Hand Evaluation

[campboy]

Now that's a way to really upset the customers. No-one wants a different imp scale. :(

Trinidad's suggestion is easy to implement without changing the IMP scale at all. Just add the four scores, multiply by 0.7, then IMP the result.

 

Anyway, I don't think anyone is disputing that it is better to use an illogical system players like than a logical system they dislike.

[Trinidad]

Trinidad's suggestion is easy to implement without changing the IMP scale at all. Just add the four scores, multiply by 0.7, then IMP the result.

That is true, but it means that the players need to do the work (over and over again: each team, for each board). Adjusting the IMP table would mean that the TD does the work once.

 

Rik

[CamHenry]

Trinidad's suggestion is easy to implement without changing the IMP scale at all. Just add the four scores, multiply by 0.7, then IMP the result.

 

Anyway, I don't think anyone is disputing that it is better to use an illogical system players like than a logical system they dislike.

 

campboy, I'm not convinced many players - ECL or otherwise - will want to multiply by 0.7, as it looks far too much like "take away the number you first thought of".

 

But yes, ECL scoring is illogical and absurd - the UBC sometimes even pulls off a win! ;)

[bluejak]

When you decide whether something is "logical" in its use in an event, I believe the wishes of the majority are important. To have something that a minority think correct because they think it logical does not make it logical. Logic means something follows from something else: I think it is logical to use something people believe in, rather than something that a few mathematically inclined people construct.

[campboy]

I do not think that something is logical simply because it follows from something else unless the thing it follows from is itself logical.

[PeterAlan]

When you decide whether something is "logical" in its use in an event, I believe the wishes of the majority are important. To have something that a minority think correct because they think it logical does not make it logical. Logic means something follows from something else: I think it is logical to use something people believe in, rather than something that a few mathematically inclined people construct.

But we already do a whole lot of things because they are mathematically right, despite the fact that the vast majority of players do not know why or even believe that it is logical - perhaps arrow switching 1/8 of the boards played in a Mitchell is the most obvious example. At least 90% of players don't understand this, yet I'm quite sure that most would understand it perfectly easily if told there's one IMP table for teams of 4 and another for teams of 8. They're already used to variations in VP scales.

 

I'm pedantic enough to diasgree fundamentally with your interpretation of the word "logical". The notion that "something people believe in" is therefore "logical"? - apply this to religious beliefs, and the many and conflicting varieties thereof, and you'll see what an absurd statement it is.

[barmar]

I'm pedantic enough to diasgree fundamentally with your interpretation of the word "logical". The notion that "something people believe in" is therefore "logical"? - apply this to religious beliefs, and the many and conflicting varieties thereof, and you'll see what an absurd statement it is.

But many people who believe would probably say otherwise. Who gets to dictate what's logical?

 

Actually, the use of "logical" for this discussion is fundamentally different from using it in a pure mathematical or scientific context. Bridge scoring is not something you can decide the truth or falsity of. You may be able to decide whether it's fair. However, in this particular case, I think logical is being used to mean "reasonable" -- in that case, belief is an appropriate way to judge it. Or it could mean "consistent with similar forms of the game", and then you have to decide how important this is.

[PeterAlan]

But many people who believe would probably say otherwise. Who gets to dictate what's logical?

I chose the example I did because there are plenty of people with conflicting beliefs out there and it can not therefore be appropriate to attach "logical" to one belief position (ie a matter of faith, without evidential support) without being prepared to do so to the other. And then you get logical contradiction; so "logical" it wasn't.

 

[This is entirely without prejudice to any religious beliefs, or lack of them, that I or any other posters may hold, about which I am saying nothing whatsoever.]

 

Actually, the use of "logical" for this discussion is fundamentally different from using it in a pure mathematical or scientific context. Bridge scoring is not something you can decide the truth or falsity of. You may be able to decide whether it's fair.

