Thank you for the quick response!

^
You're welcome, my pleasure.

So the summary is the PokerCruncher Advanced/Expert versions (iPhone, iPad, Android, Mac-Expert) are separate apps/purchases, because that's the only workable way to set it up in the app stores.

I'm committed to making sure customers get value for each purchase. I've coined a phrase "pay once and enjoy forever" in the descriptions of this app in the app stores / our website. I.e. if you own an iPhone or Mac etc. for the next 50+ years, you'll pay just the first time for that version. There are no further continuing in-app purchases/fees, and all app updates/improvements are free (and there have been many over the years).

Thanks for interest in PokerCruncher.

Hello RJ,

Recently bought both the Mac and iOS versions and its awesome.

I'm also using the app to study and construct my preflop ranges and would like to know if allowing coloring/subgrouping of a loaded range is in the pipeline soon?

This is important in range construction because it allows you to see how balanced you are, e.g.

- if for my MP range I'm opening 23% and 4b jamming the top 5% for value, I would also like to see if i am able to allocate the weakest 3-5% of my range as 4b bluffs and which parts of my range are those.

I seen in previous threads that you suggest loading/saving the subranges in other slots but that is a bit cumbersome and its harder to visualize it that way. Further more, it gets even more complicated if you're trying to construct and analyze a mixed strategy where you have limp/fold, limp/call, limp/3b for value, limp/3b for bluff, raise/fold, 4b/jam etc as parts of your range.

Thanks and as a fellow software engineer, i can appreciate the hardwork and passion that you've put into this.

Regards

Hello,
Thanks for compliments on the Mac and iOS versions, and for sending a well thought out suggestion.

Your suggestion (color coding ranges into subgroups) has come up more than a few times before and so far I've postponed it because I didn't think it was a general enough and workable enough solution for subranges. For example see my post #181 in this thread, and I'll explain more here.

I agree that the current solution for subranges (save them into the 100 range slots) is a little cumbersome. I'm planning on doing something to improve this and make it faster. I'm not sure if it'll be the coloring approach, or a "layering" approach where you can quickly select which layer (subrange) of the original range you want to see / edit / work on (without having to open a pulldown list as currently). I expect to get to this a little later this year say around the late summer or fall timeframe. But I won't promise anything yet, but it is my goal.

Definitely. The current method does let us see how balanced our subranges are pretty easily: when you open the quick-save-range and quick-load-range pulldown lists, each range's # of hand combos is written in the lists so you can check the balancing right there and do some addition verifications on the combo counts. I hope to make this faster in the future.

And these pulldown lists have other features for working with subranges e.g. adding/removing subranges; tutorial section on this:

http://www.pokercruncher.com/ipPoker...ngeRefinements

A good example, and coloring the hands/cells of the range would work well here. But I think the coloring approach works well only for these kinds of simple preflop ranges/subranges, where each cell of the range really only needs to be 1 color (in 1 subrange).

Consider your (and Villain's) ranges and subranges on the flop and turn. Here flush draws will enter the picture, and backdoor flush draws. So then a cell/hand of the range could need to have several colors on it if its combos are split up among a few subranges. It's not impossible to implement, but it could end up looking pretty messy, especially if the colors are spread out among non-contiguous cells. And it's not just flush draws that cause this; e.g. even for preflop ranges, for balancing you may want most of the combos of e.g. AA to be in your 3bet range, but 1 or 2 combos to be in your flat-call range (e.g. smooth call preflop in position). So weighting (using combos or the app's %age weights feature) can also complicate the coloring picture.

Another good example, with 6 subranges. The coloring for this would start looking messy for sure imo, especially if the 6 colors aren't all next to each other, which is what would happen in real life I think, especially postflop. And some people can't see some colors perfectly well, partial color blindness, so I hesitate to make a big feature depend so heavily on colors. I'm not sure if I'm right but I think colors can be used to enhance a feature or the UI, but having colors be the necessary / make or break thing behind a feature, I'm not sure on that approach.

Thanks for words of support, and for a very logically presented suggestion.

My summary: I'll do something later this year to improve working with subranges, but it may not be the coloring approach, could be a layering approach. I'm not sure, and won't make any promises, but it's my goal to do something on this.

Thanks for the detailed reply as usual. Looking forward to it!

Hello wonderdog78, and all who have suggested a "custom-color subranges" feature before ...

I'm still not sold on custom-coloring subranges for reasons I've given before (e.g. see my post #254 immediately above). However I've made an improvement for working with subranges (and all of our saved ranges). An app update V.11.6.1 went live today in the Mac-Expert version with a new view called "Range Manager". This view should be more convenient than using the quick-save/quick-load pulldown buttons at the top (which you can still do).

