[gwnn]

7
8
5
8
7
5

[whereagles]

Siegmund, on 2010-October-23, 13:35, said:

No, it doesn't. You misunderstand PCA.

The PCA identified 3 axes along which to classify people, rather than the original six.

The most important classifies people on a spectrum of "complicated and aggressive bidding, and heavy reliance on logic during the cardplay" to "simple and conservative bidding, and heavy reliance on carding agreements during the cardplay".

Secondarily, it classifies people on a spectrum of "serious about the game vs. serious about the postmortem," and asserts that your seriousness about the game is very nearly independent of where you land on the complexity scale.

Thirdly, it classifies people according to their cardplay agreements, separating out the factors which were confounded in the first-level classification

If you also tell us what the mean scores for each of the six categories are, we can get from our raw scores to our factor scores, and find out what our tendencies are in regard to each of the above. (And yes, we do have a very small sample, especially since many of the respondents have similar attitudes to the game.)

Thanks a lot. The interpretation wasn't very clear from the book I read. But let me ask you one thing: I didn't quite get from your statement that axis 3 is independent of axis 1. Is it or is it not? I would say yes, since the whole point of the rotation is to make things independent, right?

I have the mean scores, but let's wait for more data to come in.

[mfa1010]

CSGibson, on 2010-October-23, 15:36, said:

Here, a 3. I try to win every time out, and in every auction, but I tend to want to win on skill and judgment, not intimidation or through lawyering, especially at club events.

Being competitive is far from synonymous with trying to win through lawyering, intimidation or some other angleshooting.

[hrothgar]

whereagles, on 2010-October-24, 03:09, said:

Thanks a lot. The interpretation wasn't very clear from the book I read. But let me ask you one thing: I didn't quite get from your statement that axis 3 is independent of axis 1. Is it or is it not? I would say yes, since the whole point of the rotation is to make things independent, right?

I have the mean scores, but let's wait for more data to come in.

Principal Component Analysis is best understood as a combination of Centering, Scaling, and Rotation.

You start by centering and scaling your data.
Next, you rotate the data such that the dimension with the greatest degree of variance is your first loading.

Your second loading is the the dimension which is

1. Orthogonal to the first dimension
2. Explains the next greatest degree of variance

Repeat as necessary

Once you're performed your PCA you need to (try to) map your loadings onto something intelligible. There isn't necessarily any relationship between your loadings and your original variables. There are some good examples out there that show PCA applied to image processing with multi-spectral images which might help develop your intuition. I recall one example where they were analyzing a satellite image of a town up in the mountains. The original dimensions where the different wavelengths that the satellite was scanning.

The first loading turned out to be a measure of how reflective different backgrounds where.
The second loading was a combination of a couple infrared spectra...

For what you're trying to do classical multidimensional scaling is probably a better approach. However, in either case you don't have nearly enough data.

[jdeegan]

382745
Thanks for presenting us with a very nice self-evaluation device. It raises some nice issues that are important for all bridge players, but not often discussed. Some random thoughts:

1. COMPLICATION - make sure partner is with you.

2. BIDDING AGGRESSION - a grand slam with 13 tricks but off two aces but only one of them cashes is 50% unless the bidding gives the right lead away.

3. CARDING PHILOSOPHY - good declarers can read your signals too.

4. LEAD AGGRESSION - listen to the bidding

5. COMPETITIVENESS - bridge is a sport without any physical activity to blow off steam, so it is easy to get over-amped on your own adrenaline. Try not to press too hard. Also, you will do better if you keep focused on the object of the game (to score points) and play it one hand at a time.

6. POST-MORTEMS - it's how you learn the game, but be very careful who you learn from

[awm]

I'd hope that all reasonably good players would listen to the bidding before selecting an opening lead. However, I think there are a lot of auctions which do not strongly call for any particular lead; perhaps 1NT-3NT is the most frequent (sure, there is a slight preference for a major suit lead). On this sequence, aggressive leaders will usually lead 4th best from longest/strongest suit, whereas passive leaders will prefer to lead from two or three small (especially in a major). Of course, one can come up with examples where everyone leads from weakness (like a hand with very little values where you basically have to try to hit partner) or from strength (a sequence in a major pretty much screams "lead me" on this auction) but there are plenty of rather "typical" hands where decisions can be made.

