[VM1973]

vuroth, on 2011-September-28, 10:41, said:

I'll grant that the sample size probably isn't one anymore, but does the method really seem sound?

What method have you used all your life?

When you tried chocolate did you look at it and think, "Wow, chocolate is pretty good" or "Wow, this is crap" or did you say, "This is just one datapoint. I cannot make an intelligent decision based on that. I'd better ask a food expert on whether or not I can like this."

Or when you had your first kiss with a girl, did you think, "This is only one datapoint. I must go out and kiss men to see if that works for me better," or did you think, "Wow! Kissing girls is pretty good!"

First time you had sex?
First time for pizza?

I suppose all of those were just single datapoints that didn't justify you drawing any conclusions about them.

Personally I think drawing a conclusion and testing it in future practice and having it never been proved wromg is a very powerful argument in favor of something. Perhaps God (or what passes for him around here) will lol at the thought, but hey!

[JLOGIC]

lol youre such a complete ***** joke dude

[pio_magic]

VM1973, on 2011-September-29, 11:07, said:

"Wow! Kissing girls is pretty good!"

Right - and I don't need to kiss all girls every time just because of this


Peter

[VM1973]

JLOGIC, on 2011-September-29, 11:14, said:

lol youre such a complete ***** joke dude

As chance would have it, I was kibbitzing a bridge game last night and this exact situation came up. It so happened that, there was no functional difference between the two contracts. However, I couldn't resist weighing in and commented, "Justin Lall, man! Don't you know you're supposed to prefer the 5-4 fit?!"

To which he responded, "How darest thou blaspheme His holy name!" I quickly saw that the rest of the tables were similarly upset and I figured I'd better get out of there stat. As I was making for the door one of them tripped me and I figured I was a goner. They dragged me outside and readied stones, but a kindly old man there rescued me. He posed the question of whether a man was holier by losing important sums of money on pointless speed bets with friends in Las Vegas while stupidly using a laptop instead of having a mouse or by underleading KJ83 when you see T4 on the board instead of making the obvious surrounding play of the jack.

I thought it was a stupid question, but it worked like a charm! The whole crowd divided up on one side or other of the issue, and the man whisked me out of there. "His religion is one of peace and tolerance and they do not truly understand His wisdom," he informed me gravely.

"That may be," I said, "but wow man! You saved my life. That was a really bad situation there that I hope never to repeat."

"Nay, my son," he said. "Thou canst base thy whole philosophy off of a sample size of one. It is written in the Holy book."

And then I realized I was in the presence of a fanatic and I thanked him again, made an excuse, and got the **** out of there.

[mikestar13]

To get back to the technical merits of this question: A 4-4 fit in some types of hands is likely to play better than a 5-3 because of the discard possibilities. This can be true, though less often, of 4-4 vs. 5-4. There is, however, another scenario: the contract is set because searching for the 4-4 when you have already found the 5-4 leads to a more revealing auction which draws a roadmap for the defense. My general rule is "when you've found a trump suit, don't look for another."

[semeai]

VM1973, on 2011-September-29, 11:07, said:

I suppose all of those were just single datapoints that didn't justify you drawing any conclusions about them.

Our prior experience tells us that certain categories of actions tend to have high variance in terms of how well they turn out and certain other categories of actions tend to have low variance in terms of how well they turn out.

VM1973 said:

Personally I think drawing a conclusion and testing it in future practice and having it never been proved wromg

I don't think most of us accepted his premise that he'd really never had it be wrong. Maybe his memory is selective, or he didn't notice when there was a better 5-4 fit sometimes.

vuroth, on 2011-September-28, 10:38, said:

4-1 splits happen 3x as often as 4-0 splits (not to mention 5-0s). Basically, the difference in likelihood is around 20%, or 1 in 5. If I'm reading that right, it means that 1 in 5 times you choose a 4-4 over a 9 card fit, you've given yourself a trump split problem you didn't previously have.

Does needing a pitch as your only hope really come up that often.

pio_magic, on 2011-September-29, 01:40, said:

Can we "measure" the likelyhood of the latter and compare to the more easily calculable split probabilities? Would really be interesting....

This would be interesting. Or maybe we could just simulate the whole thing.