[kgr]

Scoring: MP

---

p-(p)-1S-(p)
1NT!-(p)-2C!-(p)-
2D-(p)-2S-AP
 
1NT=Not forcing, but can include 3cS and 4-7 pts
2C=strong, asking
2D=weak, nothing to say
 
LHO leads ♥3 (3th or 5th) and I played ♥A and took the ♠-finesse.
According to analysis in Jack it is better to play ♠A iso of finesse in trick 2.
♠A=117.8
♠J=89.2
Can someone explain why? Is playing from the top always better if you only have one entry?

[Fluffy]

I am not sure what your numbers mean, but I believe they might have something to do with the fact that spade fienese risks a diamond ruff, and even more, it risks not scoring ever the 4th diamond if the defence forces you constantly even with trumps 3-2.

[bluecalm]

Quote

According to analysis in Jack

Interesting, can you say more about how such analysis works ? I may consider buying the program

Quote

Can someone explain why? Is playing from the top always better if you only have one entry?

Finessing is better considering ♠ alone for sure.

Relevant layouts are:

Qxx to xx (6 cases)
xxx to Qx (4 cases)
xxxx to Q (1case)
Qxxx to x (4cases)
QT9xx to void (1case)

In remaining 16 (all 2-3 breaks, 1-4 breaks and 0-5 breaks) it doesn't matter)

You gain a trick by finessing in 11 of those while you gain a tricks by playing on top in only 5 of those.

[gwnn]

kgr's numbers are the average score you get on the board. If you make 140 50% of the time and 110 50% of the time, that would be 125.

[bluecalm]

What are the constraints you give for defender's hands ?
Is it possible to define that they don't have 7 card major for example ? And that they would lead stiff ♦ if they had it?

[pooltuna]

[snip]
Well if Jack is right(the program not the card) then can someone please explain how I screwed up the chart that follows

LHO RHO
hold hold Play J tr lost Play A tr lost Which wins
QT954 - 2 3 J wins
T954 Q 2 1 A wins
Q954 T 1 2 J wins
QT54 9 1 2 J wins
QT94 5 1 2 J wins
QT95 4 1 2 J wins
954 QT 1 0 A wins
T54 Q9 1 0 A wins
T94 Q5 1 0 A wins
T95 Q4 1 0 A wins
Q54 T9 0 1 J wins
Q94 T5 0 1 J wins
Q95 T4 0 1 J wins
QT4 95 0 1 J wins
QT5 94 0 1 J wins
QT9 54 0 1 J wins
QT 954 0 0 No difference
Q9 T54 0 0 No difference
Q5 T94 0 0 No difference
Q4 T95 0 0 No difference
T9 Q54 1 1 No difference
T5 Q94 1 1 No difference
T4 Q95 1 1 No difference
95 QT4 1 1 No difference
94 QT5 1 1 No difference
54 QT9 1 1 No difference
Q T954 1 1 No difference
T Q954 2 2 No difference
9 QT54 2 2 No difference
5 QT94 2 2 No difference
4 QT95 2 2 No difference
- QT954 3 3 No difference
- - 30 36

[gwnn]

pooltuna the spade suit in vacuum is best played small to jack but there are other factors which make the Ace better (see Fluffy's post). Jack was talking about the whole hand and the contract, not just the spade suit

[ONEferBRID]

Apriori, West holds Q x x 30% of the time whereas East holds Q x 20 %......
hence "eight-ever" .

[ceeb]

I don't get it either.

First, it's MP not IMP. So spurning the % play in trumps is investing some fraction of a trick, even after discounting because of the possible ♦ ruff.

Being forced out of trumps doesn't seem to me a legitimate concern. If trumps are 3-2 or the ♦J comes down, you can't be tapped out. If trumps are 4-1 and the ♦J is a loser, then you have 6 losers and are down anyway -- except if it's ♠Qxxx onside, then only NOT finessing allows those bad things to happen.

[kgr]

- I'm not sure if I can enter in Jack that it is MP's, so yes the calculation is done in IMP's.
- In Jack you can after every card do double dummy analysis (ctrl-D) or analyze the position (ctrl-A).
- Analyze the position will deal 1000 possible hands and do a double dummy analysis on them. I'm not sure what it takes into account for the possible hand, but it somehow looks at the bidding and play. (Eg if I lead from KJ52 the 2 or the 5 then that will give a different result; because Jack will think I have 3 or 4 cards in that suit). The analyze position will show and adapt the result after every hand until it has done 1000 and you can interrupt before if it is already clear. It is amazingly fast.

