i didnt try this out yet, but maybe you could let the simulation for the range equity breakdown last longer than regular calculations due to the increased variance per hand. e.g. Q6s is very bad, Q5s is better and Q2s is even better in your screenshot (if i understand it correctly).

this is a really great feature though, i think this helps a lot getting a better intuiton about why a given hand range has the equity it has.

Styx - Thanks for the +1 and feedback. Good eye on the indiv. cells' variance. I did mention this in the release notes; for this first release I couldn't find a solution.

I don't think giving an option to run the sim 2x or 5x longer will solve the whole problem, because:

* The simulation already is on the slow side (need to speed up).

* Say it takes 50000 random hands to reach an error margin of 0.3% for specific hand vs specific hand. Then you calc range vs specific hand where the range consists of 50 cells (evenly weighted for simplicity). Even if you run the sim 5x longer, 250000 hands, that's only 5000 hands per cell, which isn't enough for the desired 0.3% accuracy. I think you'd really need to do 50x the sim work to get cell accurate results! And if it's range vs range vs range then forget it! I'm fudging a bit here because the error margin doesn't scale linearly with the number of hands but should be close enough.

That said, I guess a normal/low/high num hands sim option might help anyway, eg on a fast 3gs you may always want a high number of hands. Look for this in an update!

But regardless, I think in the equity heat map we should only look for regions of red/yellow/green and can't always take an individual cell's color seriously.

Thanks for your suggestion, looks like an update is shaping up.

Still not available for Blackberry?

Sorry but no ports to other phones planned. See posts a page or two ago on this. I have limited bandwidth and my business decision is all in on iPhone platform, wrt new features and possible new apps. -RJ

Styx, and all - Re. the individual cells' equity variance and colors I just got a great suggestion from customer M.S. via email ... Instead of using just 5 (pretty different) colors use a gradiented continuous range of colors from red to yellow to green and interpolate each cell's color based on its (equity - breakevenEquity) value.

This will smoothen out the grid colors and also make them more accurate. E.g. right now if a cell is 9.9% above breakeven it's yellow, and if another cell is 10.1% above breakeven it's lightgreen. Using gradiented colors both cells will have just about the same yellow-green color. Like this suggestion? I do.

V.2.7 is just a first cut at the equity heat map feature guys, an experimental feature. Watch for it to get better in an update! -RJ

this M.S. guy is full of great ideas! i think thats an excellent solution.

i like how much time and effort you're investing in this, keep it up!

Yes M.S. is PokerCruncher's MVP right now.

Today I submitted an update V.2.71 with the continuous color range and a couple of other things (e.g. num hand combos printout for the range and for each cell in the breakdown text details). Will take about 10..14 days to go live.

Sneak peek at the new heat maps:

(1) Top50% vs. Top25%: This is the example I've been using so far. Some anomalies are still visible, but this won't go away til I can speed up the engine by say a factor of 10.




(2) Top100% vs. RandomHand:




I think these images are much better than V.2.7's, hope you think so too. Shows the benefits of community info. and ideas exchange. Thx, -RJ

Just bought this and the equity quiz program. Definitely the shiznit! Keep up the good work!

HA HA Ha Ha ha ha ha ha ha ha ....

All your smart phones belong to US!

Super, thanks for writing! I really appreciate the feedback from everyone on this forum and on iTunes.

Two updates submitted this past week:

Cruncher - continuous color range in equity heat map, show num hand combinations in detailed stats.

Quizzer - new multi-player quiz e.g. AA vs 4 random hands, top pair on flop vs flush draw vs open ender.

Both will go live about a week from now. Thx, -RJ

Why is it that the probabilty of a player making a pair (OnePair) when they have already been dealt a pair is < 100%?

See attached.

Thanks.

... Because the S stats view shows a hand histogram i.e. the OnePair, TwoPair etc. %age's add up to 100%. That is the OnePair number means the odds for exactly OnePair not OnePairOrBetter.

I guess this view could have a button that toggles between a histogram and ...OrBetter numbers, but when DealTo=Flop there's no space left in the view. So I don't want to try adding it. Maybe use a double tap.., but that's not too discoverable.

