[kchatz]

I think players should be truthful in their self evaluations, perhaps backed up by official bridge league rankings. I frequently play in speedballs on bbo. What I don't like is people purposely falsifying their credentials. I see people who classify themselves as "novice," who end up playing a lot of artificial conventions, I also don't like people considering their rank to be "private." This kind of gamesmanship is inappropriate in competitive bridge.

[hrothgar]

kchatz, on 2019-January-02, 08:05, said:

I think players should be truthful in their self evaluations, perhaps backed up by official bridge league rankings.

And I want a pony.
I think that we are both equally likely to have our wishes granted.

[barmar]

johnu, on 2019-January-01, 17:39, said:

To the extent that players in JEC matches play against random players, they would have a rating based on those random players. Say their rating corresponds to scoring ~70%. If you play those JEC players and break even, then you should be ~70% if you don't play anybody else.

What I actually meant was how would you calculate a rating for JEC and his teammates, since he only plays against other expert teams (well, except for the JEC vs BBF games, which we've finally given up on since participation dropped off). It's not fair to compare their results against experts with other players' results against randoms.

Of course, it doesn't really matter -- JEC will never show up at the Partnership Desk or ask to sit in a random game, so his rating will never be an issue. But it's indicative of a general problem: if you have disjoint communities of players, ratings within each community are not necessarily comparable. The best MBC player might only be considered "very advanced", not as good as the experts and champs who tend to play in set games.

[hrothgar]

barmar, on 2019-January-02, 09:48, said:

What I actually meant was how would you calculate a rating for JEC and his teammates, since he only plays against other expert teams (well, except for the JEC vs BBF games, which we've finally given up on since participation dropped off). It's not fair to compare their results against experts with other players' results against randoms.

Of course, it doesn't really matter -- JEC will never show up at the Partnership Desk or ask to sit in a random game, so his rating will never be an issue. But it's indicative of a general problem: if you have disjoint communities of players, ratings within each community are not necessarily comparable. The best MBC player might only be considered "very advanced", not as good as the experts and champs who tend to play in set games.

If you have completely disjoint sets of players, then there is nothing wrong with having completely disjoint sets of ratings.
There is no reason to assume that ratings need to be cardinal rather than ordinal.

[barmar]

hrothgar, on 2019-January-02, 09:51, said:

If you have completely disjoint sets of players, then there is nothing wrong with having completely disjoint sets of ratings.
There is no reason to assume that ratings need to be cardinal rather than ordinal.

If they're permanently disjoint, that's true. It becomes more of a problem when players occasionally wander from one cohort to another. They'll show up with a rating that isn't really representative of how they're expected to do in this community.

I'd have wildly different ratings if they were based on my results in robot games, acbl speedballs, or JEC games.

[johnu]

barmar, on 2019-January-02, 09:48, said:

What I actually meant was how would you calculate a rating for JEC and his teammates, since he only plays against other expert teams (well, except for the JEC vs BBF games, which we've finally given up on since participation dropped off). It's not fair to compare their results against experts with other players' results against randoms.

Of course, it doesn't really matter -- JEC will never show up at the Partnership Desk or ask to sit in a random game, so his rating will never be an issue. But it's indicative of a general problem: if you have disjoint communities of players, ratings within each community are not necessarily comparable. The best MBC player might only be considered "very advanced", not as good as the experts and champs who tend to play in set games.

I'm guessing a substantial number of JEC opponents play against at least some better random players. And those better random players play against some average random players. JEC doesn't need to play against random players to have a rating.

As far as MBC players, they normally get soundly trounced by JEC so if they had above average ratings, JEC should have very high rankings for consistently pounding the MBC teams.

[a Maverick]

it can be a prob BUT no real solution cos we all make MISTAKES/COCK UP/ take the wrong line from time to time.................Plus sometimes even the beginner( because they are) come with the killer lead/play.......thats part of the fun

[ray_p]

661_Pete, on 2018-December-31, 03:24, said:

I know this may be a 'touchy' subject and has probably been debated ad nauseam before now!

I know BBO is - for me at any rate - a site for merely playing bridge 'for fun' - but nevertheless I find it a bit of an irritation, sat facing a partner who's self-rated themself 'advanced' or 'expert' - only to find that they're no better than a beginner...

