Developing Software like PT - Any suggestions?

[PSUinDC]

Hi all... while I sit at my mundane job, awaiting a government security clearance, I found myself wanting to work on my programming skills. I started to read about Poker Tracker and was very interested in making something like this.

After a few weeks of writing and testing code, I have a preliminary version of such a product that works for UB, and soon Party Poker. I really wanted to know what kind of statistics everyone is looking for in a software package.

I am pretty much using all of the statistics (for individuals, and by starting hand) that PT uses, and am starting features for finding stats on flopping OES-draws and Flush draws, and other fancy stuff.

Also, the database that stores the information is a central mySQL server, which makes this quite a powerful tool if multiple people are inputting hands into it.

What would you all like to see in a poker hand history stats software?

[Yads]

I'd like to know how often a player sees the flop, not just VPIP.

I also have another comment for you. Does MySQL now support foreign keys and sub searches, because at one point I know they didn't. Would that not make development a bit of a hassle?

[capybara]

This isn't really about statistics, but if you make something that works for Omaha/8 on Party you'll have an instantly viable product, even if only rudimentary features are available.

The MySQL-backed approach seems desirable, especially if you can thereby avoid the 100,000 hand limit of PokerTracker.

[Rudbaeck]

That it stores stats for non-HE games. Stud, stud/8, omaha, omaha/8 would be awesome. I don't need all the depth of PT, just the basic stats.

[AAmaz0n]

I haven't seen anything like PT in a Palm or other handheld format. I'd love to have something that I could create hand histories and stats with when I play live. Even if I could just record the hands when I'm playing and then download them into a desktop application for further review would be great.

Shauna

[bigmac366]

something that kept track of basic stats for stud or omaha would be great. even if it only kept track of wins/losses, and rake paid it would still be great. if you can accomplish something like that you could make a pretty penny...

[bisonbison]

PT fills a big niche and it fills it very well. I don't see the point in replicating the huge amount of work that Pat has done.

But the suggestion of supporting Stud or Omaha is a great one. You will get a huge response if you supported these, and it would definitely make the games more popular here.

[PSUinDC]

I'm really not out to be competition... my original intentions were to just work on my own programming. I haven't played a lot of stud and Omaha... I would have to check on the rules for those games to get the right statistics, but I wouldnt see how it'd be hard to set up.

[PSUinDC]

I don't know about foreign keys, but sub searches are available. For what I've done so far, a lot of the super complex SQL statements aren't totally necessary... I've been doing a lot of the statistics from PHP code, using loops and arrays... all of the stats will be dynamically created.

[bisonbison]

Well, I'm not saying that you're out for competition, but a niche is a niche, and it would take an incredible amount of work to offer a product as useful for Hold'em as PT is. And it would take most people several months to equal PT + a basic knowledge of excel.

However, if you were to work on games that PT doesn't support, you'd get immediate positive feedback, possible monetary support and a real novel product.

I recommend not replicating PT for the same reason I recommend not replicating MS Word. Why bother? There are clearly needs that people will hail you for.

[rmr1976]

I definitely would buy a product that recorded stud and omaha games, especially if it was able also hold the records for split pot games.

[PSUinDC]

I'll look into implementing Omaha.. I dont think it would be incredibly difficult, knowing what I have in place already. I would just have to add 2 more starting card columns in my database, I'm already tracking the game type, and I'd have to look for split pots. Bingo.

[Twitch1977]

As far as I'm concerned the most attractive part of this is that it will run off a mysql db instead of an access one. That alone makes this sound like it'd be worth trying.

What I've always wanted to see is some sort of meter rating your 'luck'. Like for a session if it could show how your distribution of starting hands compared to other players distrubution of starting hands. And also a 'luck meter' based on how often you were the favorite on the flop and still ended up losing. (This could be weighted depending on how much of a favorite you were.)

I think what PT lacks is a way to discern if a losing session was caused by bad play or just an unlucky run of cards.

IMO
T

[excession]

er, it's called the 'replay hand' button

/images/graemlins/wink.gif

seriuosly my latest bet-the-pot.com article goes into detail about spotting leaks in sessions you have played - but when you are talking about just one session that is so few sample hands that statistical analyis of 'bad beats or not' is easier to do just by reviewing your top 5 losing and top 5 winning hands with the replayer

[Twitch1977]

Can you link me to this article? Looked around a bit but I'm not sure I found the right one.

And it doesn't necissarily have to be just over one session, it'd be cool to select like a batch of sessions and run it over that or something. I dunno, wtf anymore /images/graemlins/wink.gif

[KaBoom]

In addition to PT there is also Pokeroffice already out there. Not really a need for a third similar program. But like others said, a program for stud and omaha would be a killer.

[Beach-Whale]

[ QUOTE ]
PT fills a big niche and it fills it very well. I don't see the point in replicating the huge amount of work that Pat has done.

[/ QUOTE ]

I see the point clearly. The point is that the PokerTracker software sucks bigtime in software quality. It's slow, ugly, and seems just not very well programmed. Very amatureish for a software product that you have to pay for.

[excession]

http://www.bet-the-pot.com/poker-tracker-part-5ia-page67.html

[bisonbison]

I see the point clearly. The point is that the PokerTracker software sucks bigtime in software quality. It's slow, ugly, and seems just not very well programmed. Very amatureish for a software product that you have to pay for.

Alright champ. Pick this fight. Build a better product. The world waits.

[PSUinDC]

Just for clarification, that was not me... and I think Beach-Whale was being sarcastic... if not, just plain dumb.

[rmr1976]

This is also something to consider, although I'm sure it would take a lot of work to figure out. It would be interesting to see how your starting hands received compared to what would be expected through chance alone. Did you just get a bad run of cards? Or did some people draw out on you properly, because you gave a free or cheap card?

Although this might be too much programming work, it might be easier to simply keep track of loose or passive calls, hyperaggressive raises, or improper folds, based on the pot odds of each round.

Perhaps you could create a statistic that calculates the frequency of errors that Sklansky described in Getting the Best of It.

A function like this would be like the advisor function on Turbo Texas hold 'em, except it is based on real play (as opposed to computer opponents), as well as proper strategy.

[Beach-Whale]

Maybe I will release my own software to the public, maybe I wont. Either way, my view of PokerTracker stands firm. Heck - the GUI locks up while importing hand histories! Ever heard of calling Yield()?

[Beach-Whale]

Well, I guess you'd call me dumb then, because I was and am dead serious. PokerTracker does not look, feel or act like good quality software.

[Rudbaeck]

[ QUOTE ]
Just for clarification, that was not me... and I think Beach-Whale was being sarcastic... if not, just plain dumb.

[/ QUOTE ]

PT seems to be written in VB, and uses Access as a database. I don't think the technical side of PT is much to admire.

[bisonbison]

Hey, I'm not saying it doesn't have flaws. I could rattle off a few dozen things I wish were better, but the fact remains that PT doesn't have any competition as far as HE goes.

If a two or three man team wanted to create a competing product, I'm sure they could in relatively short order, but it would take one person months to replicate the basic functionality that PT users now take for granted.