s***@gmail.com
2017-05-15 02:38:28 UTC
For no particular reason I "data-mined" 40,000 hand
records (hands 77700000 - 77739999) at playOK.com.
(There are millions more of these records at the website;
I don't know if I'll keep downloading them: it took
several hours just to download this many.)
Since these are from duplicate "tournaments," there are
only 1948 unique shuffles in this dataset.
Only 46.5% of the hands were balanced, compared with 47.6%
expected, but this is within statistical bounds, so the
shuffling appears fair.
PlayOK bidders often choose Gerber Four Clubs as their
opening bid!
Two bidders opened 4C in 1st seat with A - AQ62 AKJ98542
(Both went down in slam.)
Two bidders opened 4C in 1st seat with - AKJ9843 AT743 A
(Both stopped in heart game and took 11 tricks.)
Two bidders opened 4C in 2nd seat with J3 K AQT987632 4
(Both played 6D, one making 5, one 7.)
Two bidders opened 4C in 3rd seat with AKJ53 AK AK AQ83
(Partner passed 4C at both tables.)
Et cetera.
Having gone to all this bother, I'm open to suggestions
to anything of interest that might be gleaned from this
or a larger dataset. Suggestions to the ng please, as I
don't check my e-mail.
records (hands 77700000 - 77739999) at playOK.com.
(There are millions more of these records at the website;
I don't know if I'll keep downloading them: it took
several hours just to download this many.)
Since these are from duplicate "tournaments," there are
only 1948 unique shuffles in this dataset.
Only 46.5% of the hands were balanced, compared with 47.6%
expected, but this is within statistical bounds, so the
shuffling appears fair.
PlayOK bidders often choose Gerber Four Clubs as their
opening bid!
Two bidders opened 4C in 1st seat with A - AQ62 AKJ98542
(Both went down in slam.)
Two bidders opened 4C in 1st seat with - AKJ9843 AT743 A
(Both stopped in heart game and took 11 tricks.)
Two bidders opened 4C in 2nd seat with J3 K AQT987632 4
(Both played 6D, one making 5, one 7.)
Two bidders opened 4C in 3rd seat with AKJ53 AK AK AQ83
(Partner passed 4C at both tables.)
Et cetera.
Having gone to all this bother, I'm open to suggestions
to anything of interest that might be gleaned from this
or a larger dataset. Suggestions to the ng please, as I
don't check my e-mail.