Hello rj999,

nice work on the app , it's quite easy to use and effective as hell

Quick question, maybe I missed something , but any plan to include randomly generated quizzes like SNG Wizard for practice ?

Thanks for good words on the app. Re. ICM quizzes that's certainly a good idea, would make the app a complete ICM tool. It was just a matter of what to work on first, I wanted to focus on the calc part and the two push/call/fold decision analyzers. ICM quizzes are on my long term to-do, but can't guarantee it any time soon, because right now I'm focusing on PokerCruncher-Android (development is well underway) and also prob. another OS-rewrite of PokerCruncher.

No idea if this is the best ICM app for iPhone but I bought it because PokerCruncher for iOS and Mac OS X is the nuts.

Can ICM Cruncher calculate a chop as was done yesterday at the WPT 5 Diamond High Roller event?

Discussion of deal:
- https://twitter.com/#!/SavagePoker/s...28184229429250
- https://twitter.com/#!/SavagePoker/s...30213995745280
- https://twitter.com/#!/SavagePoker/s...32057677565952
- https://twitter.com/#!/SavagePoker/s...33732886786048
- https://twitter.com/#!/SavagePoker/s...38444793069568
- https://twitter.com/#!/SavagePoker/s...84641318699008

Payouts: https://twitter.com/#!/SavagePoker/s...34733228281858

The deal: https://twitter.com/#!/JasonMercier/...41646737321984 (something seems amiss here because the chop leaves $123,920 aside for 1st and 2nd rather than $150,000 as Jason said)

So you're down to 3 players, you agree to set something aside for 1st and 2nd, then chop the rest. Can ICM Cruncher do all this, and if not, can you add such a feature for a future release?

Personally I think there's a lot of room for improvement with ICM Cruncher. I guess I'm comparing it to complex spreadsheets that offer a stack of info and are very flexible. And how about making ICM Cruncher universal for iPad?

>>> Can ICM Cruncher do all this, and if not, can you add such a feature for a future release?

I think the app can handle this now, I don't think a new feature needs to be added.

First as you say, the payout totals don't match across their Twitter posts and in their deal. In the tourney clock pic, the top 3 payouts add up to 1.930M. In their deal the payouts add up to 1.805M and adding on the 105K and 45K gives 1.955M so there's an extra 25K in their deal. So to me it looks like the tourney clock pic isn't up-to-date and 25K extra was added to the prize pool's top 3 places after that pic was taken.

So we don't know the exact payouts but I punched it into the app anyway and got a result pretty close to what their deal was.

The 1st pic below is the normal fair ICM deal.

The 2nd pic is what they did, if I understood what they did that is. I started with Pic1 and subtracted off 105K from 1st place payout and 45K from 2nd place payout and re-ran the ICM. The result is pretty close:

Their deal: "Doc got 619k I got 604k and Ike got 582k"

In the app: 612k, 597k, 576k

In the app each payout is about 6 or 7k less, but I think this discrepancy is explained by the inaccurate final payout numbers we're working with.


>>> I guess I'm comparing it to complex spreadsheets that offer a stack of info and are very flexible. And how about making ICM Cruncher universal for iPad?

This app is a pretty basic ICM app and is priced that way at 2 or 3 bucks. It's certainly not as advanced an app as PokerCruncher. Sure we can make it as complex as we want but then it wouldn't stay at 2 or 3 bucks. I've gotten suggestions to make it as powerful as 'wiz and add randomly generated ICM quizzes, etc. Haven't done it, just a matter of deciding the scope of the app.

Re. universal, this is much more of a possibility next year. Right now though I'm pretty busy on the Android version of PokerCruncher, which is in the latter stages and I can confidently say that it will ship 'soon' : ).

Thanks for writing with suggestions.


Normal fair ICM deal:
====================



Their ICM Deal: (leave out 105K for 1st and 45K for 2nd):
====================

Finally, there's an Android version of PokerCruncher now, on Android Market: link.

Have started a new thread for the Android version.

*****
Note: App has been renamed to "Tourney Cruncher" in V.2.1 (from the original name ICM Cruncher).
*****

Bubble factor is a great suggestion, thanks, I've added it in V.2.1, this update just went live today.

V.2.1 is a big update:

1) Rename the app to "Tourney Cruncher" to better reflect the expanded set of features in V.2.1.

