Feature request for cruncher.

Entering hand ranges: All-in EV has this feature where if you hold down shift when clicking on a hand like AJs, it adds in the hand range AJs+. Ditto pairs. E.g. Shift-88 becomes 88+.

Please add something like for cruncher. E.g. an Add+ button. This will greatly speed the entry of large custom hand ranges.

Thx, yes it will speed up entry, will be in the next-next update.

Right now the fastest way to select e.g. 88+ is to tap on 88 and hold down/swipe diagonally to AA. Can swipe horiz/vert of course to select e.g. AJs+ or AJo+. When hand ranges first shipped this swipe feature wasn't even there.

Also I submitted an update a couple of days ago, some small UI improvements. Thanks for ideas.

Thanks for suggesting ICM (several people). My newest iPhone app "ICM Cruncher" has gone live (thread).

This is not just an ICM calculator but also a push/call/fold decision analyzer. Of course do post feedback on the new thread if you have any. Thanks for your suggestions this year.

Another thing - funkyj's latest suggestion the Add+ button for ranges - this will be in the next update V.2.92 which should go live in about a week.

A weird thing happened today, the notification for V.2.92 went out but when you downloaded you got V.2.91. This problem is fixed now so pls. download again if you ran into this. The problem looks to have been on Apple's side; I didn't do anything except send them an email.

V.2.92 has some UI improvements e.g.:
* Add+ button in range builder screen. [funkyj]
* Improved green/red/yellow equity color selection.
* Slider in range equity breakdown heat map screen to select upper/lower bounds of the heat map colors.

Raju, I have a question. Are you married? If not, would you consider a proposal?

I am bright, hard working, very clean, and most pleasing to look at after you've had a couple of drinks.

Please let me know.

[PS By the way...I'm male, hope that's not a problem.]

Lol, of course I'm not married, how can any married person have time to work on apps at various hours, and play some hands too?

All sounded great until the "male" at the end, not that there's anything wrong with that of course : ).

Hope you enjoy the apps guys.

Hey I'm trying to buy from the itunes app store but ---> "this item is not currently available" how long will it be?

Should just be a temporary thing; I don't have any control over this. I've seen this before on my apps and other apps.

You could try buying on your device instead of on your pc/mac. Or the old last resort try, restart iTunes. Thanks.

Great app, rj999.

Suggestion: Move "Reset hand range" from Menu sub-screen to main Range screen and relabel it "Clear All". It's current location is unintuitive and requires two clicks, less than optimal for a function that is used so frequently.

Thanks, glad you like it.

I'll rename ResetHandRange to ClearHandRange. But if we move it into the main range screen, I think it would need a "are you sure you want to clear?" alert, because o.w. you could tap it accidently and lose your range construction work. So it would still be 2 clicks. If it's in the Menu we don't need this "are you sure?" alert (coloring it red to indicate a destructive command is the UI convention).

Moving it out of the Menu would improve its visibility as you suggest. But the range screen is pretty full and we'd need to move one of its current buttons into the Menu to make space. Maybe the Select and NextCell buttons aren't needed any more because you can tap/swipe the range grid so I can remove these to make space.

Given the pros/cons I'm not super excited yet about moving it out of the Menu. Also in the main calculator screen, the several Reset commands are also in the Menu. One could argue that ResetAll here is even more heavily used than ClearHandRange so it should be in the main screen. But right now the convention in the app is to have all Reset commands in the Menu. I'm open to changing this but someone will have to suggest a way to make space for the Reset commands in the main screen.

Thanks for feedback, it's an iterative process.

Obviously this is a matter of preference for users. I guess you could poll users here if you wanted. Personally I would prefer a "Clear Range" btn on the HandRange screen without the verify process. If there is room to put it up top, accidental clicks would be rare I think. But this issue is obv minor.

Another issue isn't so minor though. Could you change the way you calculate your flop texture made hand stats to correspond to the way flopzilla and other flop analysis sw does? Your current approach of counting board pairs when calc prob of hitting a pair or two pair is quite misleading. For example, your stats for hitting two pair with 75o is ~ 4.0%. But obv we really want to know prob of 2 pair using both hc's (2.02%) giving us a very strong hand and not merge that with hitting 1 card on a paired board which is a very weak hand. Similarly misleading to have unpaired cards hitting a pair on the flop 40% of the time (I wish!). A flop of KQQ isn't really that hot of a flop for 75o.