A screenshot is below.
Click the new "RngMgr" toggle button in the Range Editor view to show the Range Manager view.

Note especially the Add/Remove/Select buttons. These let us add a subrange to the current range, or subtract a subrange, or just select a subrange's cells.

The screenshot is showing some of my subranges as an example. If you name and organize your ranges/subranges well in the 100 range slots, that will speed up your workflow.

So, the ability to custom-color subranges isn't there, and I'm not sure if I'll ever be sold on that, but hopefully this new Range Manager view will make things easier wrt subranges. Do let me know what you think. -RJ

Thanks RJ, looking forward to checking this one out.

Glad to see your motivation to keep this product improving!

Hi, I really like the product and it seems to become more and more useful as I continue to use it and updates keep rolling. The adding of the range manager and being able to filter out a subrange has definitely made analyzing hands much easier.

One bug I found a little annoying is that once a range is filtered there will sometimes be 0 combos of a hand left due to card removal effects but the it will remain selected in the range visualizer. It would be great if the hand would become automatically deselected once there were no combos remaining. Notice in the attached picture that there are a few hands that have 0 combos.

Hello,

Thanks for good words on this program. The recently added 100 additional quick-save range slots should also help with our ranges/subranges work I think (200 total slots now).

A question I have in your example is, are the cells/hands that are showing 0 combos really empty (e.g. the KJs cell), or are they empty just due to card removal effects? I couldn't tell from the pic as the range's text on the LHS is cut off. Giving the scenario's text ("Txt" button) would show the situation fully.

You can check this for the KJs cell by SHIFT+clicking on it to select the cell, then look at the 4x4 Suits grid and see if any combos are turned ON. Two cases:

1) If the cell is really empty (no combos are turned ON in the Suits grid), then yes this is a bug; the cell shouldn't have a yellow(ish) color. I'll fix but I will prob. need your scenario text and rough usage steps to reproduce (if I can't reproduce it on my side right away). This would be a weird/buggy state for the range; I haven't seen the "Filter" button generate such weird states for a range yet; if a cell has 0 combos after filtering it won't be yellow(ish) color.

2) The cell isn't really empty (some combos are turned ON in the Suits grid), but is showing a combo count of 0 due to card removal effects. In this case the current behavior (showing a yellow(ish) color for the cell) is by design, but it's certainly arguable and I may make a change here after we discuss more.

For the 2) case here's my thinking for the current behavior: These cells aren't really empty, they have some combos turned ON, so they should show a yellow(ish) color to indicate they're part of the range. Another indication is that the range's text on the LHS will also show some combos for these 0 combo cells. If we make a behavior change here and un-yellow these cells, we'll also have to delete these combos from the range text to stay consistent, and this has the potential to lose info. about the range as we move on in our usage. For example if we enter a different flop, which leads to different card removal effects, then the 0 combo cells may now have > 0 combos, and we would need to bring these combos back into the range, which we can't do if we permanently removed them earlier.

So I took what I thought was the safer approach and show all of the combos as being part of the range when we're editing / looking at the range, but also give a card-removal-effect setting and show a combo count for each cell which gives us the card removal effect picture. So we have both pictures for the range, the "baseline" picture, and the card removal effect picture.

Again I think this is arguable and I still may make a change here, but first wanted to explain the current behavior to give us both some reasons for this and some more background thinking time. Thanks, -RJ

It is this second scenario that is happening. Now I understand that it was an actual design decision and I can definitely see why this behaviour makes sense. Perhaps one or the other could be the default behaviour and then the other is a toggle?

The situation where the current behaviour is causing me problems is when I try to figure out what my range will be across multiple streets e.g., I start at the flop and filter out my value bets and bluffs on the flop and then the turn leaving me with my check down to river range. I then use this range to help me pick my bluffs. The problem is that these 0 combo hands still show up as selected so I have to manually deselect them in order to have a more accurate visual picture of my range.

EDIT: Upon further thought, I don't think the problem is so much a result of the range visualizer but rather how the filtering works, which may also be a design choice. In the example given above, suppose that I decided that I wanted to bet all my top pair hands. I would filter for top pair and save it as a range then go back to my original range and filter out those hands. The problem is that when I filter for top pair hands, it doesn't include the combos that are no longer possible due to card removal. Consequently when I then filter these hands out of my full flop range, these impossible combos remain in the range.

Thanks for describing your usage path more. I see why this is happening now and have an idea/plan to take care of this.

