How many games for stats to mean something?

[BrianEDuran]

Having dinner with a friend last night who plays some poker, he claimed that after playing 10K hands you have enough deal to construct some understand of your weakness and strength. How many games of bridge to you think you need to play?

 

I'm guessing more, since with most poker hands you can bid a little, bid a medium amount or bid a lot. It seems to me in bridge there are more choice and more random things can help.

 

I did a quick study of my MP result when playing with 2 partners of 20 (480 hands) games, and got results for my weakness that I didn't believe. After 50 (1200 hands) games I got results I believed, but my STD did not go down all that much.

 

After 50 games, about 6 months, my skill level has changes a good amount, or so I hope. :)

 

So how many hands does one need to get to make statistical analysis to find weakness?

Is if possible to play that many in a short enough time frame for your skills not to change?

Should one just not care and study when they personally thing they need to improve on?

 

Thanks

Brian

[luke warm]

i personally don't have the time, inclination, or knowledge to perform a statistical analysis of my faults... so i just prefer to read and study to improve as many areas as i can... and also to play with better players, hoping they'll let me know (in private) which areas need the most work (all need work, i mean the most work :))

[paulhar]

It would seem that one would need less than 10,000 hands of bridge to identify strengths and/or weaknesses in a bridge player. After a couple of sessions, you have a pretty good idea of his bidding skills, declaring skills, wildness, etc.

[han]

I agree with Paul. In fact, if you are playing with someone much better than you, then they can probably tell you after one day (say 52 hands) very accurately what your weaknesses and strengths are. It's much harder to say these things about yourself, or about a player who is better than you are.

[hotShot]

If you play poker, it is hard to find out what a better player would have done with your cards.

 

Playing team or pair events you can compare your results with that of other players. So finding the boards where you messed up is very easy.

 

So i guess something between 30-100 boards played should be enough to get valid picture of you skills.

 

Here at BBO you can even look at movies of the other players to analyse what they did.

 

So i guess playing at BBO and analysing the boards will give you valid information about your strength and weaknesses. You can concentrate your statistical efforts on the boards you failed.

[Flame]

I dont think this is something you should solve with statistic, You can relay on your judgment and you can also ask others especially good players for it.

Also if you want to use stats then use it on book problems, 1 book problem has the value of maybe 30 real games. (although there are things you wont learn from a book)

[paulhar]

If you play poker, it is hard to find out what a better player would have done with your cards.

So true! In bridge, to give an expert a problem that you faced, he might ask for breaks in tempo in the bidding or play, or the skill level of the opponents, but this usually is a fairly minor factor.

 

In poker, it's the whole factor. The expert would not only need to know the cards dealt, but all the mannerisms of the other players, and what was true other times these mannerisms occurred (and what was true when other mannerisms that didn't occur this time occurred.)

 

I've got fair card sense and a good feel for probabilities which help me at bridge, but because I'm oblivious to table action, I'll usually be the biggest patsy at a competent poker table.

[luke warm]

but also you might be hard to read... get a pair of mirrored shades, headset with bach playing, you're all set... oh, a baseball cap, turned backwards