Rush Poker: An Ne Dream

[simulatedworld]

Anyone who plays or follows online poker should by now have heard of Full Tilt's new innovation, Rush Poker, which allows faster play than ever by moving players to a new table with new opponents immediately after they finish acting on any hand. Even if it's not your turn yet, if you already know you don't want to play your hand, you just hit the "Quick Fold" button and you're immediately moved to a new table with a new hand and new opponents.

This effectively triples the number of hands per hour that you can play, which could mean big things for the win rates of winning players.

As for strategy adjustments, the most obvious implication is that you can't collect data on people's play tendencies and use it against them because every hand is the first (and last) hand of a new table with new opponents.

I want to talk about why this is extraordinary for Ne players like myself with extremely short attention spans.

Suddenly the game is no longer about collecting detailed statistics and developing highly specific reads on particular players--suddenly the game is about looking at the entire player pool as a group and making plays that would work best against whatever the average player at this moment is doing. In Rush Poker, table image effectively no longer exists.

This is a HUGE difference and I'm already doing dramatically better than I ever did in traditional games. I've always enjoyed poker but I've never been more than a marginal winner online (except for back in the days of Party Poker when there were hundreds of terrible players running around everywhere handing out free money), largely because I get bored and let Ne start making "out of the box" plays that end up biting me in the ass--namely, getting bored from waiting too long for a decent hand and making overly ambitious bluffs that screw me out of long term profit.

I don't like watching individual players for hours on end to pick out their particular play idiosyncracies--I like playing a huge number of hands and making a lot of very fast decisions, and always having something to do.

The traditional way of dealing with the boredom is to play multiple tables simultaneously--but the problem with this is that it divides your attention and ruins your ability to watch individual players for notable play tendencies. People who are only playing one or two tables will beat you because they're paying more attention to your play patterns than you are to theirs.

Not so with Rush Poker! Suddenly I can conceptualize the game in terms of an entire group of opponents and respond according to what would work best against the average opponent, rather than having to worry about detail-oriented statistics regarding the play tendencies of particular opponents.

In short, this format effectively eliminates the Si-oriented problems in my game and accentuates Ne's strengths--being able to multitask and make a lot of quick decisions in a very short time based on generalized information about a group as a whole instead of specific, highly detailed information about individual players, and adapting on the fly to changing trends in the overall player pool based on patterns in this highly generalized information.

Can it get any more Ne?

It's incredible. Has anyone else tried this yet? Do you like it better or worse than traditional online poker, and do you think this has any correlation with your psychological type?

 

[foolish heart]

Can it get any more Ne?

It's incredible. Has anyone else tried this yet? Do you like it better or worse than traditional online poker, and do you think this has any correlation with your psychological type?

I just gave it a try this morning. Wow, and when I didn't think I could make money any faster in this game! Awesome. I will try to record how my profit rate improved but it was pretty drastic. They should have called it "Quickfold" poker.

 

[nozflubber]

well, if there is a typological self-selection mechanism going on here, you DO realize the real "Ne-cokeheads" will be highly attracted to this game over the traditional style, ?? assuming your own fascination is representative of course

IOW, if this is an ENTPs wet dream, you might have better results at the traditional tables due to the population differences.

 

[simulatedworld]

well, if there is a typological self-selection mechanism going on here, you DO realize the real "Ne-cokeheads" will be highly attracted to this game over the traditional style, ?? assuming your own fascination is representative of course

IOW, if this is an ENTPs wet dream, you might have better results at the traditional tables due to the population differences.

What do you mean about population differences?

 

[Amargith]

He means that it will attract others of your talent in the same way as it appeals to them, making the game more competitive, meaning more than Ne-power will be needed.

 

[nozflubber]

yeah, thanks amarg basically, I'm saying many others just like you will be thinking the same. So others who prefer to think/react fast will be your opponents, rather than the standard population demographics of the traditional tables.

I'm not saying it's NOT a better choice for you, I'm just saying you'll probably be facing more ENTPs

 

[simulatedworld]

True, but other ENTPs are n00bs, so I shall pwn them mightily.

 

[Nocapszy]

here's my right-brain contribution.

[youtube=http://www.youtube.com/watch?v=BbMp-TYTyWs&feature=related]rush - limelight[/youtube]

 

[Nocapszy]

as for my left brain contribution:

this format is highly accessible to bots. it's only a matter of time before flesh and blood players won't be able to win.
it'll take a while for them to be perfected, so get your face-time in now.

 

[redacted]

as for my left brain contribution:

this format is highly accessible to bots. it's only a matter of time before flesh and blood players won't be able to win.
it'll take a while for them to be perfected, so get your face-time in now.

Why?

It's possible you could make a better-than-the-average-player bot, but the good players will exploit the fact that some proportion of players are bots and win more often.

 

[simulatedworld]

Why?

It's possible you could make a better-than-the-average-player bot, but the good players will exploit the fact that some proportion of players are bots and win more often.

Not to mention bots won't be able to adapt to changing trends in play styles without being reprogrammed routinely.

There's no one perfect strategy for this game; the best strategy just depends on what the average opponent is doing, which is already changing quickly...for instance, a lot of players are 4-betting after opening the button because 3-betting open button raises from the blinds is becoming standard.

Nocap, you're oversimplifying the format.

 

[teslashock]

evolutionary algorithms.

