IMP Scoring in Club Events
Why should you care?
Part 2: Tournaments
Many different schemes were devised for reducing the luck component of
the game and any tournament director's handbook is full of
descriptions. You will have encountered Mitchell and Howell
movements for pairs events but those occur in several variations.
Then there are teams of four and individual movements. This
article discusses Teams of Four and Mitchell competitions.
The closest we come to the "real bridge" ideal in competitive bridge is
the modern Teams of Four match. I shall describe the simplest
such event possible between just two teams. There are usually
more than two teams in any competition and there are movements
which accommodate larger numbers but the principle is the same so
there is no need for me to add complexity to this discussion.
The effect of converting a raw difference in scores to IMPs is to chop off the extremes and to give more weight to the small contracts while still preserving the substantial benefits of bidding and making game contracts. Using the example from the previous article, let us invent some scores and fill out a results sheet.
We have Team Trumpers with Clara Cardace and Jack Diamond playing the N/S hands in one room while Biddie Lightner and Trixie Short play the E/W hands in the other ...
|Bd||Vul||N-S: Clara & Jack||E-W: Biddie & Trixie||Diff||IMPs
Well I ended up using an example from the Coolum Bridge Club so you can look there to see the explanations. I expanded the example far beyond what normally appears on a scoresheet in the hope that it makes things clearer but at the risk of making them too complicated. Anyway the Trumpers won the contest by 19 to 16 IMPs.
The event that I have just described is a contest between two teams
of four players. Note that it is rare for such events to be
limited to just two teams and there are movements which cater for many
teams. In those multi-team contests the principle remains that
your opponents at any given time is just one team. That is very
different from the situation in pairs events.
The most common movement for pairs events is the Mitchell and it is the one almost invariably played at the club on Mondays. The Mitchell works well for 6 or more tables. With fewer tables the tournament director may consider running a Howell movement but I do not intend to discuss those.
Here is an example of a hand played on Monday 1st December 2014 ...
N-S has a solid game in hearts losing just two spades and a club but only one pair made it. E-W also has a geme in spades losing one trick in each side suit but nobody made that.
At matchpoint scoring you get two points for every other pair you beat and one point for every pair whose score you tie. If we look at the N-S scores then pair 3 who made the 4♥ game beat the other eight N-S pairs and so scored 16 points. The next highest scores were made by pairs 5 and 9 who were each given 200 points when the E-W pairs playing in 4♠ somehow found a way to lose 5 tricks. Those pairs each beat six other pairs for 12 points and tied with each other, thus scoring 13 altogether. The next best score was obtained by pair 7 who managed to push the opposition into the unmakeable 5♠. That score beat 5 others and yielded 10 matchpoints. The four-way tie for going one down beat just one other score gains 2 for the lower score and one for each of the other three tied pairs for a total of 5 matchpoints. Finally the pair who let E-W play in the 3♠ part score got a bottom.
We can run the same analysis on the E-W pairs to show how the matchpoints were calculated. Note that a useful check is that the sum of the N-S and E-W matchpoints is always the same.
Assigning matchpoints is actually much easier than might appear from the way I just described the process. After half an hour of tallying it becomes quite routine. The simplicity of the scoring method was a great boon When I was semi-permanent tournament director at the West Australian Bridge Association in 1968. There were no BridgeMates or equivalent, no personal computing devices, no electronic assistance whatsoever and matchpoint scoring was something which could be done quickly and easily. I have no doubt that this was a major reason why matchpoint scoring became so entrenched in club tournaments. It really is so much easier than IMP scoring. Nowadays that justification has vanished. Everyone uses a computer in some form to do the scoring.
Matchpoint scoring rewards small differences in scores; IMP scoring rewards large differences. In this aspect, IMP scoring is much closer to "real bridge" which I am going to try to include in the comparisons. You may recall that in the first article in this series that I suggested rubber bridge was best played for money just to prevent silliness. Well in the 1960s when I learned bridge we'd play for stakes like 10 cents per hundred points so going down 4 tricks reduobled and vulnerable would cost a couple of dollars. It was enough in those days of penury to keep the game honest. Even now I think a dollar a hundred is sufficient and that is the basis of my assessments when I estimate the rewards and costs in rubber bridge. The bonuses awarded in tournament play (i.e. at teams and duplicate) correspond to the rubber bridge scenario when a rubber is unfinished; if one side has won a game (vulnerable) and the other has not then the bonus is 500, otherwise it is 300. For the sake of uniformity in making comparisons I shall assume that is the same situation in the rubber bridge game.
A fairly spectacular illustration of the differences between the scoring systems is offered by board 3 played at the club on Monday 1st December 2014. (They're not always so clear-cut!)
On this occasion at every table South was the declarer in a contract of 4♠ not vulnerable. The par score was 11 tricks for 450 points but pair 1 made 12 tricks and pair 5 only made 10.
The first table is a matchpoints score sheet. The second shows the raw IMP scores for the Trumpers team calculated against a par score (datum) corresponding to a contract of 4♠ yielding 11 tricks. The third table simply shows the expected value of the hand at rubber bridge when playing at the modest stakes of $1 per 100 points.
Now we are in a position to look at what the scoring systems say about bidding and play.
Overtricks matter. That can be seen in the hand scored above where N-S pair 1 got a clear top score by making that one extra trick. Likewise pair 5 got a bottom for failing to make an overtrick even while fulfilling the contract.
Partscore hands count just as much as games and slams. At matchpoints players tend to overcall frequently and engage in agressive bidding when strength seems to be divided evenly.
Favour 3NT contracts rather than a minor suit game even though the minor game may be safer.
Because of the way that the scores are ranked, the extra 10 points for playing in NT rather than in a major can mean the difference between a top and a middle score.
Bid a grand slam if you think everyone will be bidding at the 6 level and there's a 50% chance of making it. Again, this works because you're not penalised so much for missing out on the 1430 points for a vulnerable small slam as you would be at IMP or rubber bridge scoring.
Overtricks don't matter much. The difference between making 10 and 11 tricks in a major suit game is only 1 IMP so it is more important to focus on making the contract than taking any sort of risk for an overtrick or two when the difference between making and not making can be at least 10 IMPs
Similarly, the extra points for a NT contract over a major suit is pretty worthless. Along the same lines it is probably worth playing in a minor suit game rather than NT if the suit contract seems safer.
Be more cautious about bidding risky grand slams. It is worth an extra 10 IMPs to bid the grand but the cost of going down is 14 IMPs when the opposition bids and makes a small slam.
Be more inclined to bid a somewhat risky game, especially when vulnerable. Going down in a vulnerable game when the opponents stop in a partscore costs 5 IMPs but making the contract gains 10.