SOMBA
SOMBA
Release 2.19n
Call the Director

Issue #5.  

1♣  Open "Could be Short" Announcement.  

If 1♣  opener could have <= 2♣, and their convention card says they can have <= 2♣, you MUST announce the 1♣  "Could be Short".  (e.g. 5♠, 3♥, 3♦, 2♣ )

Look under Misc. SOMBA Info/Director's Tips for the latest complete list of Tips

Detecting Cheating in Bridge

by Nicolas Hammond

Review by Marty Hirschman      (from MBA's Table Talk)

    When we last met Nicolas Hammond a few years ago, he was embroiled in a dispute with ACBL over whether his software company had performed properly under a contract with the league to update the ACBLScore program used to keep score at our club games and tournaments.
    ACBL was a little vague about what the problem was.   Hammond says his new program, ACBLScore+, worked just fine, but the league was worried that it would not own the copyright on the source code for the program and decided not to use it.   Eventually ACBL paid Hammond’s company $1.5 million for the work it did, Hammond reports.
    Now he has taken his software, renamed BridgeScore+, and enhanced it in an effort to use statistical methods to determine how much cheating is going on at bridge tournaments and, to a limited extent, who the cheating pairs are.
    It is all explained in his fascinating new book, Detecting Cheating at     Bridge.
    Until now, cheaters generally could be caught only when their cheating methods were uncovered.   Hammond’s approach is different.   Basically, he compares how well players, and particularly pairs of top players, perform compared to best double-dummy play.   Over a large enough set of results, the theory goes, a pair can only do so well without cheating.   Particularly on defense.
    The last paragraph is a big oversimplification, but it captures the basic idea.   Hammond’s approach does not require knowing how the cheating is done.   That may or may not be determined later by viewing videos of hands played by suspect pairs in major tournaments, particularly international tournaments, over the past five years.
    I cannot begin to assess the validity of Hammond’s approach, partly because I don’t have the technical background and partly because he has withheld details of some of his methods.   I can, however, report some of his more startling findings and statements:
♦ His software would have flagged all four of the major international pairs that were caught cheating in 2015, plus two pairs that were caught in 2005 and 2013.
♦There were at least four other pairs, perhaps as many as eight, who were cheating in the 2014 European Bridge Team Championships.   They have not been caught.
♦There was probably a significant amount of cheating at the Bermuda Bowl in 2009, 2011 and 2013, even excluding the pairs that have since been caught.   Also at the Bermuda Bowl in 1955 to 1983.
♦Cheating in top-level bridge declined after 2015, probably because pairs were afraid of being caught, and some suspect pairs have stopped playing together.
♦At least 30 pairs among the 300 pairs who play the most in ACBL tournaments are much better on defense than declarer play, an indication of unethical behavior if not outand-out collusive cheating.
♦There are over 130 active pairs in ACBL who are acting unethically by playing undisclosed partnership agreements designed to allow the better player to declare more of the hands.
♦There is no evidence that there is cheating in computer bridge.
♦There is a 90 percent chance that as of 2016 some pair or pairs “cracked” the hand-generation program of the ACBL and thereby had access to hand records in advance of playing.
♦Code-cracking experts working with Hammond easily developed software that could take the 52-card layout on the first three deals of a set of hands and generate the rest of the hand records. (ACBL changed its hand-generation program after this was reported.)
    The book is on sale for $39.95 at www.detectingcheatinginbridge.com

Postscript
    Shortly after the Summer NABC in Las Vegas, Hammond posted the following provocative remarks on Facebook:
“Back in Atlanta from Las Vegas. I hope everyone took lots of pictures of their favorite players.   Some of these players will not be there at the next NABC…
    “A bold statement.... some top pair(s) will not be there at next NABC or major event... but you read it here first. Give me 1-12 weeks…
    “Just a question of which pair I decide to work on first.   The book shows there are several.   I'll probably start with the ones that avoided eye contact with me all last week... doesn't narrow it down much …”  12/1

    Assistant Techie

MBA-SOMBA Need Assistant Techie

Bill Landrum, who maintains the website for the MBA-SOMBA Directory, is hoping to train someone so that he isn’t the only one who knows how the software works.   If you have technical ability with websites and programming and are interested, contact:
Shelley Boschan, 248-225-1700, sehb23@gmail.com or
Satish Shah, 586-932-1817 or 586-944-6708, satishshah280@gmail.com

Recent Updates
Call the Director
11th Jan 2020 11:27 EST
Director's Tips
11th Jan 2020 11:26 EST
Home Page
7th Jan 2020 12:06 EST
Board
19th Dec 2019 18:10 EST
0 0 0 0 0 0
Pages viewed in 2020
Director's Tips

DIRECTOR'S TIPS

1.   Why am I sometimes an "A', "B" or a "C"?   There are two reasons that the strata (A/B/C) you are assigned may be different from week to week.

(A) The strata limits for all multi-site games: STaC, Inter-Club Championships, NAP, etc. are set by the rules for that type of game. The limits generally pertain to the average points for the pair. The start limits are usually: A 2500+, B 750-2500 and C <= 750.

(B) For other type of games, to maximize the number of people that receive points, we need to have have the number of pairs in each strata to be equally divided. When you sign-up for the game, we try to divide the players as equally as possible among the strata. Since we never know who will actually show up for a game, sometimes, we fail. We then use a function in ACBLScore to manually adjust the strata so that if we have 12 Tables, we end up with 4 A's, 4 B's and 4 C's divided as evenly as possible. It requires a minimum of 2.5 C tables to ensure masterpoints are awarded to C players. We also strive to have the best players divided evenly as possible between NS and E-W This only applies to Mitchell Movement games where the N-S pairs compete against each other and the E-W pairs compete against each other. This action doe not affect overalls because it is based on ranking pairs from top to bottom based on their percentage score.

2.   How to play a card.

When you play a card, lay it on the table vertically in front of you and take your hand away. When the person that won the trick turns their card over, this signals that you should turn your card over. If you are not done analyzing the trick or planning your next move, ask that everyone turn their card back face up until you are ready to proceed. If you have turned your card over, you may no longer request the trick to be re-faced. If you develop this habit, everyone will have the opportunity see all played cards and have the time they need to determine their next action.

3.  When is the correct time to plan your defense?

Your partner makes an opening lead, and Declarer calls for a a card to be played. Is this the correct time to stop and determine your defensive strategy? No!

You should play in tempo. If you stop to ponder, your action will be interpreted as you are trying to decide which card to play in the suit, thus giving unauthorized information. The correct way to play, is to play your card in tempo, and then ask everyone to leave their card face-up on the table while you plan your defense, regardless if you won the trick or not. 

4.  When has a card been played from dummy?


Law 45. B. Declarer plays a card from dummy by naming the card, after which dummy picks up the card and faces it on the table.

In playing from dummy’s hand declarer may, if necessary, pick up the desired card himself.

5.  1♣  Open "Could be Short" Announcement.  

If 1♣  opener could have <= 2♣, and their convention card says they can have <= 2♣, you MUST announce the 1♣  "Could be Short".  (e.g. 5♠, 3♥, 3♦, 2♣ )