[PrecisionL]

Interesting: 8 7 6 2 8 8 (after the game)

2: usually passive unless hints from the bidding for aggressive lead.

Partner should be willing to learn and try new things & not an unusually slow player

Willing to play with most Precision partners.

ENTF (Myers Briggs)

[daveharty]

awm, on 2010-October-24, 12:09, said:

I'd hope that all reasonably good players would listen to the bidding before selecting an opening lead. However, I think there are a lot of auctions which do not strongly call for any particular lead; perhaps 1NT-3NT is the most frequent (sure, there is a slight preference for a major suit lead). On this sequence, aggressive leaders will usually lead 4th best from longest/strongest suit, whereas passive leaders will prefer to lead from two or three small (especially in a major). Of course, one can come up with examples where everyone leads from weakness (like a hand with very little values where you basically have to try to hit partner) or from strength (a sequence in a major pretty much screams "lead me" on this auction) but there are plenty of rather "typical" hands where decisions can be made.

Agree with this, and this is what I was getting at with the inclusion of Lead Aggression: it's on those hands when the auction doesn't strongly dictate a particular lead that this really manifests. Although, it might also include those situations when the auction DOES call for a particular lead (say, a trump) and the leader can't (or won't) bring them self to do it, because of a particular holding in the suit (honor doubleton for example).

Really interesting discussion about the data analysis, thanks guys!

[mike777]

too hard,


I want part to trust me so......how can i show that?

---------




everyone tells me how I can bid on crap

[Cascade]

857727

[gwnn]

I think you're doing an injustice to loose bidders when you're saying "for the sake of causing problems to the opponents". Sure, it is one of the nice things about overcalling often, but it's far from clear whether this is the number one reason. Of course when our opponents bid like this we tend to regard them as pesky or what not but it does not immediately follow that they are being pesky for the sake of being pesky.

[manudude03]

684674 for me, though depends on the environment I'm in.

[wank]

228222

[daveharty]

gwnn, on 2010-October-25, 03:28, said:

I think you're doing an injustice to loose bidders when you're saying "for the sake of causing problems to the opponents". Sure, it is one of the nice things about overcalling often, but it's far from clear whether this is the number one reason. Of course when our opponents bid like this we tend to regard them as pesky or what not but it does not immediately follow that they are being pesky for the sake of being pesky.

I think "doing an injustice" might be a slight overbid, but you're right that I shouldn't ascribe my own behavioral motives to others with a similarly (or more) aggressive bidding style. For me, the value of the exercise is not necessarily in finding someone with the same motivations as myself; the values would be in finding someone whose behavior aligned closely with my own, or at least differed in identifiable and predictable ways. I think you would agree that strong and enduring partnerships can be formed between people that don't agree on a lot of things.

[zenko]

Asking players to rate themselves is bound to give you inaccurate results, what I consider "complex" someone else might think of as simple and vice versa. Still, to try to assess own style is somewhat useful practice, and the whole thing would work much better accompanied with example hands, sort of like when you talk with your financial advisor and he gives you risk tolerance quiz. Done right, that kind of tool would be highly useful for bridge coaches and advance class teachers. BTW bridge in general is lacking sport-science related research materials/tools. If bridge is any other sport there would be already several analytical studies available on, for example, reasons for current Levin-Weinstein dominance in top-level pairs events.

[modron]

It would appear that it depends very much on where you are on "The Spectrum" how much you need this and where you see yourself.

[inquiry]

875359

Postmortems only well after event is over.

[mrt2000]

I've scored myself as 777587 on the understanding that the postmortem question applies to discussions away from the table once a session is over!

[TheoKole]

Im a 775678.

Postmortems always should wait until after a session is over, in my opinion.

[Trinidad]

767681

If the postmortem takes places more then 48 hours after the game I am a 767688. I also think favorable about postmortems during dinner. I dislike postmortems between rounds and I hate postmortems between boards.

Rik