Home Games Classes Results Schedules Notable Articles
Contacts Directory Links History Books Map Getting Started Tournaments Conventions

Home

This afternoon there were a number of comments about the "unusual" distributions seen with the computer generated hands. My personal point of view is that extent to which you believe they were unusual is a measure of how poorly shuffled the hands have been in the past. I asked the DealMaster Pro (our card dealing program) to generate 10,000 deals (40,000 hands) and accumulate the statistics. If you click here then you will see a comparison between theoretical expectations for hand distributions and the actual ones observed in that particular set of 10,000 deals.

For those of you into a little bit more math, if you look about 3/4 down, on the right, you will see that the sum of differences between the actual and theoretical was 1.59%. To get this figure they added up all of the differences, ignoring whether the differences were positive or negative. If we increase the size of the sample by 4, a mathematician would "expect" the sum of differences to decrease by 2 (square root of N). I put expect in quotes because the process is random, but, just like you would "expect" 40,000 flips of a coin to more closely approach 50% heads than 10,000, sometimes your expectations are frustrated (it IS possible to flip 10 heads in a row, after all:). Just for fun I generated 40,000 deals to see how closely actual results follow expections. Click here to see.

The bottom line to all of the above is that the hands we get from the program turn out to conform to theoretical expections perfectly. The problem arises when you drop from 40,000 deals to 26 (or 27 or whatever). Now the variances are "expected" to be dramatically larger (over 30 times larger than those seen with 40,000 deals). Those of you that play 300 times a year for over 30 years, however, with computer generated deals, YOU will see ALL of those different variances cancel each other out:)

For those of you who played on Friday the 19th, you were the first to play a game where the hands were dealt by our own card dealing machine, a "Dealer 4". If you look at the results for 6-19, you will see a little red "HR" to the right. If you click on that, you will then see the results (which will be in a PDF file, which you can download by right clicking on it and saving to your own machine). One item of further interest is that all of the hands were analyzing by the "Deep Finesse" program and you will see the maximum contracts that can be made in each suit. "Deep Finesse" does its analyses "double dummy", which means that it gets to look at all 4 hands (so, some of the contracts it reports as makable will seem quite a stretch to those who only get to see their hand and dummy:).

There are a couple of issues with the Dealer 4....

It takes 20-30 minutes to generate the hands, feed the Dealer 4 and stuff the boards and the process works easily with 2 people, one to handle the computer and one to handle the cards. We will need to develop a group of people who would be willing to coming a half hour early, with their partner, to do the job. Don't be bashful!:)

There has been some concern expressed about those preparing the boards also playing the hands. Allay those fears, grasshoppers.....the hands are not displayed during the process. They ARE displayed later when the computer is preparing the hand record and doing the "Deep Finesse" analysis, but, this will be done while the game is going on by the director (or done after the game as was done today).

A third issue is that the Dealer 4 is VERY picky about the condition of the cards fed to it. The good news is that this means that you won't be playing with sticky, worn cards. The bad news is that we will be replacing a lot of decks on a regular basis. Part of the problem is that players can pick up oily residue when grazing at the food table. If this happens to you then it would be helpful to use the alcohol spray (on the ladies' bathroom door) or wash your hands.

It was certainly fun to see so many of you at the Ventura regional. I regret not taking my camera to record the strength of our attendance. Congratulations to all.

I'd like to thank Paul Murhpy for donating the computer table. The (intended) primary purpose of setting up the station there is to have a place to use the card dealing machine that SBBC has purchased. I don't see any reason why those of you who have bridge partners that are computer phobic or deprived couldn't use it to educate them:) (Give someone else the joy of deleting spam every day!!)

For the time being, the FAST PAIRS game on the first Friday of each month has been cancelled. Apparently we don't have enough FAST people!!

I was recently sent a reference to an article on Bridge & Aging

I just recommended an article on Barry Crane (in the eyes of many the best PAIRS player ever!) and realized that others might enjoy it as well...

Barry Crane's 11 Commandments

Hand records are posted on the results page for all of the STAC games. Those of you who didn't get a copy can download one (by clicking on the little red "HR" next to the date). They are of enormous value to partnerships to (accurately:) discuss what happened (or should have happened) at the table.

Santa Barbara SECTIONAL Flyer

There are two new games, both for 299'ers. One is at 9:30 on Mondays, replacing the 99'er game that was at 1 and a new one at 1 pm, on Thursdays. Y'all come (if you have fewer than 300 MPs:).

Feedback