Hm. Okay time to think aloud. You can skip this part and go to my answer in the next paragraph if you don't want to read my stupidity. (I'm assuming that of course if you see T fall on the first trick or see someone show out the first round, you make the obvious play.) At first I was thinking of it like the Q98 ATxxx problem or whatever. But I guess this is slightly different since in this case if you run the Q and it loses to K on right, you can just bang down AJ next if the rest of the cards split 2-1 or 1-2 (which should be likely), which gains you KTx or KT off, but now loses to Txxx on left. Okay, now let's see... if Q holds... Comparing letting 9 ride next to playing AJ next, the former lose to KTx off, but it'll gain to....KTxx on? Hm interesting. I guess in the KTxx case, you'll need to play low to J and then Ace to make sure it's only 1 loser. If Q gets covered... definitely ride the 9 next
( ... )
Comments 5
Reply
Heh. If you Google "suit combination Roudinesco" this entry is the second hit.
Reply
Reply
It at first seemed weird that low to the J isn't right, but it makes sense because you lose the 8 and the J on the same trick.
Reply
#1 Run Q then run 9 if Q was covered is 26/32: any (3-2) (20), any 1-4 exc Txxx (4), 4-1 stiff honor (2).
#2 Low to Q then cash AJ is only 23/32: any (3-2) (20), 1-4 stiff honor (2), 4-1 stiff T (1).
#3 Low to 9 then run Q is only 24/32: any T on left (16), stiff K on left (1), KT on right exc 5-0 (7)
Anyway this is close enough so that you could choose #3 if you think LHO is more likely to have the length.
Reply
Leave a comment