Any solution?

My (EBU based) NGS ranking (derived from live bridge of course) currently stands at 53% and I rank myself on BBO as "intermediate" which I think is fair and reasonable. I just wish others would do likewise...

Perhaps if players' long-term IMPs or MPs scores (as visible via the "myhands" utility) could be shown on their profiles, when you're thinking of joining a table? I know the calculation of this is not so sophisticated as NGS which takes into account one's partners' rankings - but it would be a start...

I've watched the EBU's system evolve over a number of years and accept the algorithm is reasonable and gives that ladder-like scale. However my criticism is the way that it (or any constant grading system) influences players behaviour in some detrimental ways.

Bridge players are already naturally competitive!! and having a grade, many people become overly focused on improving their ranking to the detriment of enjoying the game, and exhibiting good manners. Particularly when their desire (obsession?) to keep up their grade turns into: self-selecting only certain partners, the matches entered, avoiding hosting or new partners or anything that might affect their grade, But especially when they become less tolerant of other's mistakes at the bridge-table.

To balance this criticism, I should acknowledge that grading does server to engender more commitment to the game, by feeding on players desire to judge themselves against others and push up that ranking

However any automatic grading system must also consider the 'law of unintended consequences' when implemented, for the detrimental behavioural effect the ranking system can bring (as I have often noticed with players under the EBU system).

[barmar]

ray_p, on 2019-May-21, 19:49, said:

However any automatic grading system must also consider the 'law of unintended consequences' when implemented, for the detrimental behavioural effect the ranking system can bring (as I have often noticed with players under the EBU system).

Indeed, Lehman Ratings on OKbridge have a significant effect on who people will play with and against. Supposedly it accounts for the quality of partner and opponents (you're expected to do poorly against stronger opponents or with a weak partner, so losing by the expected amount shouldn't reduce your Lehmans), but people avoid taking the risk of doing worse than the Lehmans predict.

[Vampyr]

ray_p, on 2019-May-21, 19:49, said:

I've watched the EBU's system evolve over a number of years and accept the algorithm is reasonable and gives that ladder-like scale. However my criticism is the way that it (or any constant grading system) influences players behaviour in some detrimental ways.

Bridge players are already naturally competitive!! and having a grade, many people become overly focused on improving their ranking to the detriment of enjoying the game, and exhibiting good manners. Particularly when their desire (obsession?) to keep up their grade turns into: self-selecting only certain partners, the matches entered, avoiding hosting or new partners or anything that might affect their grade, But especially when they become less tolerant of other's mistakes at the bridge-table.

To balance this criticism, I should acknowledge that grading does server to engender more commitment to the game, by feeding on players desire to judge themselves against others and push up that ranking

However any automatic grading system must also consider the 'law of unintended consequences' when implemented, for the detrimental behavioural effect the ranking system can bring (as I have often noticed with players under the EBU system).

I was worried about this sort of thing when the EBU first proposed the NGS, but I don’t care much anymore. In my experience, people regard the scheme as no more than a bit of fun, and in fact in the North London club and no doubt other places, it is an advantage to have a lower rating.

It is a pity, though, that players cannot decide (before the game but not after, not naming names) for their game to be exempt from the NGS. This would remove the problems you are concerned with.

[pescetom]

Vampyr, on 2019-May-23, 03:43, said:

It is a pity, though, that players cannot decide (before the game but not after, not naming names) for their game to be exempt from the NGS. This would remove the problems you are concerned with.

Are you thinking about rubber bridge or similar here?
I find it hard to imagine a club exempting a pairs tournament, although it might make sense after our friday evening dinner and wine session

[nige1]

Bridge is a competitive game so most players welcome grading systems
as a rough measure of their current skills and rate of progress For example
The EBU NGS is popular and seems to work well
IMO, however, players should be able to opt out or keep their own rating private.

[Vampyr]

pescetom, on 2019-May-23, 06:37, said:

Are you thinking about rubber bridge or similar here?
I find it hard to imagine a club exempting a pairs tournament, although it might make sense after our friday evening dinner and wine session

I doubt that everyone in attendance would choose to remove their game from applying towards the NGS. But if they wanted to, why not?