Many Questions on Dealing Software I have a few questions about dealing software.
#1
Posted 2023-December-11, 04:26
(1) DealmasterPro is a well-known dealing software. How to purchase for a DealMasterPro and how much will it cost? I accessed its website: https://dealmaster.com/ and found DealMasterPro+ is under construction. Do you know any info about DealmasterPro+?
(2) I'm building a software by myself. User can customize the dealing strategy as he/she want with a single command. For example:
"I don't want 7+ suit" -> "sum(EVERY_SUIT>=7)>=1, 0%"
"I don't want north to have more than 10 minors" -> "sum(NC+ND)>10, 0%"
"I want more game boards" -> "NSP>=16 && NSP<=24, 50%"
Does any dealing software have similar functions? Do you think this funciton will be useful?
--xyux
#2
Posted 2023-December-11, 04:35
BigDeal gets used by BBO, if I got it right.
There is also a established Python library under GNU (?)
https://github.com/anntzer/redeal
Uwe Gebhardt (P_Marlowe)
#3
Posted 2023-December-11, 10:11
What you get depends on what you want.
- Do you want provably random hands, using an accepted-as-safe cryptographic process to defend against "work out future hands from previous", from the entire possible set? BigDeal - like everybody has been using for years (and some for decades) (okay, maybe not BBO). Note, you can't put constraints in - that is explicitly guarded against: "treat the operator as an attacker". You could post-screen them and throw out hands you don't want, of course.
- Is that not good enough and you also need proof of non-tampering? Use the SquareDeal addon to BigDeal. Note that this is what security experts call a "huge PITA", so unless your players have a serious expectation that someone would bribe you to cook the hands for them, it's overkill.
- Do you want a dealer that will make sample hands for you based on constraints? Deal/redeal. Note that it is *not* provably random - even without constraints - and they warn you about using hands dealt in real games. Note also that "constraints" lead to boards thrown out: "deal randomly and check if it meets the requirements"; but hands can be set ("you hold...") and at least in redeal, hands can be SmartStacked (provide some serious constraints for one hand and it will "randomly" build that hand, then scatter the rest of the cards among the other three). Note still also that Deal/redeal may be using a non C-PRNG; again not a problem for analysis, but a suitably invested player could determine future hands from previous reasonably trivially.
- Pretty much all the current options output PBN files, which all dealing machines accept. Of course, if you have a dealing machine, and you want dealing software to generate hands for it, basically all of them offer that natively. Some of them even allow certain constraints (for teaching/noncompetitive use only). You might as well use the native software, unless you need any of the above.
Dealing *machines* (Duplimate, Dealer4) are wonderful tools which, if you're running multiple games a week, will pay for themselves in short order (in frustration if nothing else). However, for home use, they're "specialist tools" and economy of scale just isn't there - they run about USD 3k. (There's a "HandyDup" that is Very Slow - 60s vs 6s/hand - and almost unusable if you're left-handed - that is much cheaper (about USD 300), if you only need 10 hands or so, or you're "shuffling and dealing" at the table.)
If you're writing your own software, ensure you know what you are doing it for:
- If you're writing it to generate hands for tournaments, don't. Use BigDeal. Securely and provably randomly generating bridge hands is like any other cryptographic problem, and you "don't roll your own crypto."
- If you're going to do that anyway, just to solve the problem and learn about this, do ensure that (barring constraints work) you ignore any other way to "shuffle and deal" besides "generate 96 bits of randomness, look up the matching hand in 'The Big Book'(*), and if the 96 bits is higher than number of bridge hands, throw it out and start again". They're wrong.
- If you just want tools that will run out 100 "1NT opposite a minor slam try" hands, use deal/redeal (or BBO's constraint system, based on van Staveren's "dealer"). You'll have to learn a language (Tcl or Python), but not all that much, unless you want to do some crazy stuff afterwards (analysis; haven't needed numpy (NUMerical PYthon library)/pandas yet, but I'm sure I will). I think the basics of python are such that that's easier to get into, but I don't know your current language set.
- You'll want analysis tools - Bo Haglund's DDS (Double Dummy Solver) is available. I *strongly recommend* you don't try to redesign and reimplement that. Dealing is "simple" (as long as you're willing to admit it); DD is hard (and has lots of "oopsie" corner cases to avoid).
(*) Not literally. Re-implement the Book in software, using either Thomas Andrews' ordering or Richard Pavlicek's. Obviously :-)
#5
Posted 2023-December-12, 10:18
barmar, on 2023-December-12, 08:41, said:
I mixed it up, ..., the same author, but dealer is it.
Uwe Gebhardt (P_Marlowe)
#6
Posted 2023-December-12, 15:36
xyux, on 2023-December-11, 04:26, said:
Dealmaster Pro is/was an old program that was never fully converted to a modern Windows UI. Even so, I think a lot of bridge clubs use the software to produce deals and hand records for their games. The interface is a lot easier to use without having to learning a scripting program like some of the other deal and double dummy analysis programs that are also available, but it wasn't free.
The developer of the program was an older man who stopped development quite a while ago, who died pre covid. His successors announced several years ago that they were going to modernize and upgrade the program, and at one point, offered beta trial versions to players at one of the ACBL NABC championships from what I've read. They've been silent for years now, website notwithstanding.
#7
Posted 2023-December-12, 16:43
I've mentioned that the hands are completely random, I don't know what else to say.
I probably won't convince others that a black cat isn't unlucky either.
#8
Posted 2023-December-12, 16:48
#9
Posted 2023-December-12, 16:50
jillybean, on 2023-December-12, 16:43, said:
You could start by saying that they are on average more interesting to play.
And add that at least nobody in this room could possibly have influenced them, intentionally or otherwise.
#10
Posted 2023-December-13, 16:45
pescetom, on 2023-December-12, 16:50, said:
And add that at least nobody in this room could possibly have influenced them, intentionally or otherwise.
Tell them to buy lottery tickets instead.
While you're at it tell them there are 10,000 tickets and they can choose one of the following three numbers: 1, 2,754 or 10,000.
Let me know how many choose 1 or 10,000.
#12
Posted 2023-December-14, 15:43
pescetom, on 2023-December-14, 06:52, said:
Are you taking the ...?