Same for straights and fh's on the board - technically they should be removed from flop hits, but these happen so rarely its almost irrelevant. And when calc to turn if you miss the board completely but there are 4 of a suit or 4 to a straight on board this shouldn't be calc as a flush/straight draw.

Flopzilla has the best flop analysis layout I've seen as well as a fairly user-friendly interface. I imagine you've checked it out already, but if not it really does have some of the nicest implementations I've seen for flop analyses.

Just want to say I'm impressed with your dedication to improving your software and the speed with which you incorporate your users suggestions. Good business model -wish more of the business world took this approach!

Agree, let's skip/postpone the minor ClearRange button issue and get to the meaty stuff : ) ..

I completely agree that settings like "must use both hole cards", "must use at least 1 hole card", etc. would improve the flop analysis stats. It's on the todo list but I haven't gotten there. A few people have suggested this, the last I remember was customer M.S. (whom I also talked with in person in Vegas later on this issue and other heat map analysis features).

For now, the app has stats that approximate "must use both hole cards" that some customers may not have noticed yet. E.g. if you enter AKo and set DealTo=Flop, the OnePair stat is 40% as you say (includes the cases where the board pairs), but if you tap the "->" S button in the stats view you'll see stat "HitBoard'sTopCard" which shows 32% (tap it once more to see the stats for combination draws). This stat measures the hole cards' "pure" hit %age and doesn't include the cases where only the board pairs. Similarly below this stat is PocketPairAndNoOvercard.

What I'm saying is I should implement a "must use both hole cards" setting as you suggest for generality across all stats not just TopPair, but currently there are some minimal stats to help give you a picture of a "pure" hit.

Yup I've seen the excellent tool FlopZilla and hope PokerCruncher's flop stats can match that tool's someday. But FlopZilla is a special purpose tool with a specific focus, which may be one reason why it does what it does so well (along with, it has desktop real estate to work with). PokerCruncher may be trying to be too much of a jack of all trades - 10 player fully general hand range equity calc, flop texture analysis and stats, DealTo=xxx, heat maps, etc., on a small screen, with a wimpy processor, at a fraction of the price!

But being a general jack of all trades tool does have advantages - e.g. FlopZilla requires 3 specific flop cards to be entered, but in PokerCruncher you can enter just 1 flop card or leave the flop all blank (random), set DealTo=Flop, and see what made hands and draws a given hand range or hole cards can flop and how often.

Again I have a 100% positive opinion of FlopZilla and didn't mean to get into comparisons, but it came up.. In the end the comparisons will improve PokerCruncher (e.g. after I implement "must use both hole cards") so I'm happy. Thank you dfan for writing and bringing up these issues. It's customers like you who make things happen (and there are many on this thread). Best, -RJ

(and stay tuned for an app update early next year)

When will we see a PLO quiz program from you

Good idea, the existing Quizzer app can be leveraged a lot, but can't promise it ! (next year's work plans not set yet).

Check out this "ouch!" 1-star review posted on iTunes this week:



I can't repro this, I get the RoyalFlush as 100% equity. If anyone has seen similar problems could you please post? Felt ok posting this review here because the review writer also said he wants to ask people if they're seeing this.

I added a note to the app description saying I can't repro and maybe DealTo or another feature was set to something he didn't want. Thx.

=====
PokerCruncher V.2.92

(Equity Tie Win)
Player 1: 0.00% 0.00% 0.00% [Ah,Ad]
Player 2: 100.0% 0.00% 100.0% [Ks,Qs]

Board: [Ts,Js,As,Ac,7d]
Deal To: River
Dead Cards: []

Complete Enumeration: 1 of 1 possible hands
=====

Obviously the reviewer can't replicate it either, since he says "seven or something" in the review. I'd say there's a 98% chance that he's using two-color cards and did this:



The small iPod screen makes it a lot easier to mix up those black cards. Or he could have been drunk and/or dyslexic and mixed up the players in his head...

Nice guess. The 4-color suggestion really helped with distinguishing the cards.

I'll add a guess to the list: he accidently set DealTo=HoleCards, which means a 2-card game (sim doesn't even look at the board cards):

Hmmm, now that I look at it, I think your guess is more likely than mine. Hopefully the reviewer will realize his mistake and edit the review.

