(Close Window)
Topic: Something wicked...
Message: Posted by: AMcD (Sep 19, 2014 12:56AM)
Er... this way comes!

[img]http://www.arnoldmcdonald.org/temp/1.077.462.png[/img]

(just a small overview, sorry)

Full pix here: http://www.arnoldmcdonald.org/temp/1.077.462.png
Message: Posted by: JFX (Sep 19, 2014 02:17AM)
I have absolutly no clue about coding (unless you let MatLab or VBA count), so I can only imagine how much time and effort you put into this. It looks great.
Message: Posted by: AMcD (Sep 19, 2014 08:51AM)
It's quite some work yes...

But what you can do with it (or will be able to do, it's far from being finished) is just second to none.
Message: Posted by: Daggers (Sep 19, 2014 04:17PM)
What exactly is it and what does it do? Looks really interesting
Message: Posted by: AMcD (Sep 19, 2014 06:49PM)
It's a program, let's say a "gathering" of all I have coded about cards in the last decades. More or less.

When finished, it will allow virtually anything you can do with a deck of cards. But on a computer/tablet.

Manipulations: from Faros, Reverse Faros (full or partial) or cuts to switches, transfers or dozens of shuffles (peel-offs, reversing, Aussie shuffle, overhand/riffle, etc.). As you can see in the picture above, it's very "visual", thus it's a good tool for devising routines, finding new effects. You can't see it on a screenshot, but you can change the back of any card, you can "track" up to 16 cards during the various manipulations (their position into the pack). You can simulate deals using heaps, and there are tons of ways to gather the cards back. It would take 1 hour to describe what that first module can do. For instance you can study deals. Say you wanna know who is gonna win if dealing a round of Omaha for 7 players. Or Hold'em. Or Stud, or Draw Poker. I list all possibilities for all cuts, all players... There are tons of statistics, tables. There's a live clipboard collecting all data. In a click you can get a nice table, just in case, like me, you use such tables in your writings :-). Of course, you can export the generated pictures.

But this is crap. Indeed. The rest is 50 times better! You will be able to compute odds (and not the ones everyone does, trust me). You will be able to study Faros on every possible angles, like analyzing cycles, or other properties. The most impressive module being the one about stacks, allowing to find out (or just listing, not sure I want to release everything yet) cyclic stacks for instance. Trust me, and those who I share such things with for years could confirm, it's just outstanding. I have found things you would hardly believe.

There will be a shuffle analyzer too which, "teaming up" with the capture module, well... you should get my drift.

Again, I gave you just the thumbnail, the full specifications are not definitive yet and may change. But, at the moment, there are ~125 windows like the ones in the picture above. Not a misprint, 125! It's a crazy project :-).

And before some idiot ask, NO, it won't run on an Iphone nor won't be sold $1.99. It won't be for free either. Sorry for using the word "idiot", but the new generations are so stupid, so used to steal, share, having all for free, that I prefer to set the limits from the outset.

If enough people are interested, I'll post a few screenshots once in a while.

Thanks for your interest.
Message: Posted by: Daggers (Sep 20, 2014 12:26AM)
That sounds quite comprehensive! It seems to me like it's a program to shuffle and deal cards virtually and be able to see the position of any card at any time. Instead of doing it all physically and having to look at which cards are where. But please pardon my ignorance, what does analyzing shuffles and deals do? If you're handling cards and using faro's and whatnot, shouldn't you know which cards are where and be in control? Or is this more of a simulation for advantage play in casinos?
Message: Posted by: JFX (Sep 20, 2014 05:31AM)
[quote]If enough people are interested, I'll post a few screenshots once in a while.[/quote]
Although it's way out of my league, I'd love to see how this project evolves.
Message: Posted by: vindar (Sep 20, 2014 07:46AM)
Hi,

Indeed, this looks like a very ambitious project. I find the GUI really nice and you obviously put quite a lot of thinking in the layout of the windows (although I am wondering how you display a deck of, say, 54 cards with the jokers ? just kidding :-) ).

You say that you can compute odds with it. How do you achieve that ? Do you make a formal calculation, exhaustive search (whenever possible), or do you just use Monte-Carlo simulations ?

Anyway, this looks interesting for finding hidden properties in memorized stacks. One thing that would be very helpful is the ability to create and run simples scripts (e.g. computing all the possible ordering after a sequence cut/faro/cut/faro and look for a particular sequence automatically for instance).

Good luck with your project !
Message: Posted by: AMcD (Sep 20, 2014 09:49AM)
About the Jokers, well, it took me some thinking. Including them or not? I'm not a Magician and the games I played, well, Jokers are never used. It's kinda tricky to use Jokers in a software like that as well. If you consider Poker variants, their values, what they "mean" can change 100% from one game to another. What about games where a standard card is used as "wild"? It would have been a crazy job to do and I'm not sure it would worth the pain. There are also cases with 3 Jokers (French decks). So, after all those considerations I decided not to use Jokers. Hence the full name of that app, which is DeckStudio52 :-)

About shuffle analyzing. The point is to check how well shuffled is a deck. I've always disagreed about the famous "7 times" for instance. So I plan a module where you can track a series of genuine shuffles and you would get statistics, visual indicators, etc. We can check all kind of consecutive series, suits orders (CHDS, CHSD, CDHS, etc.), and all sort of sequences (A-3-5, T-J-Q-K, etc.). The "Capture" module would allow to capture such a series of shuffles. Of course, we could also simulate them.

Same thing with Deals analyzing. Out of a shuffled deck, who is gonna win the next round of Omaha? Where to cut in order to make #4 a winner? Etc.