What was throwing me off earlier was the "range is filtered" wording ("once a range is filtered there will sometimes be 0 combos of a hand left due to card removal effects"), which I took to be using the "Filter" button. But this button never generates such cells with 0 live combos. I see now that what's causing this is not a filter operation, but a filter-out operation or negative-filter or just range-subtraction.

What I'll do is, create a new command/button somewhere that "minimizes" or "cleans up" the range you're looking at i.e. removes all un-assignable combos wrt card removal effects. You'll want to make sure the original range and your subranges are saved into slots so you can go back if needed. Don't know what exactly to call this button/operation yet, maybe also "normalize" a range, and don't know where to put it in the UI but those are just details.

This will be in the next app update in about 1..2 weeks.
Thanks for writing in.
-RJ

Thanks for the quick response, keep up the great work!

Thanks blitztime.
Your suggested new operation is now live in V.11.9.1:

* CTRL+[Filter] = filter hand range on all assign-able hand combos (if "View Card Removal Effect" setting is ON).

This removes all un-assign-able hand combos from the hand range to minimize/clean-up the hand range's hand combos.

As this is part of the filter-hand-range operation, you can use the <- and -> buttons under the "Filter" button to go back and forth and compare the before and after versions of the range.

It appears that PokerCruncher's calculated odds differs from those presented in a WPT tournament video. It was heads up, Jacks (hearts and diamonds) vs. Ace clubs, King hearts. The WPT odds are 56.3 for pocket jacks (my own Monte Carlo program gave the same within the uncertainty). PokerCruncher gave me 62.3% for the pocket jacks.

Is there something likely that I might have done incorrectly when I set the calculation up?

Thanks to anyone who can help me resolve this problem?

I get 56.85% equity for JJ in PokerCruncher-Expert-Mac (my scenario text is below).

I checked this on cardplayer.com's very basic online calculator and got 56.85% equity there too.

Not sure why the WPT result is 56.3%, along with your own personal program's result. Maybe it's a Win% vs. Equity thing. Or maybe the WPT result includes some known/dead cards that other players folded. My money would be on 56.85% equity being the right result.

==========
PokerCruncher-Expert-Mac V.11.9.2

(Equity, Win, Tie)
Player 1: 56.85% 56.65% 0.411% [JhJd]
Player 2: 43.15% 42.94% 0.411% [AcKh]

Board: [? ? ? ? ?]
Deal To: River
Dead Cards: {}

Monte Carlo Simulation: 12776425 trials

Notes:
==========

The best way to get to the bottom of these things is to give your scenario's text ("Txt" button at the top of the LHS in PokerCruncher-Expert-Mac), as I did above.

I'm guessing you set DealTo = Turn by mistake, instead of (the default) DealTo = River. This would end the hand after 4 board cards, not 5, giving JJ more of an advantage in this matchup. Deal-To is a very important feature of this program e.g. setting Deal-To = Flop and entering in a range lets us do flop texture analysis against a range.

Thanks for using this program.

Thanks much. Deal to Turn was the difference. And now I have another feature of the program that I can explore and use! Nice!

My Monte Carlo result was 0.568 +/- .009 so it was consistent with either 56.85 or 56.3%, but not 62.3%.

Great, we're in sync on the right equity result (PokerCruncher, cardplayer.com calculator, your personal program).

+/- .009 is roughly +/- 1%; that means you must have been doing a fairly low number of Monte Carlo trials in your personal program, say on the order of just 100,000 trials. Roughly what I've seen is that on the equity stat Monte Carlo gives an error margin of +/- 0.1% after about 1 million trials, and +/- 0.01% after about 10 million trials. In a desktop program, I think 1 million would be the right minimum number of trials for Monte Carlo.

So I think the 56.3% WPT result that you saw is still a mystery and looks incorrect to me, unless they were taking folded/dead cards into account or something (which we can model in PokerCruncher if we knew what the folded/dead cards were).

Yep. 800,000 trials. I was just trying to get a quick result. My Mathematica program is great for programming ease, but runs inefficiently. I haven't used the program much since buying PokerCruncher so it took a while to remember how to set it up.

I suspect WPT did account for folded cards. That information was available.

A question that I have been meaning to look up. Does my iMac purchase allow me to use the iPhone app? I seem to recall the answer is no. I prefer a computer to a phone, but it might occasionally be nice to have it on my phone when I am out playing poker.

