Open Side Menu Go to the Top
Register
WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot
Two Plus Two Forums
WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot

05-06-2015 , 03:36 PM
Quote:
Originally Posted by AdamSchwartz WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot
Update on the totals please!
Totals are always available on the claudico extra stream
http://www.twitch.tv/claudico_extra
05-06-2015 , 03:44 PM
Quote:
Doug really seems tired and a bit in idc mode and just wants it to be over. Folding 3k hands should be -225k fwiw . He had periods were he was just rushing through hands on the first days, too, but I feel like his attitude has really changed (starting yesterday).
No surprise it's incredibly high volume for intellectual sports. Take chess for example: they play 4-5 hours a day and usually have a rest day every 4-5 days.
The conditions they are playing in are way better as well, if anything starring at computer screen is more tiring than looking at the chessboard and being able to walk around/eat in between moves.
I think stamina is huge issue and while they are dealing with it well it still influences their performance.

Last edited by punter11235; 05-06-2015 at 03:54 PM.
05-06-2015 , 03:45 PM
Quote:
Originally Posted by JPR WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot
At this point, I think there is a much bigger than slight chance that Claudico gets a statistical tie vs Doug and Jason.
That would only be 40k hands though. Considering the 200bb stacks and the fact claudico likes to 5x shove rivers every 50 hands I bet the stdev in these games is really high. You really have to view the humans as a team here. Picking on jason because hes down a few buyins after 15k hands or whatever is dumb. Over the last two weeks I've been checking the streams and I swear everytime I put on Claudico_vs_JasonLes he's on the river with a weak two pair facing some 4x river shove with beads of sweat rolling down his face. lol.
05-06-2015 , 03:56 PM
Quote:
Originally Posted by Loctus WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot
Totals are always available on the claudico extra stream
http://www.twitch.tv/claudico_extra
First session today: Bjorn +68k, Doug -82k, Jason +48k. Dong's stream was either black or off (but if I heard it correctly he was close to break even).

So humans are up +708k + Dong's 1st session result. I bet the scoreboard on claudico_extra will be updated soon.
05-06-2015 , 04:51 PM
Everything is working according to plan. WCGrider mentioned that he like the team to build a decent lead at the beginning because he anticipate the bot to rally at later stage of the challenge.
05-06-2015 , 04:56 PM
Very much impressed with Bjorn's play the last two days juxtaposed against Doug's seeming indifference/fatigue. He looks like he cares about giving it his all.
05-06-2015 , 04:57 PM
Quote:
Originally Posted by e i pi WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot
That would only be 40k hands though. Considering the 200bb stacks and the fact claudico likes to 5x shove rivers every 50 hands I bet the stdev in these games is really high.
Hence the statistical tie
05-06-2015 , 05:05 PM
Is this really Doug? Unenthusiastic, sleepy, and wtf is with the hair bruh?

Is Dong 5x the player Doug is? The results say, yes.

edit: Kind of funny how everybody said Doug put on a good front for poker and now he's doing the complete opposite. I get why he's doing it now but it's kind of ironic to turn into the stereotypical miserable grinder.

Last edited by .isolated; 05-06-2015 at 05:14 PM.
05-06-2015 , 05:57 PM
terrible post
05-06-2015 , 06:16 PM
Are these guys getting paid to participate in this?
05-06-2015 , 06:19 PM
Quote:
Originally Posted by DarkMattersMan WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot
Are these guys getting paid to participate in this?
they'll split 100k
05-06-2015 , 07:40 PM
Quote:
Originally Posted by The4thFilm WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot
they'll split 100k
Don't forget Doug's Gunnar Optiks' sponsorship money
05-06-2015 , 09:12 PM
Quote:
Originally Posted by potpotpondo WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot
Jason Les should be expelled from humanity.


Quote:
Originally Posted by .isolated WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot
Is this really Doug? Unenthusiastic, sleepy, and wtf is with the hair bruh?

Is Dong 5x the player Doug is? The results say, yes.

edit: Kind of funny how everybody said Doug put on a good front for poker and now he's doing the complete opposite. I get why he's doing it now but it's kind of ironic to turn into the stereotypical miserable grinder.
i assume you mean Bjorn and not Doug.

and the amount they're beating Claudico for over the course of 17.5k hands each is hardly a relevant indicator of who the better player is between the 2 of them.

Bjorn is a beast though.
05-07-2015 , 01:29 AM
I feel strange about this thing. On one hand it's clear that this thing has the potential to kill poker, to essentially solve the game and to remove any challenge or mystery. On the other hand, I feel that if that's to be so - let it be. Even though it would effect me as a pro and I might have to find something else to do. If the game can be reduced to being solved by a machine, well then maybe it's not that great of a game after all and it deserves that. Does anyone else feel like this?
05-07-2015 , 01:42 AM
chess bots are better than the bests, isnt it still a great game?

was limit holdem a bad game once bots solved it?
05-07-2015 , 02:18 AM
Quote:
Originally Posted by Kirbynator WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot
chess bots are better than the bests, isnt it still a great game?

was limit holdem a bad game once bots solved it?
I do think solved games are generally less enjoyable and yes LHE has seen increasingly less action as it has moved towards being solved.