About simulations, in many easy cases you can get exact values. But if you wanna know what hand is gonna beat your JJJ with 4 Hold'Em players around the table, simulations are a faster way to get a fair odd. I may rely on simulations only.

@vindar

Can you detail your last sentence please?
Message: Posted by: Daggers (Sep 20, 2014 05:07PM)
Oh wow I get it now! That's super cool! Does it have the ability to account for human errors as well? It sounds like a really powerful piece of software! I too, really like the GUI quite a bit. I don't know if you're into card counting at all, but I bet you'd be able to make one hella good piece of software for that. Plus it would be easier on the eyes, the others aren't quite as smooth looking.
Message: Posted by: Magic-Scott (Sep 21, 2014 08:32AM)
Very interesting and a bit above my pay grade, but I too look forward to seeing how this evolves...
Message: Posted by: AMcD (Sep 21, 2014 09:41AM)
For sure it's a bit tiresome to code, but it's not rocket science either. You guys are too humble.
Message: Posted by: AMcD (Sep 22, 2014 10:24PM)
So, after a few emails, looks like Jokers are important for some of you... Argh, it's a lot of work :-(.

PS: talk here guys, answering emails is a pain, I receive dozens per day.
Message: Posted by: vindar (Sep 23, 2014 06:23AM)
Hi AMcD,

When I say "running simple script" I am thinking about a way to do some intensive computation and look for remarkable arrangement in a deck. To give an example. I would like to know if there are interesting sequences obtained from a deck starting in Mnemonica order after a sequence of cut/faro/cut/faro. There are more than 2000 such ordering hence it cannot be done "by hand"... So, I was thinking about a simple language allowing to run scripts looking something like

put deck in order XXX
for every transformation [cut/faro/cut/faro]
search for [some property]
endfor

But I now realize that this is not really the purpose of your program (hum, I might program that myself when I have some free time)....

You also say you that disagree with the "7 shuffles rule" and want to do statistic to prove it. So you think there are still "usable" information hidden in a deck after 7 shuffles ? All right but will this not strongly depend on the shuffle procedure you use ? How do you do simulate a "genuine" riffle shuffle ? For instance, the classical modelization for the riffle shuffle assumes that the cut between the two packets follows a binomial law. It is a very convenient assumption from a mathematical point of view but I do not know how closely it matches the real distribution of a "real riffle shuffle". And since you are looking for very fine statistics, it seems to me that even a tiny variation in the shuffling procedure might alter the result significantly.

Anyway, good luck with your program, it looks very promising.
Message: Posted by: AMcD (Sep 23, 2014 09:03AM)
@Vindar

In fact, the goal of that program would be -exactly- to solve questions like yours :-).

About the "7 shuffles rule", my main point of disagreement is about HOW they shuffle and cut the decks in their studies. In Diaconis, etc., the way they simulate rifles, is far from the way a genuine/standard riffle looks like.

Theory is one thing, facts another one. Modern scientists have the bad tendency trying to find facts that match their theory. Me, I'm old school, I prefer to study facts and to infer the theory after :).

"If the facts don't fit the theory, change the facts." - Albert Einstein.
Message: Posted by: AMcD (Sep 24, 2014 08:29PM)
OK, despite all the work required, I may reconsider things.

- What games you know use Jokers, and how many of them?

- What games you play use less than 52 cards, or more. I know games using 32 cards or 54 (52+2 Jokers)

Please, name "famous" games or at least games spread enough. I don't need a variant of Chicago Stud using 3 wild cards and 17 rounds of betting...
Message: Posted by: AMcD (Sep 28, 2014 09:12AM)
No one is playing cards games with less than 52 cards?
Message: Posted by: Albatros (Oct 7, 2014 11:22AM)
Well, in Germany we like to play Skat, which uses 32 cards...

http://www.pagat.com/schafk/skat.html

All the best,
Sven ^^
Message: Posted by: AMcD (Oct 8, 2014 12:32AM)
It's a huge extra-work, but I have decided to add the possibility to use decks made of 26, 32, 40, 48, 52 or 54 cards (in fact any type of deck with an even number of cards). And Jokers.

When it's done I'll post a couple of screenshots. Not sure many of you are interested, but anyway.
Message: Posted by: uhrenschmied (Oct 10, 2014 09:45AM)
In Germany there is also a Doppelkopf variant that uses 24 cards. Just saying ^^

Best regards,
Message: Posted by: AMcD (Oct 10, 2014 10:18AM)
24? Geez... You have a description of that game somewhere?

Thanks.
Message: Posted by: uhrenschmied (Oct 10, 2014 01:18PM)
Actually its Schafkopf, which is quite similar to Doppelkopf.
http://en.wikipedia.org/wiki/Schafkopf

It is played with a kind of a Skat deck (32 cards)- since it is basically an Bavarian game they of course use a different set of suits *g*. For the variant the 7s and 8s are removed.


Best regards,
Message: Posted by: AMcD (Oct 10, 2014 02:06PM)
Cheers.
Message: Posted by: AMcD (Oct 13, 2014 01:55PM)
OK, the new interface. Now I can handle even decks from 16 to 54 cards, and Jokers!

[img]http://www.arnoldmcdonald.org/temp/newgui.png[/img]

I hope no one is gonna find out a game with an odd number of cards...
Message: Posted by: vindar (Oct 14, 2014 02:12PM)
[quote]On Oct 13, 2014, AMcD wrote:

I hope no one is gonna find out a game with an odd number of cards... [/quote]

... "Le pouilleux" (http://en.wikipedia.org/wiki/Old_maid_(card_game)) :-p
Message: Posted by: AMcD (Oct 14, 2014 02:33PM)
...

I quit. I'm on the next flight to planet Mars.