After a few months of private beta testing I am happy to announce the release of public beta version of Pokerazor! Visit our website: http://pokerazor.com to get the public beta. Check out our blog to see some examples. Pokerazor is a software of new generation which analyzes EV of diffrent strategies and answer most math questions you may have when playing holdem. It's not possible to describe all the features of it in a short post so I will give an overview. Here is what it can do:

- calculating frequency of events like: "How often do you flop a set", "How often do you flop any pair or any draw with suited connectors", "How often do you have at least' 2pair+ with given range on given flop", "How often will you have at least top pair by the river with range X on flop Y". It answers many questions of this kind, I did some examples here:

- calculating EV of multistreet strategies; it's possible to specify any strategy and settings (stack sizes, pot size, ranges, actions on diffrent streets etc);o ne example is:
"You raise to 40$ in 5/10 game with 33 and your opponent three-bets to 120$ with: TT+,AQ+,AJs,KQs. You call and check-raise all-in on the flop if you flop a set and fold otherwise. Your opponent always makes a continuation bet of 75% of the pot and goes broke with top pair or better, but folds otherwise; here is a screenshot from above mentioned calculation:



It's possible to analyze almost any strategy. I did more examples on my blog:

1) playing underpair in 3bet pot :
http://pokerazor.com/blog/2008/02/04...-reraised-pot/

2) raising for free showdown in limit:
http://pokerazor.com/blog/2007/11/21...howdown-limit/

3) playing back preflop against a loose 3bettor:
http://pokerazor.com/blog/2007/11/16/5/ (inspired by old CTS post)

You are welcome to add any questions to blog analysis or post your own
strategies on our forum.

- full statistical range analysis; it's possible to see how often your opponent has at least top pair or a draw on given flop; how his range changes if he plays any pair any draw on the flop and top pair+ on the turn; how often your hand is bets against his range on any street; what's actual equity etc. :








You can select range using a selector which is similair to pokerstove standard, but with much better UI:



To specify different conditions for which you choose diffrent actions or want to see frequencies you can use condition window:



Now it's possible to solve almost all poker problems analyzed in poker literature in just a few clicks, seeing how small changes in assumptions (stack sizes, looser/tigther range, bet sizes etc.) require just a few seconds and clicking START or CALCULATE again. I believe this software makes writing math equations obsolete (or will make in near future).

Our goal is to make ultimate poker software. We are aware that there are any possible features yet to be implemented, many things to improve and many bugs to fix. We invite you all to play with the software and share your feedback with us. Don't hesitate to write to us with any thougths/ideas/problems/bug reports you may have.

For general things you can contact us at : pokerazor@pokerazor.com
To report bugs or problems with installation please use our forum.
To share your simulations and ask other users what they think use strategy forum.
You are also invited to post comments on our blog or to pm me directly here on 2p2.

Any info on when its gonna be commercially released and what other features that version will have?

Wow this looks impressive.

gTg

Public beta will last at least 3/4 weeks we don't have final date for commercial release just yet. We plan to focus mainly on fixing bugs during that period but some features will be added for example "bluff evaluator" , "all-in evaluator" and "value-bet evaluator" are almost implemented. Those features will make it very easy to ask about EV's of these plays on given street (they will also show possible calling ranges and make a nice graph to see against what we are bluffing/value betting etc).
Other things depend on feedback we receive from our users.

In fact we got some respected 2p2'ers (and one not-so-respected BBV Hero) begging us not to publish it

looks very promising. will definitely try it out..

qft

Tell me how someone gets the time and energy to do this amount of work that doesn't pay the bills? There's a dozen personal projects I'd love to do if I could get off the "work to live, live to work" treadmill, LOL!

gg poker. can't wait until there are stickies with config files in each forum.

If this software can be linked to the PT databases, then there's just a small step left to build NL deepstack robots, right? I mean, if from PT you can extract villain's strategies for a given board and action (with enough sample hands), then you could use this software to find the strategy that maximizes the EV. Would this be possible?

punter,

sick program.

wowza... this thing is pretty awesome

wow
now i really have to mix up my play ^^
ABC poker will die soon,
this would be best to use offline like pokerstove.

Wow this is so sick.

Awesome looking program - although I think this is really starting to push the boundaries of what should be considered "acceptable".

as long as its worked offline and not in realtime there wont ever be a boundary of whats acceptable imo.

This is offline tool. It neither does nor will make any decisions for a player. It's not integrated with any online room in any way either.
It's for analysis only so "acceptable" is not an issue here.

Punter,

any idea of target price when this goes retail?

oh, and will it work in WINE?

Wow, this sounds like the program Sushiglutton proposed. I remember that some High-stakes players said this would be the death of poker.

To extract a villians strategies and produce a model of how they play "X hand in Y context" would require knowing their hidden cards and even if you could somehow deduce the hidden information from what's shown down (it's very easy to create simple strategies that cannot be deduced from observation - see my old post here), for deepstack NLHE it's even harder because very few hands actually go to showdown and it would require millions of observations to create anything near human comparable.

The alternative is to create action frequency models and shown-down hand distribution models, both of which don't require any knowledge about your opponent's hidden information. You can then use these to predict how they will act and if they show down a hand; what distribution of hands they will have. For a tool like the OP though, this method wouldn't help.


One another note, this looks like a great tool! Will it work with $EV (ie: using ICM) for use with SNGs or just with cEV? Something like this would make analysing stuff like stop-and-gos and post-flop play vs limping donks a whole lot easier.

Juk

We haven't decided yet. We will try to talk to people who might have better idea about potential market than we have. We would be grateful for any info or suggestions on the topic.

I am not sure. It's coded in C# and use standard click once installer. I don't think there are any obstacles but it' yet to be seen

I wrote about this idea like a year ago here on 2p2. It was a long way to get it going finally. While this is for sure very powerful tool it's far cry from killing or solving the game.
It makes the math part of the game easy (well.. at least much easier than before) you still have to insert ranges and strategies though and you need to be good hand/mind reader to do that correctly. While it's possible to see how diffrent strategies work against diffrent type of players you still have to know who you are up to in real game
. I thinnk our tool will make general population of players much more math-aware and playing will be much more in psychological area. Also there is a long-way from being able to doing. As with any tool the most dedicated/smartest players will benefit the most.

Thanks for this post Jukofyork. I hope it will calm down some concerns
As to your question abotu ICM model. It's definitely possible but it's not our priority right now. A lot of future features depends on feedback we are receiving now. For sure "tournament version" is one of the options.

Not true at all. You are about to see real ABC poker soon though :-)

I know. What I was trying to say was exactly that

You might want to take a look at the TurboTexasHoldem software, as even though it was a bit clunky, I found it pretty helpful when tuning my limit strategies years ago.

The ability to simulate your own set of strategies against others was prolly the most helpful feature of TTH and I couldn't tell from the OP if this was possible?

Another idea would be to allow the strategies to be encoded by the user as a decision tree. TTH tried to some extent to do this, but it was a pretty clunky GUI and inputting a strategy from scratch was a real pain (also it wasn't a "free-form" decision tree and had a fixed structure which wasn't ideal). Even so, being able to start of with some very basic (modifiable) strategies like "nit", "calling station", "set miner", etc (each encoded into a fairly simple decision tree) would help massively when trying to fine-tune your own counter strategies.

Juk

sry, after trying it tonight
i was wondering what's the main different comparing your product and pokerstove?

Level?