[Fluffy]

with only 1000 hands analyced I'd say you got a set that favours ♠A because the difference seems very huge to me.

[Free]

I think the difference between going down and making the contract = or +1 is a big factor here. If you go down it's -100, if you make +110 or +140. So your average goes down a lot for rare situations like ♠Q stiff behind or a possible ♦ ruff. You can lose a ♠ pretty safely and still make your contract, but not if you lose it while it's not necessary. Oh btw, you can even go down a lot more than 1 if there are multiple ♦ ruffs!

Can you try this NV please?

[kgr]

Fluffy, on May 31 2010, 10:44 AM, said:

with only 1000 hands analyced I'd say you got a set that favours ♠A because the difference seems very huge to me.

Most of the time when I do this the result will remain the same if I do it again. 1000 seems to be a big enough sample.
A bit more samples:
SK: 115.2 SJ:85.9
SK: 116.6 SJ:88.3
SK: 117.4 SJ:87.5
SK: 117.6 SJ:86.4
Remark: If RHO plays on trick 1 the ♥7 iso ♥2 then the figures change a bit (low is encouraging):
SK: 111.3 SJ:84.5
SK: 111.3 SJ:87.8
SK: 113.0 SJ:85.9
(BTW: LHO had ♠Q singleton)

[kgr]

Free, on May 31 2010, 11:03 AM, said:

Can you try this NV please?

NV (probably closer to MP's)
SK: 118.3 SJ:101.1
SK: 119.6 SJ:103.8
=> It could well be that SJ is the best play at MP's.

[Free]

kgr, on May 31 2010, 10:09 AM, said:

Free, on May 31 2010, 11:03 AM, said:

Can you try this NV please?

NV (probably closer to MP's)
SK: 118.3 SJ:101.1
SK: 119.6 SJ:103.8
=> It could well be that SJ is the best play at MP's.

Since the average score for ♠J goes up and the score for ♠K remains pretty much the same (slight increase), I suspect that the frequency of going down has a considerable influence on the average score.

If the J wins you get a good MP score
If the J loses you get an average or poor MP score
Bluntly speaking you might conclude that the J will be a good MP score around 50% of the time, will be average 25% and will be poor 25%. This means it's a good MP choice to play the J (62.5%).
At imps, if you go down your average sinks to the ground, at MP it doesn't. The fact that Jack doesn't count in MP but in points explains the difference in play.

[ceeb]

Free, on May 31 2010, 11:03 AM, said:

At imps, if you go down your average sinks to the ground, at MP it doesn't.  The fact that Jack doesn't count in MP but in points explains the difference in play.

This may explain Jack's thinking, but I don't think it is correct analysis.

Finessing loses the contract only if it loses, the opponents then get a diamond ruff, and they still score two diamond tricks. Unlikely the opening leader (East) has a singleton diamond, or will manage to shift from ♦AJxxx. Hence the danger is substantially the chance that West can win a ♣ shift and get a ♦ ruff. That's 4-5% assuming the imaginative ♣ shift from East.

Finessing gains the contract when ♠Qxxx(x) is onside and there are two natural ♦ losers. Counting only ♠Qxxx, that's about 7.5%.

Hence even if overtricks were negligible compared to the big fat partscore bonus I'd finesse as the best chance to make 2♠. But really, even the overtricks from finessing may well overwhelm the small chance of a ♦ ruff defeat.

[Fluffy]

hey ceeb, I said this earlier, but I'll remind you, if you have 2 diamonds losers, spades are 3-2 and ♥K is offside 5th, you get tapped and you do not ever make your 4th diamond. losing 1♠, 2♣ 2♦ and the last trick in hearts.

[ceeb]

Fluffy, on May 31 2010, 10:18 AM, said:

hey ceeb, I said this earlier, but I'll remind you, if you have 2 diamonds losers, spades are 3-2 and ♥K is offside 5th, you get tapped and you do not ever make your 4th diamond. losing 1♠, 2♣ 2♦ and the last trick in hearts.

How so? If trumps are 3-2, we're 3 longer in trumps than the opponent. If we let them in 3 times with ♠Q, ♦A, ♦J and get tapped each time, we'll have equal length (all trumps gone assuming we pulled trumps when able) and we'll be on lead. At that point we cash the long ♦.

Hence the only cost from being tapped out is a possible extra down-trick when ♠ are 4-1.

(I did see your first message in which you mentioned generally the danger of being tapped and I acknowledged that claim in my initial analysis.)