[helene_t]

jdonn, on Apr 1 2008, 04:32 PM, said:

What I meant was there is a 95% chance that opening it averages a better result in the long run than passing it, given that both results are 2 standard deviations away from 50%. I wasn't saying it will work better 95% of the time. Of course if they aren't standard deviations, as you say, then never mind.

Sorry I misunderstood. Your conclusion is correct, then.

(Strictly, confidence at the 95% significance level is not the same as a 95% probability but you probably need to be a statistician to be exited about this)

[awm]

Yes, the conclusion is that opening is better than passing on these hands at MP. It seems there are enough trials (and enough difference) to be pretty sure of this. The conclusion at IMPs is the opposite, although this is not quite as certain.

Of course, there are always some issues with this kind of bridgebrowser test. In particular there is the question of whether good pairs (i.e. pairs with a lot of partnership experience and/or pairs consisting of two good players) are more likely to open or pass than random pickups. For example, I am sure that a high percentage of people who are playing a strong club system open and these pairs probably average better results (because they have more systemic agreements than average and are unlikely to be novices if they are playing strong club). So there is some inherent bias, but I suspect that the vast majority of pairs on BBO are playing something pretty standard and are not long-standing partnerships, so hopefully this effect doesn't have too much impact. I may try a few tests to measure this effect later tonight.

I can't answer Gerben's question because: (1) there were only something like 8 1NT openings on these hands so the results will be very inaccurate (2) anyone who is playing weak NT probably has a bit of an edge on the field (a high percentage of these are established partnerships, although surely a few are pickup british pairs playing acol).

[han]

Can't you restrict the search to people who's rating is 50-51 or something like that?

[neilkaz]

With no aces, this 5332 11 count is just a bit weak for me to open inspite of me loving to get the ♠ suit in quickly. Give me the 10 of ♠ and I'll open this playing SAYC or 2/1

.. neilkaz ..

[ArtK78]

Interesting discussion.

When I first saw this problem (today) I voted to open. It was almost a "wtp." Experience tells me that opening this hand is right. Jlall says his experience is the same. The bridgebrowser result confirms this in a statistically significant manner (at matchpoints).

I would not get upset with a partner who chose to pass this hand, I just don't think pass is the winning action.