Scoring the bidding contest

[inquiry]

[hv=w=saqjxxhaxxxdqxxca&e=stxhdakt9xxxckjxx]266|100|This is a hand posted by hanp, I didn't want to hijack their who to blame thread, but this is a good one to try to illustrate the principle[/hv]

 

In han's post, he noted that 7♦ makes while 7NT did not. We can gather from that fact that the spade king was offsides (else 7D, 2C, 1H, 3S). In the bidding contest, the spade king has to be somewhere (when bidding is over, you can see the king). So if this hand was in the contest, you would say "yes 7NT is top spot" if the king was onside, or whoops 7D is better if the spade king is offside.

 

The fact is, ideally we (me?) will try to come up with the best spot for the hand, and the placement of cards in the actual NS hands shown at the table are immaterial.

 

So for this hand, I would ignore the possibilty of 8-0 club suit (♦ ruff at trick one) but would at least consider a possible 6-0 spade split, but I think I would not factor that in either. So 7NT is top (tie for) or bottom 1/2 time each. So I would award an "average" of 6 out of 12 for 7NT.

For 7♦ it makes all the time, being the best contract half the time, and being a very good second best the other half the time. I would award it a 10 out of 12.

 

Then 6NT is never optimal, and in fact, when it makes, 7NT makes (some squeeze chance when spade King is offside by ducking first heart or not getting a heart lead makes it easy). So I would reward 6NT just less than 6D, but 3NT beats 5D, and all games beat any partscores... Then you have to start factoring in 6S contracts. You could win heart ace and duck a spade, type of thing. And here the possibility of a diamond ruff becomes very real (either opening lead or after losing spade hook). There are even long spades to the king to consider or if someone can hold up the spade king for a round or two... So I would give 6S a lower score.

 

Off the of my head I would come up with somthing like this:

 

7D = 10

6D = 7

7N = 6

6N = 5

6S = 4

3N = 3

5S = 2

5D = 1

others = 0

 

Obviously if you were bidding this hand in the contest, you might reasonably think something is underscored or overscored. For instance, you might feel 6S is certainly going to score better than 6D h higher percentage of time and deserves to be moved higher than 4 on average. OR I might decide 7D is rare enough to award it 11 (not 12 because sometimes 7NT will make). Or maybe you feel that at matchpoint 6S is safer than 6N, making sometimes when spade king is offside, etc.

 

I have not doubt I will get many wrong, but it is a thankless job. I have already had two messages lobbying for certain contract to rated higher (than they anticipate I will score them). I feel like this type of discussion will be great fun when the hands are posted. So mabye we might want to delay the start of next round until the discussion of last round reach a conclusion about the scores. Because I couldn't commit the kind of time needed to calculate exactly (or even close to exactly, because opening lead will affect the scores on many hands) the odds of different contract.

 

I just fall back on this is meant to be fun, and provide us with entainment value... and if not, well and I am use to having stones cast at me.. I move slow but duck fast.

[Lobowolf]

I nominate Justin to the Appeals Board.

[JLOGIC]

I nominate Justin to the Appeals Board.

I'd be happy to do it obv but since I'm in the contest there is definitely a conflict of interest. In fact, I was one of the people who messaged ben about a hand and how I thought there might be difficulty scoring it!

[Cyberyeti]

If the max is 10/12 on some boards and 12/12 on others, shouldn't you scale by 1.2 on the former ? so all boards are equally important.

[jdonn]

If the max is 10/12 on some boards and 12/12 on others, shouldn't you scale by 1.2 on the former ? so all boards are equally important.

No. Just like real bridge, some hands are flatter than others. Also a 12/12 means you reached a better contract than a 10/12 even if nothing higher than a 10 was available (often in the sense that the 12/12 contract was harder to reach).

[P_Marlowe]

Hi,

 

just an suggestion - you could try to estimate how good a contract is using GIB / DD,

of course if the result of the simulation is completly oppossite to your first idea, than

one would need to have a closer look, most of the time the parameters for you simulation are wrong.

 

Given that this is also time consuming, I would only do this in close cases, ..., and

after all it is just a "fun" contest.

 

With kind regards

Marlowe

[hanp]

I like the idea of discussion on the forums, but in the end I think it best if Ben just decides what the scores are. That also means that us contestants should not complain too badly if the scores are not entirely to our likings.

 

I think that the scores should be close to what you'd expect to get in real life, not what you'd get in a double dummy simulation. For example, if a slam is 50% on a heart lead but 80% on any other lead then that's quite a good slam to reach. The exact odds are hard to determine, especially since we want the score not to depend on the auction. Maybe it should be treated as a 65% slam.

 

Pff, glad I'm not making up the scores!

[Hanoi5]

I think it best if Ben just decides what the scores are. That also means that us contestants should not complain too badly if the scores are not entirely to our likings.

I agree completely. You can contest a score by giving reasonings and it might be taken into account but in the end it has to be the host decision. After this contest is finished you can run the next if you feel you can do a better job (and I believe no one will do that, because it's quite time consuming and because with Ben's experience and method the scores given by him will be highly accurate).

[JLOGIC]

Yeah, maybe if you don't agree with a score you can appeal it to Ben and give some reasoning and then he will just have the final say whether it stands or he changes it.

[hotShot]

I would suggest to re-shuffle the 26 opponent cards a number of times and check the impact on the contracts.