 

[simulatedworld]

Get on that!

 

[teslashock]

Someone will. You're oversimplifying AI if you think bots will happen and that shit won't, son!

 

[Nocapszy]

Why?

It's possible you could make a better-than-the-average-player bot, but the good players will exploit the fact that some proportion of players are bots and win more often.

you can only exploit the bot if you know it's a bot.
with rush poker, it's completely anonymous.

Not to mention bots won't be able to adapt to changing trends in play styles without being reprogrammed routinely.

There's no one perfect strategy for this game; the best strategy just depends on what the average opponent is doing

and you say that a human can do that better than a computer?

Nocap, you're oversimplifying the format.

i'm not, but i realize that i've burst your bubble so i don't blame you for retaliating this way.


look at it this way: because it's an enclosed group of players with a retroactively measured state of play, the bot can make EV decisions based on those statistics.
sure, your'e going to lose sometimes, but as long as your bot plays within a safe EV and you play a long enough session, you will come out ahead, unless the deck is running exceptionally hot for everyone except your bot, in which case your bot will lose a buy in or two before going on tilt and short-circuiting.

this game is even more mathematically beatable than an ordinary ring game. since rush poker calculates the statistics of everyone and not just one players at your table, you have an enormous sample size, and so the math is more reliable. you wouldn't even have to program perfect or even decent theory.

you can figure about how tight/loose the play pool is by the avg. plr/flp.
it's obviously counting hands that never go to a flop, because the minimum percentage of players at a table to see a flop can be is 22%
[2/9 = 0.22]
if it's less than 22%, then naturally, you're going to have a pretty tight game. people aren't defending their blinds etc.
i'm looking right now and i see two pools with 19% plr/flp. that means most hands never even see a flop, and 3+ way flops are a rarity.
i'm also looking at a pool with a 30% plr/flp. that means more often than not, the hand not only sees a flop, but sees more than just a limper and the big blind. that's going to be a really loose game.
i don't know how you'd figure out the average hand range based on that, but someone can, and they will.

if you wanna get really fancy, you could look at the average pot size, and figure how many of the pots are being raised/limped.
though that has no relevance to the pool's hand selection and can't help it to calculate its card's equity vs. the average hand in the pool, it can help to determine which pots it wants to enter.

and the thing sw said about 3betting - well you can just program the bot not to get involved in marginal situations with potentially high cost. after all, a few unlucky reraised pots can be devastating to stack size and ruin hours of work.
what i mean is, as long as your bot is playing with such general information, small ball will be key. a slow trickle - like the casinos love.

not to mention, your HUD is still going to work. sure, it will be useless for a while, but after a enough long sessions, you're going to be able to identify the regulars and pin their play.

Not to mention bots won't be able to adapt to changing trends in play styles without being reprogrammed routinely.

the only way i know how to respond to this is "you'd be surprised"

the bot has it both ways in this game, as long as you make sure to play long sessions to make up for the losses inevitable to something using such broad calculation.

the biggest thing it boils down to is this: it will be hard, but bigger feats have been accomplished. and as long as people can exploit mathematics they will. it's not just that it works - it's fun.

 

[Munchies]

that sounds really homosexual. i like to play poker and figure people out, then play mindgames. this makes that impossible.

 

[simulatedworld]

Nocap:

Are the statistics given by Full Tilt (% of players per flop, average pot size, etc.) enough to derive a consistently winning strategy? If not, the bot would have no source of statistics to adapt to changes in.

I guess more than anything this depends on quality of opponents, but I imagine FT has methods of looking for bots. You can't just run the bot from the same IP address 24/7, for instance, or they'll pick up on the fact that you're not really human.

I'm sure dedicated bot programmers will find ways to circumvent that, but even if bots do win, they won't prevent the better players in the pool from also winning, so I'm not too worried about it.

 

[Nocapszy]

Nocap:

Are the statistics given by Full Tilt (% of players per flop, average pot size, etc.) enough to derive a consistently winning strategy? If not, the bot would have no source of statistics to adapt to changes in.

yes. they are. i know enough about inference on mathematics and averages to say conclusively that a consistent winning strategy can be derived from that single ratio.

I guess more than anything this depends on quality of opponents, but I imagine FT has methods of looking for bots. You can't just run the bot from the same IP address 24/7, for instance, or they'll pick up on the fact that you're not really human.

uh... no shit. for arguments sake, does it really escape you that programmers can think of that themselves and remedy it with only a few lines of code?

I'm sure dedicated bot programmers will find ways to circumvent that, but even if bots do win, they won't prevent the better players in the pool from also winning, so I'm not too worried about it.

naturally. but they will undermine the winnings of the players.

 

[simulatedworld]

naturally. but they will undermine the winnings of the players.

Sure, but no moreso than other winning players.

Also, wouldn't bot programmers use their winning bots in the highest stakes available (which for Rush is currently 1/2)? You should be able to avoid the bots in this case by simply playing levels lower than the highest ones available.

 

[Nocapszy]

Sure, but no moreso than other winning players.

this is getting really speculative.

Also, wouldn't bot programmers use their winning bots in the highest stakes available (which for Rush is currently 1/2)? You should be able to avoid the bots in this case by simply playing levels lower than the highest ones available.

answering this requires more knowledge of poker and programming than i have.

again, getting speculative.