Stefan_O

Hi Tim, On page 3 in the pdf, it says I can probably estimate the tricks in my own hand... :) But how to estimate the combined tricks with partners hand? Should we just assume pard has 0 tricks, and place the contract? Or what is the idea here?

I would think 99.99% or so of all pairs, generally dont bother or have methods to find it :) Some 25+ years ago, I played a relay-system with a guy very smart in bidding-design, where a strong-enough hand (both as opener and responder) could always initiate a relay-sequence to ask for partners exact honour-strength (A/K/Q's) and suit-pattern. Also, a dominating principle was to always try to place the stronger (often totally unknown) hand as declarer. It was fun and we had excellent results with it for our level. But after all the crackdowns and suppression against "highly unusual methods" over the years internationally, today it's virtually impossible to even practice such systems in real, duplicate-bridge, which is obviously a big reason why so few play such methods.

Out of curiosity, another idea/suggestion... :) If you use double-dummy-analysis results on the deals, instead of actual play results, would your eval-system come out significantly different, or would it be mostly the same?

Once you have all deals+scoreboards in digital form, you can let the computer scan the deals. For each pair of hands, check the popular contracts how many tricks the majority of pairs make, and compare to how many tricks your eval-method predicts for the hands.... Then run the same comparison on Milton/LTC/Zar/whatever..., and see which is more accurate in the long run...

Umm.... "Zar points"... "major system"... That one never really took off, did it...? :)

@M1cha ummm... I see... yes, makes sense, then :)

By the way, where do you find all those 400k tournament recorded deals? Are they freely available somewhere?

Just wondering also on page 3, under "Trump model" => "Number of cards in trump suit" table... If the pair has, say, 4-4 in trumps, I understand from the table that each player should add 3 points (total 6 points) for the trump length -- is that correct? Assuming 3 points = 1 trick, this amounts to at least 2 extra tricks. (In addition, you also add more extra points later for any shortnesses.) But if you have 4-4 trumps, I think you mostly only gain 1 trick compared to NT-play? (3 rounds to draw trumps + 1 ruff in each hand) So, do I understand your table correctly?

Hi Tim, Minor detail: On page 4 in the pdf, the example says: That looks like a typo? AKQT6 is "3 out of 4" and "4 out of 5" top cards -- but not "4 out of 4".

Aah, got it -- yes, much clearer :) Thanks!

Were you comparing you evaluation-method to double-dummy-analysis par-contract? Or using records from actual tournaments played? Or how did you do this analysis?

Hi tnevolin, Interesting initiative! :) Some examples would be helpful, though... Under NT-evaluation, holding: KTxxx, I get: 3 for the K, +1 for KT, +2 for 5-card suit => Total = 6 points. KTx, I get: 3 for the K, +1 for KT, +1 for 3-card suit => Total = 5 points. Since you still only require the usual 25 points for 3NT, this seems too optimistic(?) so I guess I misread your guidelines?

OK. But how do you know this? Have you been involved with the Gib source-code?

How can you even tell, whether a robot-bid was based on simulation or not?

1. Yes, hard to find an example in this specific situation where Blackwood can lose you points, but that's just not how Gib works (i.e it does not exhaust all the possible actions and checks "what cannot cost?"). From the description, it seems there is a rule in the database, suggesting: with 16+hcp bid 7NT. On top of that, it probably makes a simulation, and sees that (double-dummy) the contract makes in, say, 97% of all cases (assuming South has 20+hcp), which is considered good enough to just place the final bid. Something like that, I'd guess, is how the algorithm works. 2. Additionally, I do not think in Gibs system 2NT-4NT is Blackwood -- when there is no potential trump-suit, Gib usually (always?) interprets 4NT as quantitative. So it would then need to start with Stayman or something else, which gives opps options to Double for lead or bid a suit (which theoretically CAN cost something).

Mmm, yeah, forgot abt that, sorry... But looking closely, it looks like the 4♠ description somehow comes out of the bidding-database, because the 4♠ bid has a different description than the previous North bid (4+♣ vs 3+♣, 14+HCP vs 11+HCP, etc). But "14+HCP" is also a blatant lie, since East evidently has only 11HCP. Hard to understand what is actually going on inside the robot-brain... :) If a bid is based on simulation, it would be better if the description could say so (or just "natural/competitive" or similar), rather than attaching an incorrect description. It's quite common that Gib does not have what the description says (for example "4+♥" when it has only 3♥, etc)

AFAIU, Gib does not even have any concepts of "offensive" or "defensive" values -- it just searches a (huge) set of rules based on HCP, total-points and distribution and picks whatever bid they dictate. If you examine the scary structure of gib's bidding database, http://orig.gibware.com/bidding/ you will realize that fixing defects in this area must be extremely hard, particularly in exotic sequences like your example here. After seeing this, I no longer think we will ever see much of improvement in gib's general bidding, unless the whole monstrosity is rewritten from scratch in a much more readable and maintanable way.

Yeah, this is a good example how broken Gib really is when bidding moves "out of book" :D Why does East suddenly throw 4S into the fire, after E-W passing the previous round?

I guess this is the answer: "Mixed Raise and Fit Showing Jumps - A call which has both constructive and preemptive properties." http://www.bridgehands.com/M/Mixed_Raise.htm

OK, got it :) Re monopoly, though, I think it would only help computer-bridge development as well as the bridge-community at large, if some other provider wanted to step up and offer a better/stronger robot-pgm for online competitions (incl ACBL points). It's getting very obvious that BBO is struggling hard with the maintenance and improvements of Gib, and to me seems unlikely after all these years, that it will ever really move somewhere...

Just wondering -- why are you happy abt it? I much prefer robot-bridge over random pickup pds, but what value does the ACBL blessing contribute? Particularly, if they are granting BBO a monopoly...?

why is it called "mixed?" mix of what?

Playing 2NT as part of the transfer structure (rather than natural invite) has a very distinct drawback, in that you then need to bid 2♣ stayman with a balanced invite, even if you have no 4 card major. This often causes the opener to reveal information that is useful only to opps. Additionally, you allow them to make a lead-double of 2♣ or overcall at the 2 level over 2♣. I strongly advice against that approach. Much more efficient to play 2NT as natural invite.

Interesting. What are the criteria that guide your feeling?

Which suit do you generally open, when you have opening strength with 5-card major + 6-card minor? The major, or the minor? Does it matter how strong the hand is? Does it matter if MP or IMP? Does it matter in which seat you open? I'd be interested to have some input/discussion around this.... EDIT: And -- by the way -- what about 5 spades + 6 hearts?