I tried the quads vs royal and it works as it should for me too.

Hey, a few questions:

1. Have you thought about offering a (slightly-discounted) all-in-one package?

2. w.r.t. optimized hand evaluators, have you searched the forums here? e.g. this thread had some great comments from brecher, prock, jukofyork, etc.

3. out of curiosity, how long did it take you to figure out the colors for your heat map? I actually thought the color map for this old jam-or-fold calculator was the hardest bit to code!

Otherwise, looks very cool. Thanks!

Does the oddsquizzer quizz you about hand vs ranges or just a hand vs specific hands?
Also if it's not just ranges something in the line with:
Player 1 has XY flop is ABC, Player 2 has TP, FD or Air, what's Player 1 equity?

Hi -

1) Not yet but will consider making a 4 apps in one total odds package app for say $8 instead of the sum $10 (unless combining the apps proves to be too much of a technical pain).

2) Yup I've seen that thread. But I consciously went with adding more features and stats over simulating more hands per sec. Respected posters on this thread agreed with this plan many pages ago. Plugging in a faster evaluator is on my long term list but I still want to do things like the recent "pure flop hit stats" suggestion. Customer M.S. has even emailed me code for a fast evaluator.

But look at the big picture - currently the error margin is a few tenths of a percent. Will getting the error margin down to one tenth of a percent really make a huge difference, other than from a purist standpoint? How does this compare to the additional features beyond PokerStove? - Deal to flop, flop texture analysis, histogram and draws stats, heat map, etc?

3) The heat map colors really didn't take a lot of time and are implemented with only a half dozen or so lines of code. The colors aren't hardcoded but are computed at runtime by interpolating the RGB values based on (cell'sEquity - breakevenEquity). This runtime interpolation also allows easy implementation of the colors min/max spectrum slider in the heat map screen. I should probably shut up and not tell potential competitors how I did this, but, whatever, go for it!

Thanks for your suggestions and questions.

Hi rj,

re 2: I certainly see what you mean about the desire to add more features. I just wanted to point out the thread in case you hadn't seen it. I've played around a lot with this kind of stuff on my own, and having a faster hand evaluator always allows me to add even more features. If I recall correctly, there was also some nice discussion in that thread about the merits of various different Monte Carlo strategies, something that often gets overlooked (the state space for poker hands can have some really weird edge cases so that your Monte Carlo code works really well most of the time, but bungles some important things).

re 3: Heh. Yeah, I just meant that I spent *forever* fiddling with the color interpolation code and then Prock spent another full day fooling around with it until it looked good. It's not like it was deep or conceptually tricky, but I was certainly surprised with how much tweaking it took. Looks like your color-sense is just better than mine .

Cheers

Yes Hold'em Odds Quizzer has hand range quizzes, pre-flop and post-flop quizzes, 10 questions each.

E.g. post-flop, like you describe, here's one of the scenarios - TopTwoPair on flop vs. range:{OverPair or Set or TopPairTopKicker or FlushDraw or OpenEndStraightDraw}, given a specific flop and player1 cards and player2 range.

E.g. pre-flop, QQ vs. range:{AA-KK, AKs, AKo}. This is a standard dilemma you can face with QQ pre-flop if you find yourself re-raised big by someone that can play AK aggressively.

Thanks I'll be going through that thread again. I appreciate you guys sharing your findings.

Re. heat maps, the only differences I see between yours and PokerCruncher's are:
1) PokerCruncher uses darker colors for the endpoints of the color range for more contrast.
2) PokerCruncher uses one continuous color range instead of two, one for positive equity and one for negative equity. So in PokerCruncher if one cell is 0.1% above breakeven and another cell is 0.1% below breakeven then both cells will have almost identical yellowish colors, which makes sense because the cells are almost equally profitable. In your heat maps one cell will be yellowish and the other grayish. In PokerCruncher the Monte Carlo margin of error can easily move a cell across the breakeven boundary, more reason for using one continuous color range.

Regards, -RJ

RJ,

First, love the app. Great stuff! Second, can you add a feature that will show you only the range of hands that have the desired equity vs villains range?

The range breakdown feature comes close but since some of the colors merge it can be a bit hard for me to tell which hands really are profitable. What I'd like to see added is a button on the range breakdown window that will show you only the hands that are >= the desired equity, and then be able to export it.