[jimger]

Why use cross IMP scoring at all in the 8 board 30 min tourneys?

From what I read, it gives an unfair advantage to players who do not complete the 8 boards in the alloted time.

Just use the standard IMP scoring in these type tourneys.

[barmar]

jimger, on 2013-October-12, 06:08, said:

Just use the standard IMP scoring in these type tourneys.

What is the "standard IMP Scoring" for IMP Pair games? AFAIK, the only choice is between Cross-IMP and Butler (IMPing against a datum).

[hokum]

Which type of scoring would be fairest for a 3-7 table IMP pairs event?
I'm trying to get more players in, but in the meantime I'd like to use the fairest scoring possible. Cross IMPs, or IMPs vs datum (into VPs?)? And if we use datums with so few tables, is it logical to use all scores when calculating the average (ie, leave in the top and bottom results)?

Thanks

[helene_t]

hokum, on 2014-September-25, 02:27, said:

Which type of scoring would be fairest for a 3-7 table IMP pairs event?

X-IMPs is better. Others have explained why.

Maybe matchpoints or Patton is best but if you want to make it IMPs, go for X-IMPS.

[Trinidad]

hokum, on 2014-September-25, 02:27, said:

Which type of scoring would be fairest for a 3-7 table IMP pairs event?

All scoring methods are equally fair. In a socialist philosophy it might be considered fairest to give everybody 10 points, irrespective of the result. So, "What is the fairest type of scoring?" is the wrong question. (I am saying this, because others have pointed this out in their posts above.)

An example of a possible right question would be: "What method, for a 3-7 table IMP pairs event, would best resemble team of four IMP scoring?" In that case, I would answer that Cross-IMPs does the job best.

Rik

[hokum]

Cheers

[helene_t]

Trinidad, on 2014-September-25, 07:17, said:

In a socialist philosophy it might be considered fairest to give everybody 10 points, irrespective of the result.

OK, a philosofy teacher would tells us that "fairness" is a weasel word. But what hokum probably asked is which scoring system gives the most consistent winners. A system that leads to particular "strong" pairs winning consistently may be rewarding the wrong things (such as a talent for gaming the scoring system) but at least they must be rewarding something. A system that produces more random winners would be less "fair". Maybe "reproducible" or "consistent" would be a better word.

I find it an interesting question whether matchpoints or IMPs or even total points is "fairest". I don't think that it can be resolved conclusively using mathematics alone since we don't know whether slam swings are more skill-related than overtrick swings and therefore deserve to be rewarded more. But I think it would be possible to address the issue empirically. If for example it turned out that the XIMP hands played in the main room would give more consistent winners if they had been scored as MP, it would be a strong case for MP. Obviously any scoring system that makes the stakes higher at red/red than at white/white is suboptimal but that doesn't prove that IMPs is always less fair than matchpoints.

The arguments in favour of X-IMPs are partly fairness related (butler scoring must be more noisy than X-IMPs), and I think it should be possible to quantify this difference using a purely mathematical argument (or simulations).

[NickRW]

helene_t, on 2014-September-26, 02:55, said:

If for example it turned out that the XIMP hands played in the main room would give more consistent winners if they had been scored as MP, it would be a strong case for MP.

From the point of view of a club owner or committee member, within reason (not sure how to define), the more inconsistent the result (i.e. the more different winners over time) the better as more of the punters get their day in the sun.

As MPs count all boards equally and IMPs do not, probably MPs gives the more consistent result. Equals most clubs should be doing things by IMPs

[Trinidad]

helene_t, on 2014-September-26, 02:55, said:

The arguments in favour of X-IMPs are partly fairness related (butler scoring must be more noisy than X-IMPs), and I think it should be possible to quantify this difference using a purely mathematical argument (or simulations).

In practice I agree with this. Butler must be more noisy.

But suppose that we have two phenominal bridge pairs in a Butler competition. One player is more phenominal than the other: Not only does he bid and play perfect, but after each board he can predict exactly how all the other players in the field (the Mrs Guggenheims, and the Unlucky Experts) must have played the board and how they must have scored.

On one board, this player is in a 2♣ contract. He can play safe for 9 tricks or take a finesse for 10, risking that he will take only 8 tricks. He looks around the room, calculates and sees that if the 10th trick isn't there, because the finesse loses, the datum will be 96, rounded to 100. If the finesse wins and the 10th trick is there, the datum will be 113, rounded to 110.

The player realizes that he has to take the finesse for the 10th trick: the +130 will give him 1 IMP if the 10th trick is there, the +90 will cost him 0 IMPs if the finesse loses. In the other phenominal pair, the star declarer isn't able to calculate what the field is doing and he estimates (wrongly) that it is better to play safe for 9 tricks. As a result, they finish second.

Seems fair to me. (But yes, it is far fetched.)

Rik

[aguahombre]

NickRW, on 2014-September-26, 06:58, said:

From the point of view of a club owner or committee member, within reason (not sure how to define), the more inconsistent the result (i.e. the more different winners over time) the better as more of the punters get their day in the sun.

Part of the ad campaign?

[Vampyr]

NickRW, on 2014-September-26, 06:58, said:

As MPs count all boards equally and IMPs do not, probably MPs gives the more consistent result. Equals most clubs should be doing things by IMPs

As there has been discussion of "fairness", it should be pointed out that IMP pairs is a very "unfair" way of scoring. If you get a bad result because your opponents got to a tough game or slam, there are no teammates to protect you. Also an arrow-switch in IMP pairs will benefit some pairs and harm others.

Much better, if you want to spread the wins around, is to use handicaps.