I don't know what you mean about the truth or falsity of bridge scoring, but it's certainly something about which you can say what is mathematically appropriate. I find the thrust of bluejack's argument very strange - it's precisely on such technical matters that ordinary players expect the regulatory authorities to take a lead and sort out the relevant technical issues on their behalf. Instead, he wants to abdicate that leadership role, apparently on the grounds that "100,000 lemmings can't be wrong". [And before anyone goes there, I'm fully aware of the questions about the veracity of lemming suicide.]

 

However, in this particular case, I think logical is being used to mean "reasonable" -- in that case, belief is an appropriate way to judge it. Or it could mean "consistent with similar forms of the game", and then you have to decide how important this is.

Indeed. So use "reasonable" or "sensible" and don't go on arguing that "logical" is right.

[Trinidad]

IMO, what is logical is not so relevant. I would think that "being consistent" is relevant.

 

If you want to be consistent when it comes to the fundamentals of IMP games, you use an adjusted IMP scale.

 

If you want to be consistent when it comes to the numbers on the teams of four IMP table, you also use this table for teams of eight scoring.

 

Whatever you chose is fine with me. Both methods are fair, as long as the contestants are informed ahead of time of the conditions.

 

But it would be good to realize that chosing the second option is equivalent to saying that a train has 64 wheels, because the 9:15 train from Reading to London has 64 wheels. The only thing I have done is hand a method to calculate the number of wheels on the train, depending on its length, making sure that the train looks like a train when it has 2 cars and when it has 24. Some might think that this complicates life. I think it makes life a lot easier when the number of wheels on a train depends on the number of cars. But what is needed for that is the notion that it is not the number of wheels that makes a train a train, but rather the amount of wheels per car and the fact that it moves on a track.

 

Rik

[Trinidad]

Players find it easiest and reasonable.

 

Customer service is important in running this game.

Now, essentially what happened in this thread is that you got a minor customer complaint ("the illogical way"). You asked why this particular customer felt this way (excellent customer service). And because of this, you got constructive feedback.

 

Then, you seem to discard the feedback, "because your customers don't feel like that":

I direct and used to run a major inter-County team of eight. Last time we tried feedback, one of the questions was whether we should change the form of scoring from teams-of-eight to two teams-of four. No change was nearly unanimous.

Whereas in reality, your customers have only said that they preferred the bad thing (adding the scores, IMPing using the team of four IMP table) over something even worse (dividing into two teams of four matches).

 

Maybe the next time you ask for feedback you could ask the customers whether they want to use the IMP table with the numbers for a team of four match which leads to "bending of the IMP scale" at places where it wasn't meant to bend or whether they want to use the IMP table that is specifically designed for a team of eight match where the principle of IMP scoring is kept correctly, but the actual numbers are different from the team of four table (equivalent to the different VP tables for different numbers of boards).

 

Rik

[barmar]

I don't know what you mean about the truth or falsity of bridge scoring, but it's certainly something about which you can say what is mathematically appropriate.

Scoring a game is essentially arbitrary, it can be done any way the game designers want.

 

However, if they declare a specific set of goals for the scoring system, you can determine whether the scoring system achieves those goals. That could be a "logic" of scoring.

 

For instance, when the rules for doubled undertricks was changed, it was because players and administrators felt that the old rules made it too easy to sacrifice. That feeling was a popularity issue, but the scoring changes were a logical consequence.

[AlexJonson]

Scoring a game is essentially arbitrary, it can be done any way the game designers want.

 

However, if they declare a specific set of goals for the scoring system, you can determine whether the scoring system achieves those goals. That could be a "logic" of scoring.

 

For instance, when the rules for doubled undertricks was changed, it was because players and administrators felt that the old rules made it too easy to sacrifice. That feeling was a popularity issue, but the scoring changes were a logical consequence.

 

One hates to be argumentative, but you are equating the word reasonable with the word logical. Not a good idea.

[PeterAlan]

Scoring a game is essentially arbitrary, it can be done any way the game designers want.

 

However, if they declare a specific set of goals for the scoring system, you can determine whether the scoring system achieves those goals. That could be a "logic" of scoring.