Thinking about spots and discussing them is a large part of what makes these games enjoyable. It is particularly fun when different styles clash and we can debate whether a higher 3bet is more profitable or not for example.

It is far less enjoyable when you will just be able to say to a friend I am unsure if I played this hand correctly and he will say well PokerBot says correct play was bet $50 there, end of discussion. Personally I do take less enjoyment from Chess/backgammon knowing that I can get the optimal play on my mobile phone in an instant and that if you want to become world class you may be best served by memorising lines as opposed to critically thinking about them.
05-07-2015 , 02:20 AM
the correct answer was limit holdem has always been a bad game
05-07-2015 , 02:28 AM
Quote:
Honestly, we had just assumed computing exploitability is hard
So you guys are using an algorithm to approximate something you can't measure?

Quote:
, since we use imperfect-recall abstractions
Calculate ranges at the end of every line (up to turn), you can do that cause you play flops/turns fast.
Solve rivers with those ranges (or recreate them in any other way are doing during play but just solving them with assumed ranges is the most obvious way), profit. You can use even 5 bet sizes on the river, it will be fast anyway, especially on the kind of hardware you have.
As it's depth-first-search you don't need much memory for that either.

Quote:
It's not clear to me if it's doable, and what the scientific value is of computing a bound on our exploitability like this.
Is there any scientific value in approximating something you don't have a measure for? I mean the results can be w/e and you will not know. If you can't assess it you can't experiment, you can't say if a tweak is good or not etc.

Quote:
chess bots are better than the bests, isnt it still a great game?
Chess is way harder than poker though and top programs are nowhere near optimal play yet (although way stronger than humans). The problem with Holdem is that it's possible to solve it or at least approximate it to make it essentially solved (like Alberta team did with limit). It will happen to HU NLHE sooner than solution to chess happens.
Good thing though is that multiplayer games are unlikely to be solved and equilibrium in those isn't as valuable anyway.

Last edited by punter11235; 05-07-2015 at 02:35 AM.
05-07-2015 , 02:42 AM
I think I understand why HU can be solved, but I've never seen an explanation why multiplayer games can't be. Can you describe it simply?
05-07-2015 , 02:57 AM
Quote:
Originally Posted by punter11235 WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot
So you guys are using an algorithm to approximate something you can't measure?


Calculate ranges at the end of every line (up to turn), you can do that cause you play flops/turns fast.
Solve rivers with those ranges (or recreate them in any other way are doing during play but just solving them with assumed ranges is the most obvious way), profit. You can use even 5 bet sizes on the river, it will be fast anyway, especially on the kind of hardware you have.
As it's depth-first-search you don't need much memory for that either.
I would assume that this is exactly how the bot currently works, so estimating distance from NE the same way would yield a useless exploitability lower bound of 0.

ignatius
05-07-2015 , 04:21 AM
Quote:
Originally Posted by NoMeansYes_ WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot
I think I understand why HU can be solved, but I've never seen an explanation why multiplayer games can't be. Can you describe it simply?
Besides the fact that the game tree is exponentially larger, players can intentionally or unintentionally cooperate, which causes all sorts of problems.
05-07-2015 , 04:29 AM
Quote:
Originally Posted by psloth WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot

Nice interviews with Dong and Jason, Sam. Much more enjoyable than muted Doug/Bjorn streams.
Anyone know on which stream and approximate times this occurs at?

Thanks!
05-07-2015 , 04:45 AM
Quote:
Originally Posted by punter11235 WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot
...
Chess is way harder than poker though and top programs are nowhere near optimal play yet (although way stronger than humans). The problem with Holdem is that it's possible to solve it or at least approximate it to make it essentially solved (like Alberta team did with limit). It will happen to HU NLHE sooner than solution to chess happens.
Good thing though is that multiplayer games are unlikely to be solved and equilibrium in those isn't as valuable anyway.
Let's not turn this into a chess thread, but surely Komodo and Stockfish can't get like 10% better, but rather 0.5%. They are close to optimal.
05-07-2015 , 04:53 AM
Quote:
Originally Posted by NoMeansYes_ WCGRider, Dong Kim, Jason Les and Bjorn Li to play against a new HU bot
I think I understand why HU can be solved, but I've never seen an explanation why multiplayer games can't be. Can you describe it simply?
Simply, HU only deals with two positional situations and never more than two players competing in a pot. This doesn't mean HU can be easily solved or that anyone has even come close.
05-07-2015 , 05:00 AM
Quote:
I would assume that this is exactly how the bot currently works, so estimating distance from NE the same way would yield a useless exploitability lower bound of 0.
No, because you calculate ev of current strategy of the bot (for both sides) and max exploit vs those. Then you subtract.

Quote:
Let's not turn this into a chess thread, but surely Komodo and Stockfish can't get like 10% better, but rather 0.5%. They are close to optimal.
Current consensus among chess programmers is that there is at least 200-300ELO points to go. This is way more than "10%" better w/e way you define it.

      
m