Article 7Q41 by Richard Pavlicek

About Instant Matchpoints

Instant Matchpoints

Setember 1993

Just kidding. Instant matchpoints are determined by the frequencies of actual results when the deals were played in a previous tournament. Of course, the tournament must be foreign (and preferably not recent) so the chance of anyone having played and recognizing the deals is remote.

For the last seven years I have done the analyses and prepared the matchpoint tables for the ACBL Instant Matchpoint Pairs (usually sponsored by Royal Viking Line). I receive a list of frequencies for the results on each board. To illustrate the process, say Board 1 was played 16 times with these results:

Matchpointing on a 15 top produces:

This is easy so far, but what about scores that are not listed? For example, what should a score of plus 430 receive? Obviously it should be between 11.5 and 14.5, but what exactly? Or, worse yet, how about a score of minus 100! Since minus 50 gets a zero, is there anything worse?

There are a number of ways to do this. The simplest and most elegant is to imagine a “mystery result” being added to the data, increasing the top by one. Each actual result is matchpointed as if it had tied the mystery result, and each gap is matchpointed directly as if it were the mystery result.

The following table shows this adjustment from a 15 top to a 16 top with all gaps properly matchpointed.

Note that the effect is to increase all existing matchpoints by a half. Also note that the gap score between +420 and +400 is omitted because it is impossible to score +410. In theory this gap would receive 10 matchpoints; but who cares!

The final column shows the matchpoints converted to a percentage (as if 100 were top). All percentages are rounded to the nearest whole percent.

To illustrate the methods, I have greatly reduced the amount of data. Typically, each board has hundreds of results, so the adjustment to a 100 top is actually a reduction from some huge actual top. I wrote a short computer program to do all the calculations (else I would definitely prefer the Dart Room).

Do I edit the matchpoints in any way? Yes. When two (or more) results yield the same matchpoint score, I either combine them into a single gap or adjust one (plus or minus 1) so the scores are different. I also remove ridiculous results — you know, like the guy who goes for 1400 on a partscore deal.

Lastly, when analyzing the deals, I occasionally make minor adjustments if I feel the matchpoints wrongly reflect what would happen at an American tournament. For example, suppose a laydown slam receives 95 matchpoints, yet I feel it should be reached by sound standard bidding. Clearly, this is an injustice (especially to the opponents) so I would make appropriate adjustments (including neighboring scores) to lower it to, say, 87 matchpoints.

Sometimes I would like to make drastic changes (such as lowering my preceding example to 65 matchpoints) but I resist the temptation. One must not forget that the frequencies are based on facts, and it’s not nice to fool with Mother Nature.

Instant Matchpoints

September 1997

Just what are “instant matchpoints” anyway? This article should provide some insight into this unique scoring method and the preparation behind it.

Most matchpoint events are scored by the actual results of each participant, i.e., each section (or group of sections) is matchpointed to determine the awards. An instant matchpoint event is different: The matchpoint awards for every possible result are predetermined. For example, if you bid and make 4 S on a particular board, you will look up your score in a chart to find your matchpoint award immediately.

Which is better? On a strictly comparative basis, the regular method is superior because it is based on the actual results of the participants. But instant matchpoints are fun! Besides being able to look up your score right away, these events are usually accompanied with written analyses. You can read about the deals you played and often gain tips to improve your game.

The most difficult task for an instant matchpoint game is to prepare the scoring awards. No matter how this is done, it cannot be completely fair because no one can predict the future; it can only approximate fairness. In most cases (as in the upcoming ACBL event) the scoring awards are based on actual results when identical boards were played in another event. Of course, this has to be a foreign event and preferably long past to minimize the chance of anyone replaying and remembering the boards.

How are the matchpoint awards prepared? How can you determine an award for every possible result? I will explain the steps I follow using a hypothetical example. Assume Board 1 was previously played 180 times with the following data, which has been matchpointed in routine fashion with 179 top:

The above matchpoints will not suffice on a general basis. What if a N-S pair scores +500, or -150, or any other result not shown? What do they get? Well, you could refund their entry fee and tell them to try again next year — or you could adjust the table to account for the gaps.

The method I use is to assume a “mystery score” as part of the original data, thus increasing the top by one. Every existing score is presumed to have tied the mystery score (thus adding 1/2 matchpoint to each), then every possible gap is matchpointed in normal fashion. The table then becomes:

Note how the top gap gets 180 (the new top) since it beats every real score, and the bottom gap gets zero. The gap between +420 and +400 would be worth 24 matchpoints, but there is no way for anyone to score +410 so it is eliminated from the table.

This method would produce different tops on many boards, so the awards are converted to a percentage (or 100 top) for uniformity. Fractional percentages are rounded to the nearest whole percent, producing:

Oops. The percentage for -100 rounds to zero, just like the bottom gap. This is acceptable, but it seems wrong and is esoterically displeasing to give the same award for different results. Hence I would smooth out the last three awards to become 2, 1 and 0.

As a final step, I occasionally adjust the matchpoints if I feel the results are biased by non-American systems, or if I feel our players would generally bid better. For example, say a N-S slam contract happens to earn 96 percent (giving the poor E-W victims only 4 percent) and I think the slam would be bid more often; I would use my judgment to lower the award slightly. I make these adjustments sparingly, and never by more than 10 percent.

Even with a protracted attempt to produce fair scoring awards, one cannot please everyone — so what else is new? Therefore, to eliminate potential bias, the ACBL goes one step further: The N-S and E-W fields are scored independently to produce two separate winners. Thus, if a board turns out to be biased, it will affect everyone in your field equally.

So mark September 17 on your calendar! Get out and play! I can’t give you any previews of the deals (ha), but I’m sure you’ll have a great time.