A software I would love to have - Page 4

Sushiglutton · #31 09-25-2007, 05:01 PM

[ QUOTE ]
yeah wtf. i loled when someone said $40 = reasonable.

[/ QUOTE ]

Hey, I'm poor eating noodles all day long.

Sushiglutton · #32 09-25-2007, 05:16 PM

[ QUOTE ]
line analysis is the future of online poker.

and whoever said there are no nl bots, way off by several years.

[/ QUOTE ]

Do u mean that there are allready NL-bots that beat top players? Last thing I heard on the bot thing was that Phil Laak beated some limit bot.

euroglot · #33 09-28-2007, 01:17 AM

Very interesting thread

Matt Flynn · #34 09-29-2007, 12:35 PM

[ QUOTE ]
[ QUOTE ]
line analysis is the future of online poker.

and whoever said there are no nl bots, way off by several years.

[/ QUOTE ]

Do u mean that there are allready NL-bots that beat top players?

[/ QUOTE ]

i don't know but strongly suspect not.

not what i was saying.

Sushiglutton · #35 09-30-2007, 07:28 AM

[ QUOTE ]
i don't know but strongly suspect not.

not what i was saying.

[/ QUOTE ]

Since you are an oracle I guess you don't need to be more precise than that. But a bot is really only interesting if it can beat humans. It's not hard to build a 'randomly clicking the buttons'-tool.
But I guess you mean it can beat medium skilled players?

Sushiglutton · #36 09-30-2007, 10:25 AM

My free river tool is now available for downloading. Enjoy!

You need one excel file

and one word file

It's very simple and primitive. I hope it may inspire some computer wiz to make a better version. You are allowed to make any changes, updates you want. Comments, bug reports etc please pm me.

Matt Flynn · #37 09-30-2007, 04:23 PM

[ QUOTE ]
[ QUOTE ]
i don't know but strongly suspect not.

not what i was saying.

[/ QUOTE ]

Since you are an oracle I guess you don't need to be more precise than that. But a bot is really only interesting if it can beat humans. It's not hard to build a 'randomly clicking the buttons'-tool.
But I guess you mean it can beat medium skilled players?

[/ QUOTE ]

not trying to be oracle-ish, just not dumb enough to give specifics.

the first time i became reliably aware of a bot beating humans in nl was about four years ago. since then, i have become aware of much better bots. in general, i don't know much about bots. but i know for sure they exist and that some do well. no big deal, lot of people know the same.

Sushiglutton · #38 10-01-2007, 11:45 AM

The other link seems to be some casino/xxx kind of trash. My apologise.

Try this instead:

http://pokerz.attractedtopsychos.net

emerson · #39 10-02-2007, 04:37 PM

Here is what would work. Rather than program the computer to play certain lines, create additional queries that Poker Tracker could do to analyze the actual return on the lines that you filter.

That's step 1.

Step 2 should have already been set up by someone. A large combined 2+2 database with millions of hands from Poker Tracker.

With sufficient filters you could get more real information, and better, than you could get from computer sims.

Example: you could first filter for players who had a win rate of at least 1BB/100 with a sample size of at least 50,000 hands. Then you could filter for hands when they were the pre-flop raiser, in position, against one opponent. Then you could filter for various defined lines.

An alternative to this big 2+2 database would be to have Poker Tracker changes so that it could track hands from Turbo Texas Holdem. You then run sims with millions of hands. Then, again, you run queries.

Absolution · #40 10-03-2007, 02:09 PM

[ QUOTE ]
Here is what would work. Rather than program the computer to play certain lines, create additional queries that Poker Tracker could do to analyze the actual return on the lines that you filter.

That's step 1.

Step 2 should have already been set up by someone. A large combined 2+2 database with millions of hands from Poker Tracker.

With sufficient filters you could get more real information, and better, than you could get from computer sims.

Example: you could first filter for players who had a win rate of at least 1BB/100 with a sample size of at least 50,000 hands. Then you could filter for hands when they were the pre-flop raiser, in position, against one opponent. Then you could filter for various defined lines.

An alternative to this big 2+2 database would be to have Poker Tracker changes so that it could track hands from Turbo Texas Holdem. You then run sims with millions of hands. Then, again, you run queries.

[/ QUOTE ]

This is case-based reasoning and is probably the proper way to handle this problem.