Handicap Pairs with Butler Scoring

MikeK

Following the 'success' of our Handicapped team events, I have been volunteered to run a handicap pairs!
The idea behind the event is to try and integrate our Improvers Group more into the Club.
Most of the Improvers aren't EBU members so don't have an NGS.
But I think there is a spreadsheet with all their historical results on it.
As some Improvers do have NGS, hopefully I can use that data to 'key' the rest.
Then I can estimate an NGS for everyone.

I think that will work, and I could score in MatchPoints.

But just to make things a bit more complicated I thought I might use Butler Scoring.
(The Improvers are just pleased to make a contract, gunning for overtricks is a bit subtle!)
Does anyone have any suggestion as to how I could handicap this?
For Teams IMP scoring I have been using 0.3086*difference in NGS grades.
(I found that in the NGS Guidance.)
But other factors also appear in the NGS Guidance, I am confused!

Senior_Kibitzer

There's information about Butler and Cross-IMP Pairs in the main text of the full guide. I suggested on a previous thread here a lazy way by waiting till the next day when the results will have been processed for NGS and seeing which pairs has had the best improvement in NGS from the event.

But let's say you want to use Cross-IMPs (More popular nowadays than Butler).

The guide says that a gain of 1 IMP at Cross-IMP Pairs is equivalent to an increase of 6.48% on the Match Point score for a board. So if you are playing 24 boards, each % difference in the average NG of the two players in a partnership is worth 24 divided by 6.48 IMPs (=3.70 IMPs)

If you are playing 27 boards, each % difference in the average NG of the two players in a partnership is worth 27 divided by 6.48 IMPs (=4.17 IMPs)

It would be a reasonable approximation to 4 IMPs for every 1% difference in the average NG of the two players in a partnership.

Indeed at my club we have Cross-IMP Pairs every Tuesday evening and I compare the IMP results to equivalent MPs (for NGS) using 4 IMPs per 1% and so, in "near enough" terms, +4 IMPs is worth 51%, +20 IMPs is worth 55% and +40 IMPs is worth 60%.

MikeK

Thanks Barrie.
Not sure whether I will use Butler or X-Imps.
With the latter outlying/outlandish results will be discounted which might be 'fairer.'
I expect we will only play 18 boards.
(Some of the Improvers are not that quick, and we are having a bit of a party too.)
So I will use either 2.8 per 1%NGS for X-Imps or 2.3 if Butler, assuming I have got my sums right.

Senior_Kibitzer

@MikeK said:
So I will use either 2.8 per 1%NGS for X-Imps or 2.3 if Butler, assuming I have got my sums right.

I agree re Cross-IMPs but I think you have a typo for Butler. I make it 3.3 IMPs per 1% NG difference