No. The different versions of this app (iPhone, iPad-specific, Android, Mac-Expert) are separate apps/purchases. The main reason is that is has to be this way because all the versions are sold on the various app stores. The MacAppStore and iOSAppStore are separate app stores with no app-bundling possible between them. With a set of iPhone apps bundling is possible, and there is a discounted 4-app PokerCruncher apps bundle in the iOS app store.

Another reason I think separate purchases makes sense, at least from my/our point of view : ), is that making the different versions of this app, at this quality/power level, has been and continues to be a good amount of work over many years. The different versions aren't simple UI scale-ups, but are complete UI redesigns to nail the app/interface on each particular device.

Also there are no continuing / in-app fees; each version is only the initial purchase, with free app updates forever, so hopefully they will be worth it long term.

Thanks. I thought that was the case. No complaints. It is reasonable. You are doing a lot of good work.

Thanks; appreciated! -RJ

Hi,

While tinkering around with pokercruncher, I came up with a few feature ideas:
* make hand distribution view more interactive by:
1. ability to filter range to only include hands above certain equity %
2. ability to select and deselect hands in range while in distribution view
* filter for top X% of range by equity instead of top X% of all hands

I think these features would be useful when trying to pick which hands to bluff with, float or bluff catch.

This program already has a "Filter Range On Equity" feature, the "E" button to the right of the "Filter" button in the Range Editor view. Have you tried this out yet?

This "E" button does almost exactly what you're suggesting. The difference is that it does cell-level filtering on Equity (using the heat map's data which is cell-level). In the next app update I will add another option to this button to do combo-level filtering on Equity using the distribution graph's data, which will give what you're suggesting.

I'm not a big fan of this because I don't see it as being convenient. The distribution graph shows the range at combo-level resolution, so there are usually dozens or even hundreds of combos in the range. It's hard to select that many combos by hand, and even if we do, the Range Editor view can only show cell-level selection, not combo-level selection. I think the best bet is to just do a filter on Equity operation in the Range Editor view to select/narrow-down to the combos you want, or do two filter operations, to pick off the combos-range that you want e.g. all combos that have between 75% and 100% equity.

I think we'll be able to do this without adding anything new: in the distribution graph, when you click on a point/combo on the graph, under the graph a line like the following is written:

24% of combos have >= 81% Equity

As you click on different points/combos on the graph from top left to bottom right, you can stop when you hit your TopX% target in the first number above. Then look at the second number to see what Equity that corresponds to. Then do a filter on Equity operation in the Range Editor view at that Equity value.

Fully agree that this would be useful. It's one of the reasons I added the distribution graph in the first place. The shape of the distribution graph can tell us a lot about the range, for example if the range is polarized or non-polarized, and what that means for betting, bluffing, and check/calling decisions. I added a note about this some years ago in the distribution graph section of the tutorial; I hope my thinking was right when I wrote that.


Re. the next app update where I'll improve the "E" button to do combo-level filtering, it will be at least a few weeks away (prob. first half of Jan.). The reason is that I'm in the middle of a big round of updates on all the iOS and Android PokerCruncher apps and versions. Then I'll get back to the Mac-Expert version. Thanks for suggestions and ideas.

Hello RJ,

Thanks for this awesome software and for continuing to update. Great stuff!

Troubleshooting question: My PokerCruncher seems stuck at Version 11.5.2. The App Store does not provide me any way to download the newest version.

Any help would be appreciated!

Best regards,
boricua

That is odd, there must be something weird happening on your machine because updates are happening OK for everyone else AFAIK (I get app update counts daily).

First some questions: what exactly is happening? When you click the Updates tab in the MacAppStore you don't see the PokerCruncher app listed there? Or do you see it listed there but when you try to update the app the update fails (error message?, stuck in the middle?, does't even start up?)?

The app purchase/install/update process is handled completely by the App Store's software, so we're in the same boat on this issue guessing why this is happening, but here are some ideas to try:

1) In the MacAppStore app, open the "Store" menu at the top and verify that you're sign-ed in correctly (using the same id/password that you used when you bought this app). In any case, sign-out, then sign back in, to reset things. Then check the Updates list again and see if this app is there.

2) If still no success, try restarting your Mac, and try again.

3) This may seem like a weird thing to try, but if still no success, try using a different internet connection, e.g. your phone's hotspot service or a friend's, or a coffeeshop's wifi etc. This may seem totally unrelated, but I heard from a customer via email recently that they noticed a difference/glitch w.r.t. downloading apps in general (not just the PokerCruncher app) in the MacAppStore because of their particular internet connection. Longshot but worth a try.

If still no success, try a google search "can't update mac app store app"; prob. others have run into this and posted online.

In any case do let me know how it goes and we'll keep trying to come up with something.