'd trust to draw at noughts and crosses let alone pull off a Dongmin. I don't reckon you'd ever get to see a full game of chess as a deathmatch, it just takes too long and isn't very interesting television, there's a reason they used a miniature version of shogi rather than a larger version.
|
# ¿ Oct 7, 2015 06:36 |
|
Weird choices of games to try to make versions of. Monorail is already an existing game that presumably has some sort of licence over the concept and I'm not convinced Black & White can be won at a rate greater than chance.
|
# ¿ Nov 4, 2015 13:55 |
|
Shakugan posted:I'm really curious to know if they'll adapt the half celebrity half random people strategy that the Genius Korea always used. Because the UK is one of the only countries I can think of that does have a sort of similar celebrity culture. Stephen Fry, David Mitchell, Bill Bailey (basically, a bunch of the people who've appeared on QI) could all be good contestants. I think there is little chance they have half celebrities half normals, it'll be one or the other. They won't want to end up with "Hi I'm Victoria Coren off the telly and I'm competing for Great Ormand Street Hospital" alongside "Hi I'm Ian off of the public and I'm going to spend my winnings on holidays and some cocaine". All the people you mention are smart because they have a lot of knowledge, I don't think any of them would be very good at it, similarly I don't think Jinho would be very good on QI. There's plenty of UK celebrities that would be good but most are BBC, Victoria Coren is a pro poker player, Dara O' Briain has a physics degree, one of Armstrong and Miller I forget which has a physics phd, Rachel Riley is like a sexy rainman, Brian Cox is an irl particle physicist, Marcus du Sautoy is an irl maths professor, Tim Key loves cheating and so on.
|
# ¿ Nov 16, 2015 15:14 |
|
I love the bald manc twat but he'd be loving useless at this even if the main game was guessing the answers to his own rockbusters.
|
# ¿ Nov 18, 2015 22:49 |
|
The Saddest Rhino posted:Genius facebook appears to suggest a surprise is in store the start of building a new set?
|
# ¿ Nov 23, 2015 00:39 |
|
Ditocoaf posted:Isn't the core talent behind designing and producing The Genius leaving? If I remember that correctly, then ending the show is the proper thing to do. I prefer when a great show doesn't outlive its greatness. Not every good thing in this world must be driven into the ground. The staff don't really matter imo as long as you have interesting contestants and good games. If anything having some other people design the games could be a good thing as the last two seasons the games got very samey.
|
# ¿ Dec 12, 2015 15:51 |
|
Can Of Worms posted:You forgot the part where Jinho loses becauses they go to a tiebreaker when Jiwon goes all in, making his loss entirely luck based. Corn Thongs posted:The worst part of the Jiwon/Jinho match is when a hand ends in a tie, the bets automatically carry over into the next hand. Jinho called on the first tie, but that rule doomed him and is the stupidest rule they had in the game To be fair, its no more luck based than the game itself. You just have to take it into account with the odds. The hand that Jiwon went all in on had both players with a 2, and 1 8 on the board. It was the first hand so of the 37 unknown cards each player could have, 34 are winners, and 3 would draw, so thats a ~92% of winning, and an ~8% chance of drawing which translates to another ~4% chance of winning the following round, so each player knew they had a ~96% chance of winning and the only way they could lose is a draw, which isn't really any different from having an actual 96% chance of winning a single hand with no draws. The draw rule effectively makes it so there are fewer totally safe bets. It does give the appearance of a loss based on a coin flip which is lovely but in reality its just a continuation of the previous hand's game choices.
|
# ¿ Aug 19, 2016 14:37 |
|
Max posted:This doesn't take into account a player actively keeping track of what has been played, which Jinho is known to do and what got him the win against Gura in the first season. There were no cards already played, it was the first hand.
