Quote Originally Posted by DSchles View Post
Would have been nice, and a very valuable study, I believe, but I don't think it will happen, because Eric isn't interested.

Don
Sure I am! I should maybe clarify what I am/am not interested in: I think it would be useful to provide an upper bound on achievable performance, and I think we have a *limited* capability to compute such bounds, and I'd like to help with said computation. Those limitations are (1) support only for rules prohibiting resplits, and (2) the expense of a lot of CPU hours. The former is due to us simply not having more sophisticated algorithms, and the latter I'm willing to eat. My only "lack of interest" is in any weird constraint on posting results of said computation. I'll acknowledge that this forum can seem like a train wreck at times, but it *is* possible to read the wheat surrounded by all the chaff; I have repeatedly found very useful additional insights (or bugs, etc.) from the comments, even from the questions, raised in this forum.

(I should also re-emphasize, as I have in the past, that I'm *not* interested in disclosure of any truly private information, e.g., the detail's of Tarzan's count, which I don't have access to anyway. I look at it this way: any of the results that *I* have posted in the past, are results that, at least in principle, *anyone* else was free to generate for themselves and post/publish/whatever, since all of the source code I have ever used in such analyses is, and has been, publicly available.)

In short, if there is an established list of setups (number of decks, rules confined to no resplits, penetration, etc.) that Gronbog has analyzed or plans to analyze, that I can help do the corresponding optimal play calculations for, let me know.

Eric