Thanks for getting the app! Do fire off any more questions. -RJ

Can you please comment on why the differences exist between making certain hands (indicated below) for Jacks versus Aces?
- One pair (see my post above)
- 3 of a kind
- (straight = ok; should be easier to make straight with jacks)
- Flush
- Full House
- 4 of a kind
- (straight flush = ok; again, easier with jacks)
- Royal flush

Mathematically, there should be no difference in the hands I've indicated. I assume it is based on the mechanics of your algorithm?

Thanks,

TC

Got it: this is a zero sum table of ex post realizations.

How you display it is another story: you've done a wonderful job with making these probabilities clear in this app, so I'm not sure I'm able to help ascetically! I might put an "N/A" in high card rather than zero -- helps ever so slightly to reinforce the point that that particular realization is not an option with a dealt pair.

... Because the app does Monte Carlo (random sampling) simulation. So there's a few tenths of a percent margin of error. However the app does do exact computation if all player cards are specific cards and all flop cards are specified. But with all the app's features e.g. hand ranges and DealToFlop and the various stats and odds for flopping draws, exact computation all the time would be very hard and I don't think worth it.

In your AA and JJ case, as you point out the diff in the straight %age is real and this steals a tenth here and there from the other hands. Other than that the other numbers look within the expected error margin.

I do have a long term goal of speeding up the engine. This will bring the error down to say 0.1%. But right now I want to keep adding new features, that you can't find elsewhere like the heat map.

Thx, RJ

Re. the N/A, yeah that would make it stand out but it's a lot of work to get right generally. The starting with a pair case isn't the only one. If the first 4 board cards are rainbow then a flush isn't possible so would need N/A here. Or if the flop is trips e.g. 333 then all players have at least 3OfAKind so need a few N/A's per player. I think too hard to know if you've covered all the cases.

So we could do this - if a %age is 0 then just write N/A instead. But due to Monte Carlo the app could show 0.00% even if the correct answer is 0.01%. So this doesn't work.

I think I'm going to invoke the [clean, easy, high value] feature principle here and not risk destabilizing for the N/A : ). Hope that's ok. Do keep sending suggestions guys; many features in the last half dozen updates were from you guys on this forum or emails. thx, RJ

Hi RJ...
Is it still available on AppStore? I couldn't find...
Tks and congratulations for your success!

Thanks CM! Small success so far, I'm not done with the app store yet : ).

You bet it's available, do a search for "pokercruncher". Noticed you're in South America. The app is English-only, so make sure you aren't looking on a local to your region/language app store. Maybe you need to go to the "worldwide" store to see it or change your language setting. I see sales outside the U.S just about every day so should be visible.

Do let me know if you still can't find.., thx, -RJ

RJ,
Tks for your response...
Still without success...My language is set to English.
As a matter of fact, I did a search over AppStore and I found several Poker related apps, such as PodOdds, Gamblers Anonymous, PokerHand, PokerOddCalculator...
Strange, isn't it?
I'm sure I'm doing something wrong...lol
I'll figure out!
Cheers!

RJ,
I found it...
My language was set to English, but my store was a brazilian store...
Sorry!!!
And thanks!

Or I haven't set up the app right for intl stores. But just confirmed on the app's configuration page and it says "this app will be on sale in all app stores worldwide".

Can you try looking on both your phone and on your pc/mac? And in the games->card and games->casino top 100 lists? (hadn't fallen off last time I looked) Or on a friend's phone/pc to double check?

Bummer, not good for the app ; ), hope we can get to the bottom of this.

Super! But according to the app's settings it should be visible in all app stores worldwide. I'm going to turn on each app store manually instead of using the blanket "all stores" setting. Prob. will take half a day to take effect. Maybe you've helped me get more intl visibility for the app, thanks.

And do let me know if you see improvements in the app. Thx, -RJ

RJ,
I don't have an Iphone yet, but I've already orederd one. It is expected to arrive this week... I hope at least...
Probably, you app will be my first one... I'll let you know my thoughts!
Tks!

Awesome. 1st would be an honor : ). Thanks.

I had problems finding it via search. I found it by visiting the official site though and following the link to the app store.

Poker Quiz comment: When hero is the underdog the odds are displayed in non-standard form: I.e. if hero has 13% equity the quiz answer says hero has 0.15:1 odds as a favorite rather than 6.66:1 against. It would be more useful if the odds were always shown in the "x:1" form where x >=1 and you say whether hero is a favorite or an underdog.