2) New app icon.

3) Calculator for your M & Q ratios (measure the health of your chip stack as the tournament progresses).

4) Calculator for Bubble Factor (measures how big an ICM effect a player has against a particular opponent).

5) Help screens/explanations for M & Q ratios and Bubble Factor (the "?" button in the M & Q ratios screen and the "Help: Bubble Factor" menu button).

Thanks for your suggestions and support of the app (if you're so inclined pls. do write a review on the app store, would help the app/me greatly ).

Here's the app store link to the app: link, which has the latest screenshots.
(Looks like the latest screenshots aren't there yet on this page, but should be updated there sometime today.)
Thanks, -RJ

The App makes simulation for push/fold Chip EV situation? I mean with no payout structure.

And simulations with a raiser before the hero?

No, at least not yet. This app is tournament focused, and currently the push/call/fold decision analyzers take ICM into account, and for ICM considerations you need a payout structure (o.w. you're in a cash-game type situation). I think early in deep stack long structure tournaments you're almost in a cash game situation. But as soon as you enter the middle stages of even these tournaments, ICM comes into play to at least some extent and you can't make decisions based just on Chip EV, must consider $ EV.

That said, for very early in most tournaments, maybe a good feature addition to the app's push/call/fold decision analyzers would be a "turn ICM off" mode. This way you can compare the results with "ICM on" mode. I'll add it to the to-do list, but won't promise, thanks for the suggestion.

Yes. The decision analyzers have a column "Chips Put Into Pot" for each player. You can simulate any number/amount of raises and re-raises in this field. E.g. see the below screenshot. Hero bets 200, Villain raises to 400, and action is back on Hero who's deciding push all-in or fold.

For example, a torney begin with 2k players and the ITM is only for 10% of the field ( first 200 players ) , after some time we have 300 players left and we are short-stack, so how you recommend to configure the payout structure?

This app handles up to 10 payouts, and I don't think it will ever handle 200+ payouts. One reason is that ICM calculation time grows very fast as the number of payouts (and players) increases (definitely not a linear function). So around 10 is the maximum to compute the result in a reasonable amount of time. So this app isn't suited for your use case.

But maybe you can approximate your scenario in the app as it currently stands. What if you "chunk" the remaining 300 players in chunks of 30. So treat payouts 1..30 (average these payouts) as Payout1 in the app, payouts 31..60 as Payout2, etc., and similarly chunk the player stacks. Then you wind up with 7 payouts in the app, and 10 players left. However, I don't know if this would produce accurate enough results to trust, and I'm still saying that this app isn't suited in its current form for your 300 players example. Just an idea to try.

I understand... Nice idea, ty.

I find this app provides useful info when discussing chops, but it would be really helpful to be able to calculate $ev of everyone's stack at final 2 tables when chop/pay the bubble discussions often start happening.

I wouldn't mind waiting a short while for the app to run the calculations.

The problem is that "a short while" could be many hours.

Handling up to 20 players and 20 payouts (or 18) isn't feasible because the ICM algorithm is a very expensive recursive algorithm. Even for 10 players and 10 payouts it takes 5 or 10 seconds and I've added a "please wait" message. I estimate for 12 players it would take on the order of minutes. For 15 or 20 players and payouts I think we're talking hours, and for 20 prob. many hours.

Even the website ICM calculators I've seen handle only up to 10 players and payouts (have you seen one that handles 20?), and they have the advantage of being able to use a fast server machine.

A suggestion is to use the same "chunking" technique in my reply to "A9s" above. If there are 20 players left, sort the stacks, then average the 1st and 2nd, then the 3rd and 4th, etc. So then you have 10 stacks. Do the same for the payouts if needed. This reduces the 20 players/stacks to at most 10. But of course the ICM calc you get will only be an approximation.

On my iPhone 5 10 players, 9 payouts was calculated nearly instantly.

Is there a way to do the icm calculations for 10+ spots in way where computational requirements are reduced but we get roughly the correct result? ie rounding off to fewer significant digits, doing fewer iterations, etc.

Btw, what would the situation be w say 15 players and 10 spots paid?