 

For instance, when the rules for doubled undertricks was changed, it was because players and administrators felt that the old rules made it too easy to sacrifice. That feeling was a popularity issue, but the scoring changes were a logical consequence.

You seem to be confusing the underlying scoring structure of the game, which for argument's sake I'm prepared to accept as being wholly arbitrary, with the issue here, which is the appropriate way of combining a collection of such scores, whatever they may happen to be, to produce a single score.

 

When the underlying scoring system is such as it is in bridge, and the independence of the events is as it is, this process is one that is reasonably well understood mathematically (statistically, if you prefer). What's being said here is that if this particular way of combining N scores meets certain desirable statistical criteria, as we suppose is the case with the standard IMP scale, then combining 2N such scores in the same way won't continue to meet those same desirable criteria, and there's no point in pretending that it will.

[Vampyr]

What's being said here is that if this particular way of combining N scores meets certain desirable statistical criteria, as we suppose is the case with the standard IMP scale, then combining 2N such scores in the same way won't continue to meet those same desirable criteria, and there's no point in pretending that it will.

 

Why does this matter if it meets other desirable criteria, chiefly that the players are used to it and like it?

[barmar]

You seem to be confusing the underlying scoring structure of the game, which for argument's sake I'm prepared to accept as being wholly arbitrary, with the issue here, which is the appropriate way of combining a collection of such scores, whatever they may happen to be, to produce a single score.

I was using the underlying scoring structure as an example, but didn't think I was equating them.

When the underlying scoring system is such as it is in bridge, and the independence of the events is as it is, this process is one that is reasonably well understood mathematically (statistically, if you prefer). What's being said here is that if this particular way of combining N scores meets certain desirable statistical criteria, as we suppose is the case with the standard IMP scale, then combining 2N such scores in the same way won't continue to meet those same desirable criteria, and there's no point in pretending that it will.

True, it doesn't meet the same criteria, but who says it should?

 

I'm not familiar with teams-of-8 (AFAIK, they're not played much on this side of the pond). But it sounds like the controversy is not too dissimilar from whether IMP Pairs should be scores as cross-IMPs or versus a datum. For many years the datum method was mostly used, I think simply because it was easier when scoring by hand; computers can easily calculate cross-IMPs, and it's considered more consistent with the design of the IMP scale, so it's now most common.

[PeterAlan]

Why does this matter if it meets other desirable criteria, chiefly that the player are used to it and like it?

Because those are probably not informed views (can we accept that this is not a patronising statmenet? - see discussion above). What I'm suggesting is that there should be an honest acceptance of the shortcomings; until that's forthcoming, statements like this are, to mix my metaphors, just sweeping the issue under the carpet and hiding behind skirts.

[Vampyr]

Because those are probably not informed views

 

Informed of what?

[Trinidad]

Why does this matter if it meets other desirable criteria, chiefly that the players are used to it and like it?

The point is that this discussion got started because one of the players clearly did not like it.

 

In addition, there is the fact that people have already been thinking of a solution. After all, there must have been a reason why people were polled with the question whether they wanted the format changed from team of 8 to 2 teams of 4. I obviously don't have any hard evidence, since I was far from involved, but it seems to indicate that there must have been some kind of problem with the scoring. In general, people do not try to look for solutions if there isn't a problem.

 

Because those are probably not informed views

Informed of what?

Uninformed about the issues that f0rdy mentioned in post #13 and uninformed about the fact that there is a very simple way to correct these issues.

 

It may well be that the field of players is divided into two parts:

1) The happy oblivious: The part that is completely unaware of the oddities of the use of the team of four IMP scale in team of eight events.

2) The uneasy aware: The part that realizes that there is something odd with the IMP scale in team of 8 events, dislikes it, but accepts it since they can't really see what causes this oddity and since they are unaware of the solution.

 

It might well be that in the future there will be a growing third:

3) The annoyed enlightened: The part that knows what causes the scoring oddities in team of eight events and knows that there is a simple solution, but is unhappy with the fact that this solution is not used.

