[130012]

It's about rating a player, an idea followed in most of the commercial websites. Similar to "Like" and "Comment" in Facebook

1. A player can be rated on a 10 point scale by other players to get a better idea of their skill level.

2. A feedback box which can contain the comments of other players with their name visible.

3. The player should be able to see above 2. But cannot Edit it.

4. The right of edition of objectionable comments should remain with the concerned authorities, the player may request them for edition.

[1eyedjack]

A novel variation on an otherwise well-worn theme, so for that reason worthy of serious discussion, I guess.

Some minor refinements to suggest:

1) The votes of someone who himself has a high rating should carry greater weight than those with low rating (for "himself" read "herself" as appropriate, of course)

2) Older votes should perhaps depreciate in weight if not at some point drop off entirely

3) You may wish to consider a simple system where you can just "up-vote" or "down-vote" a player depending on whether you think their current rating is too low or too high, contrasted with a more complex system where you vote a number in the rank 1 to 10 which is considered in the weighted average. Simple sounds good.

4) However tweaked, you would not want an individual vote to have a measurable effect, but it would require a hundred or so consistent votes to change the displayed result.

Objections? Well, several, really, among which:

A) I don't like the idea of implementing a system that is likely to result in a drain on BBO management when they get inundated with edit/reset requests. OKB allows this, but then you pay an annual sub for the service.

B) Anyone who doesn't like their current rating can just set up a new account and undermine the entire process. I don't like the idea of implementing a system that encourages the proliferation of new accounts and discourages stability. This objection equally applies to any rating system that you might wish to suggest, automated or otherwise.

c) I am perfectly capable of forming my own opinion on the quality of another player, after just 3 hands. I have little enough trust my fellow BBOer's judgement of other players as it is for me to place more credence in such a rating than I can form for myself.

D) Players will inevitably down-vote a colleague for a number of reasons, including both table manners and play quality. It may be no bad thing that the final tally is a mix of skill and other factors, as long as those interpreting the figures do not assume (which they will, of course) that it is all about play quality.

E) As Hrothgar mentioned in another thread, the absence of a rating system is not explained by technical difficulty. So providing an alternative technical solution in the belief that BBO cannot otherwise implement an automated method rather misses the point. Its absence is socio-politically motivated.

[helene_t]

I thought of a friends-of-friends concept like in g+ and facebook, but it requires people to use it with discipline so I don't think it is realistic. One problem is that lots of "friends" of mine are just good players worth watching, not someone I would realistically think I could partner. Others are bad players who I like to chat with but wouldn't play a speedball with. So the information that someone is a "friend" (or "upvoted" or whatever) of mine isn't useful to other people.

It may be a good idea to allow users to tag other users as "good player", "slow player", "rude" or whatever. I strongly believe such tags shouldn't be visible to anyone (other than the person who made the tag) but they might be useful for compatibility calculations. It would take some research - some people would try to get at other players by tagging them as rude, clueless, slow and unethical, so the system needs to be made robust against such things.

Maybe one place to start is to try to make a model that can predict whether a newly formed pair will continue playing together and whether they will tag each others as friends. Then once such a model has proved succesful it could be used for compatibility calculations.

[mgoetze]

Each year at my company we do an employee satisfaction questionnaire. The answers are on a 1-5 scale, but the consultants doing the survery throw out the exact data because it's too noisy and only consider whether the answer was a "top box" (4-5) or not.

But I'm sure on your hypothetical 1-10 bridge rating scale everyone will give very careful consideration to and have the same understanding of the difference between a 6 and a 7.

[steve2005]

Most people aren't a good judge of other's abilities (a waste of time). If we can't trust a person's rating of self why would we be believe their rating of others?

[130012]

steve2005, on 2014-December-03, 16:46, said:

Most people aren't a good judge of other's abilities (a waste of time). If we can't trust a person's rating of self why would we be believe their rating of others?