Yes 10 players, 9 payouts is almost instant (wasn't on older iPhone's). But for 10 players and 10 payouts it took 2 sec., just tried on my iPhone 5. And as you go higher from 10/10 the exponential nature of the algorithm really kicks in and I think my guess of hours for 15/15 or 20/20 is a good one.

>>> Is there a way to do the icm calculations for 10+ spots in way where

Really good problem to think about. As a worst case approximation the app could do the same chunking technique I wrote about, i.e. reduce the 20/20 problem to 10/10 by averaging, and give a warning alert about the approximation/accuracy. A concern I have with this is generating inaccurate results, even with a warning note. I wouldn't want to give people the feeling that the app doesn't always give the correct results.

I won't make any promises but I'll think about it. But still, I haven't seen even a web-based ICM calc that supports more than 10/10, so this could be out of scope especially for a mobile app.

Not sure if I understand correctly, but e.g., the tourney started with 100 players, you're down to 15, 5 away from the bubble (or 4 away). This isn't a chop/chop situation yet but you may still want to run ICM calcs or the app's push/call/fold decision analyzers or bubble factor calculation. That is, ICM effects should be pretty strong here even 5 away from the bubble.

Yeah exactly. Also I'd like to calc change in $EV when paying the bubble or otherwise changing the payout schedule.

To calc change in $EV, you can use the "Save Scenario" feature. First set up the current "normal" payout structure and calc your $EV, and save as Scenario1. Then make the proposed change e.g. paying the bubble and calc your $EV again, then save as Scenario2. Can then go back and forth and compare. It would be nicer to be able to see the comparison side by side at the same time, but, ... maybe someday or if this app were a full blown desktop ICM app.

The limit of 10 stacks/payouts is still the main thing I think. I'll try to go beyond 10 someday but won't make any promises. At least for small tourneys e.g. nightly 7pm's or weekday tourneys you usually have fewer than 100 players which means fewer than 10 payouts so if you get down to say 10 players and 7 payouts the app gives you the ability to calc ICM and "what if" payout decisions even before you reach the bubble.

Also, the ICM algorithm is an exhaustive technique (look at all the cases to compute exact results), not a Monte Carlo approximation, so can't do fewer iterations or fewer digits.

You may be interested in the thread below. It contains a description of an iterative algorithm to approximate ICM equities for larger instances:
http://forumserver.twoplustwo.com/15...ments-1098489/

First reply there also contains a link to my (exact) 20/20 website calculator.

Thanks!
Great work by all on that thread.
I will add the new algorithm to my to-do list. Am working on updates to the PokerCruncher versions right now so it's probably some months away.

Is the algorithm of your optimized 20/20 calculator public code somewhere?, or will I need to do my own 'homework'? : )
Could it be implemented in a mobile app?, or does it require a server computer, large lookup tables, etc.?
Thanks for info.

It's not public at the moment, sorry. Can't think of any reason why it could not be implemented in a mobile app, memory requirements for 20 players are around 10mb iirc.

***

Happy Holidays RJ999 - I found a way into the forum once more, but I'm sorry to report my poker playing is near nil today. With what happened in the USA and losing most of my funds on sites that went out of business quickly (and rather dubiously) - I only have some $ on one site still today, and there are few games there. Anyway, the reason I really wrote you is because I found the thread still open, and I just got on the site again after almost 7 years. I was curious as to whether you are still active and playing? I still have your software on my iPhone, and look at it once and awhile - I can't bring myself to delete it - but I'm not sure I'll play anymore. Too bad, I really did love learning.

Take care of yourself. Just wanted to reach out and say hello.

Kevin

Hello,
Happy holidays.
I see, looks like you were username "k9_scruz" years ago? ..., welcome back.
Yes I'm still playing, maybe not as much as years past, but want to stay sharp.

>>> to delete it

Such blasphemy! J.k. I understand that our hobbies and interests can change over time, and sometimes we can return to old hobbies like poker.

Years ago you may have asked about training videos re. the apps. PokerCruncher does have a tutorial video now, made by an independent poker pro/coach who's using the PokerCruncher Mac-Expert version in his work. There's a link to it at the top of the app's tutorial & videos webpage.

There are also links to some other pros/coaches who use PokerCruncher, in the "Further Education" section on the webpage. Maybe these additional links/videos will help get you back into poker and these apps full steam.
All the best. -RJ