My first thought was double but now I think I pass. I certainly might have less strength than I have so far shown so passing seems odd, but maybe it's right.
"The five level belongs to the opponents" is usually attributed to Ed Manfield. The same thought can apply to the three level. And we could bring in LOTT, preferably in its original form rather than the Cohen re-formulation. Suppose partner is 3=4=5=1, so we have an eight card diamond fit and they have a nine card club fit. LOTT says that the number of tricks we can make in diamonds plus the number of tricks that they can make in clubs is 8+9=17. So, if we can make 3
, they are off one in 3
. If they can make 3
, we are off one in 3
. Now the statistical justification for LOTT is that on average it comes out right, sometimes over-estimating, sometimes underestimating. Sometimes it is right on, but far from always. And it also uses double dummy analysis. And it assumes we know the total lengths of our suit and the opponent's suit. Not something you want to bet your life on, but something to consider in close calls.
A further problem is this: If partner indeed is 3=4=5=1, and with minimal values (explaining his pass over 3
) then if I double he is, I think, going to bid 3
not 3
. His thinking will be "Well, I of course would have bid 2
in response to the double but after the 3
on my right I passed with my minimum. Now that partner has doubled again, I guess he is prepared for me to bid whichever major it is that I have."
I suppose if partner has minimal values, say a 12 count, with 3=3=5=2 shape he will bid 3
over a double. Maybe we make it, maybe not. LOTT now predicts only 16 total tricks, suggesting that if we can take nine tricks then we can set them two in 3
.
I suspect we get a small plus if I pass. If I go on, either with X or with 3
, we might get a slightly larger plus, we might also get a minus.
So I pass. If we miss a game, I take the blame.