[badderzboy]

For a bit of fun, I've been writing an Excel spreadsheet (it would be better in MS Access but I thought most people have MS Excel but not MS Access) that can score both Duplicate & Cross-Imps for 32 boards maximum.

It accounts for pairs not playing a board etc and seems to work against a sample of BBO scoreboards I've played.

It needs more testing of course and some comparative runs in my local club.

I have set it to handle a max of 32 pairs so up to 16 tables playing the board at present due to size constraints on the file.

Question 1:
I've only played at a local club where we have had 16 tables once. In people's experience, is it common for more than 16 tables to play the same boards or do clubs set up sections when numbers get up higher?

Once I've tested it and ran it past the TD's at our club I would probably release it as shareware somewhere tbc?

Do u think TDs would be interested in this sort of automation or using Cross-IMP scoring?

Steve

[Free]

I'm not sure, but I think there are already programs which handle almost all scoring methods, for lots of tables. If your spreadsheet would be able to cover as many scoring methods as these programs, and for at least as many tables, AND it would be cheaper, then you might have some success.

Btw, it's always great to have something of your own ofcourse

[whereagles]

Actually, what sort of "cross imps" are used in BBO tourneys?

As far as I recall, cross imps means comparing your score against every other pair, not against a datum score (butler scoring). Being so, I would expect scores of +200 imps if I bid and make a slam no one else makes. Instead, when I check my score sheet on a cross imps tourney, I see a mere +10 for it.

Does cross imps mean butler scoring, for BBO purposes? It would definitely seem so to me...

[gabika73]

Cross-imp means averaging the IMP you achieve against all pairs playing the same side.
So making a vulnerable slam that no one else bids gives you the expected 13 IMP.

It is like playing a team match versus all other tables, and averaging the IMP's on each hand.

I would be pretty much surprised if BBO used a different method.

btw, it is a better form of IMP-scoring than a butler (average of IMP's vs IMP against an average)

[Free]

I think you're wrong there. Imo cross imps is comparing your score to all other scores individually. So if you have 5 tables, and you make slam which nobody finds, you'll get an expected 4*13 = 52 imps. This is much heavier to play Mistakes cost a lot, and overtricks can gain some imps...

[MickyB]

Cross-Imping refers to either method, whether you choose to take the sum or the average of the scores is irrelevant. You may have lost 20 times as many imps by missing a game, but you'll get 20 times as many back for any good score. I prefer to take the average so that I have a better feeling for how good or bad a score is.

[whereagles]

Ok I it see now.

[inquiry]

Free, on Sep 22 2004, 11:13 AM, said:

I think you're wrong there.  Imo cross imps is comparing your score to all other scores individually.  So if you have 5 tables, and you make slam which nobody finds, you'll get an expected 4*13 = 52 imps.  This is much heavier to play    Mistakes cost a lot, and overtricks can gain some imps...

Excuse me, but this is utter nonsense. It cost exactly the same. Well, this isn't precisely true... but all you have done is change the scale for everyone. Let's say there are 15 other tables (16 is all). And you eveyone plays 4S and everyone makes exaclty 10 tricks but you. You make 11.

In one method, you wil 1 imp (average one imp better than all you opponents), in the other you win 15 imps. But is 15 really worth more than 1? Of course not. While 15 is a bigger number, all you have done is change the lowest obtainable imp number from 1 to 15. You have changed the scale, that is all. You are still just one overtrick ahead of the field. The data is the same, just normalized to a different scale.

There are, in fact, two different ways to figure out imps. You could imp your score across the filed, or you could calculate the average score for all boards and then imp your result with that average. Let me show how this works, using a trhee board example... ...

TABLE 1 3SN= +140
TABLE 2 5DEX-2 +300
TABLE 3 4SNX= +790

What would be the Imps of the players at the three tables if you calculate total imps and divide by 2?

Table one loses 4 imps to table 2 and 12 imps to table 3. This is a total loss of 16 imps over two comparisons for average -8 imps.

Table two loses 10 imps to table 3, and wins 4 from table one (average -3 imps), and Table 3 wins 10 from 2 and 12 from 1 for average 11 imps. And you no what, -8 + -3 = -11 for tables 1 and 2, while table 3 is +11. Balanced. Free's total imp method table one would be -16, table 2 -6, Table 3, +22. But as you see the scale is just doubled. The effective difference between the pairs is really the same (percent difference, what you need to do on the next board to catch up, etc).

The second method is to average the three scores (+140 +300 +790) which comes to an average of +410. Now you compare this average to each score. USing this method, pair one gets (410-140=-270 or -7 imps), pair two gets (+410-300 = -110 or -3 imps), and pair 3 gets (790-410=380, or +9 imps).

A couple things to see here. First, in this second example, the bad scores are not as bad and the good scores are not as good. Second, the plus imps and the minus imps no longer add up. The BBO uses the first method.

Ben

[mink]

badderzboy, on Sep 22 2004, 04:40 PM, said:

...
I've only played at a local club where we have had 16 tables once. In people's experience, is it common for more than 16 tables to play the same boards or do clubs set up sections when numbers get up higher?

At my local club we play 26 boards. If there are 19 tables or less, there is only 1 section. With 20 or more tables, we make a 13-table-section and another section with the remaining tables. Boards are duplicated during the first round, so we can compute one single result for both sections.

Quote

Once I've tested it and ran it past the TD's at our club I would probably release it as shareware somewhere tbc?

Do u think TDs would be interested in this sort of automation or using Cross-IMP scoring?

No, programs dedicated to this task do their job better than a spreadsheet application probably could.

Karl