|
# ¿ Aug 19, 2016 15:22 |
|
Max posted:Honestly, Jinho shouldn't have taken that bet right off the bat and seen the strategy for what it was. The problem then is, if you won't call an all in when you have a 96% chance of winning, when will you? 100%? Thats unlikely to happen until the last few hands of the deck and by then Jiwon could have bled you dry by going all in every hand. If Jiwon thought his strategy through which he probably hadn't, he also could potentially fold the final 2 or 3 hands of the deck to lose the minimum amount to Jinho's card counting skills. Jiwon got very lucky that essentially the perfect hand for his strategy came out first time but Jinho made the right decision to call imo.
|
# ¿ Aug 19, 2016 15:47 |
|
Exercu posted:Funny thing is, when Hyunmin clowns on Jinho in the BTS - when Hyunmin says he's won and Jinho keeps playing anyway, Jinho could have won from that position. The winning move even uses the shape Jinho was talking about at the beginning of the BTS. Its weird how bad all the players are at monorail even though its the only deathmatch which is feasibly solvable.
|
# ¿ Aug 27, 2016 05:19 |
|
Exercu posted:Yeah, I'd put the death matches in four rough categories. I'd be interested to see your Monorail solution if that implies you have one. I found a player 1 solution, technically 4 solutions if you count rotations of it and I'm pretty sure there are more. 12 Janggi as another perfect information game is theoretically solvable certainly but there are about a billion game states so no-one is going to memorise the entire game tree
|
# ¿ Aug 27, 2016 15:16 |
|
First move:code:
code:
I think if you are the first player and you don't have a solution the best strategy is to play one tile, as the more fixed the track is the easier it is to come up with a winning move.
|
# ¿ Aug 27, 2016 17:29 |
|
Propaganda Machine posted:I haven't seen The Genius in a while, but I backed the Kickstarter to get the Black and White and Monorail card games and I might actually get to use them this weekend. Kyunghoon's monorail strategy is a losing one so you don't have to worry about that. Check my post history in this thread if you want the move to beat his move and also an actual winning player 1 strategy. I'm interested to know what the alternate cards are though? I guess thats how they got around the fact that monorail is already an existing board game who's name I've forgotten. I think the best way to try to "unbreak" monorail is to start with one tile randomly placed so its less likely a player can use a planned winning player 1 strategy, instead having to solve it on the fly. Black & White I remain unconvinced that its possible to win at a greater rate than chance without meta-game strategies.
|
# ¿ Sep 22, 2016 20:39 |
|
tbp posted:Finally finished up all four seasons of The Genius. One of the best TV shows I've seen in a long time, and tough to say whether itd be replicable in the west. Sangmin probably my favorite character of all four seasons, his comedic moments were amazing. I'm glad you enjoyed it fellow footy poster.
|
# ¿ Jun 22, 2017 02:16 |
|
|
# ¿ Mar 29, 2024 08:11 |
|
Artelier posted:Is that the one where Dongmin knew it couldn't be solved, played a move anyway which psyched out his opponent so hard, even though she thought it couldn't be solved, she played on anyway? And when it went back to Dongmin he immediately went "Can't be solved." Kinda, Dongmin massively hosed up, it wasn't intentional, he played a dumb move that left it unsolvable. After that his only choice was to act as if it was still solvable even though it clearly wasn't and hope that his opponent wilted, which she did. Propaganda Machine posted:I thought somebody broke Monorail in season 4. Wasn't it titty god? Nah his opening move is a loss given perfect play. There is a opening move win though. NowonSA posted:lame poker poo poo taking advantage of their dumb "when it's a draw keep flipping cards and high card wins" rule to eliminate one of my faves... not so much. The more I think about that the more I think it was actually a really smart play. In gambling games the superior player's advantage becomes larger with a bigger sample size of hands/bets, if you know you're inferior and are more likely to lose the longer the game goes on, trying to make it as luck based as possible by going all in whenever your chance of winning the hand is >50% is a good strategy.
|
# ¿ May 3, 2018 20:34 |