 

The easiest way to introduce innovations and get them accepted is to implement them at the point when the third group is large. The best way to introduce innovations is as early as possible (once the innovation has been developed) while making sure that the third group grows rapidly to facilitate the acceptance of the innovation.

 

Rik

[PeterAlan]

Informed of what?

As I said, see discussion above. This was an excellent summary, if you don't want to scroll back.

 

F0rdy and campboy clearly showed that, by applying the team of four IMP table to teams of 8, you are changing the IMP scale. They showed that the scoring gets "warped" when you use the team of four IMP scale for team of eight games. The team of four IMP table was never meant for teams of eight. (Just think what this would do if someone would organize a "team of 400" match ...)

 

My post was showing that it is very easy to adapt the existing IMP scale for teams of 6, 8, 10 or 400, in such a way that this "warp" is compensated for. This doesn't make it a different IMP scale. It is the exact same IMP scale, but now for another number of participants on the team.

...

Mathematicians, statisticians and engineers know these kind of methods as "scaling". ... I can see that scaling in such an abstract thing as bridge scoring may appear confusing, but it is simply a question of doing things right.

[gordontd]

When I've run a heat for a national teams-of-eight competition which is scored by adding all four scores together, I've provided the players with the appropriate IMP scale. Most of them don't notice. None of them (including Stefanie) have ever said they would prefer to have the unsuitable team-of-four scale.

[Trinidad]

When I've run a heat for a national teams-of-eight competition which is scored by adding all four scores together, I've provided the players with the appropriate IMP scale. Most of them don't notice. None of them (including Stefanie) have ever said they would prefer to have the unsuitable team-of-four scale.

Darned, I was soooo hoping that in the future I would see tournament announcements with a phrase like: "Form of scoring: Teams of 8, according to the (well-known, famous, brilliant, ...) Trinidad scale".

 

This means that I will have to work harder on earning the Nobel Prize to achieve immortality.

 

;)

 

Rik

[f0rdy]

Scoring a game is essentially arbitrary, it can be done any way the game designers want.

 

However, if they declare a specific set of goals for the scoring system, you can determine whether the scoring system achieves those goals. That could be a "logic" of scoring.

 

For instance, when the rules for doubled undertricks was changed, it was because players and administrators felt that the old rules made it too easy to sacrifice. That feeling was a popularity issue, but the scoring changes were a logical consequence.

 

Of course. There are at least 3 (?) common methods of getting from raw scores (420, 1430, etc) to comparisons for duplicate, and everyone knows that they require different approaches to aspects of bidding and play. My concern about the "teams-of-eight with standard IMP table" is that I think most people sit down and assume they're playing "IMPs", ie ordinary teams of four scoring, and that they should bid and play accordingly; this isn't entirely true, particularly with regard to large scores from slams or bad sacrifices.

 

I suspect that nearly everyone playing in this competition for Cambridge UBC falls into either the "happily oblivious" or the "enlightened" (probably with varying degrees of annoyance). The question of course, is which of Trinidad's categories members of the other 7 counties fall into, and whether the "happily oblivious" would be resistant to change; I agree entirely with Bluejak that these events should be run with whatever keeps the players happy, within bounds of fairness (which I agree this doesn't affect).

 

Peter

 

PS Yes, this is analogous to the "how to score IMP pairs" question, but I think this is less bad than IMPing against a datum, which is completely barmy; I actively avoid playing in Butler-scored events unless required to.

[gwnn]

Who gets to dictate what's logical?

Logic?

[StevenG]

The question of course, is which of Trinidad's categories members of the other 7 counties fall into, and whether the "happily oblivious" would be resistant to change

 

"Happily oblivious" to a man, I'd suggest - certainly at B & C team level. The IMPs scale is just a lookup table; I can't imagine anyone caring if it was changed slightly.

[gwnn]

BTW why would people be so resistant to a change in IMP scale (w.r.t. number of people on a team) when they are completely fine with a changing VP scale (w.r.t. number of boards)?