"MANY" people aren't a good judge of other's abilities, not "MOST"

[130012]

1eyedjack, on 2014-December-03, 00:14, said:

A novel variation on an otherwise well-worn theme, so for that reason worthy of serious discussion, I guess.

Some minor refinements to suggest:

1) The votes of someone who himself has a high rating should carry greater weight than those with low rating (for "himself" read "herself" as appropriate, of course)

2) Older votes should perhaps depreciate in weight if not at some point drop off entirely

3) You may wish to consider a simple system where you can just "up-vote" or "down-vote" a player depending on whether you think their current rating is too low or too high, contrasted with a more complex system where you vote a number in the rank 1 to 10 which is considered in the weighted average. Simple sounds good.

4) However tweaked, you would not want an individual vote to have a measurable effect, but it would require a hundred or so consistent votes to change the displayed result.

Objections? Well, several, really, among which:

A) I don't like the idea of implementing a system that is likely to result in a drain on BBO management when they get inundated with edit/reset requests. OKB allows this, but then you pay an annual sub for the service.

B) Anyone who doesn't like their current rating can just set up a new account and undermine the entire process. I don't like the idea of implementing a system that encourages the proliferation of new accounts and discourages stability. This objection equally applies to any rating system that you might wish to suggest, automated or otherwise.

c) I am perfectly capable of forming my own opinion on the quality of another player, after just 3 hands. I have little enough trust my fellow BBOer's judgement of other players as it is for me to place more credence in such a rating than I can form for myself.

D) Players will inevitably down-vote a colleague for a number of reasons, including both table manners and play quality. It may be no bad thing that the final tally is a mix of skill and other factors, as long as those interpreting the figures do not assume (which they will, of course) that it is all about play quality.

E) As Hrothgar mentioned in another thread, the absence of a rating system is not explained by technical difficulty. So providing an alternative technical solution in the belief that BBO cannot otherwise implement an automated method rather misses the point. Its absence is socio-politically motivated.

You suggestions are reasonable, A button named "Player statistics" can be provided, It should contain all the relevant information about the player

[130012]

Let us see some examples here, consider criket and football, How do we know Sachin and Messi are best?? Their ratings!!
Here we have a verity of players in a very large numbers. A rating scale will definitely help the other players to understand their partners. And this what Bridge is all about, understanding your partner and opponent's bid.

[barmar]

steve2005, on 2014-December-03, 16:46, said:

Most people aren't a good judge of other's abilities (a waste of time). If we can't trust a person's rating of self why would we be believe their rating of others?

There have been studies that show that even though individuals may not be very good at estimating a variety of things, if you get a bunch of them to make these estimates, and then average them, this is quite often very accurate.

http://en.wikipedia....om_of_the_crowd

[diana_eva]

130012, on 2014-December-03, 18:17, said:

Let us see some examples here, consider criket and football, How do we know Sachin and Messi are best?? Their ratings!!
Here we have a verity of players in a very large numbers. A rating scale will definitely help the other players to understand their partners. And this what Bridge is all about, understanding your partner and opponent's bid.

I like the idea of a way to measure players' "likeability" (which in my opinion is not a skill thing, rather a combination of niceness, flexibility and to some extent - of skill) but the way you describe it, you seem to wish a rating system. That can probably be done without asking input from users. As for understanding your partner, you just need to agree on a common system.

[mycroft]

I don't want 2-4 at all. It's bad enough what partners say to me at the table because they're "anonymous", but everybody else I get to play with getting to see them? *Including me?* Never mind the vendettas and gang-razzing (anyone remember the chaos around a certain G- a few years back?) No matter what the validity of anyone's comments, the vitriol and the illfeeling caused by it all was worse than any potential bad play/learning/teaching. Now imagine if one of those packs started and *people who had never even played with you* start badmouthing you in comments visible to everybody and locked to your profile because one of their pack did and you're today's source of amusement.

And you couldn't *pay me* to be the person who gets all the "reset/this isn't valid/these are old, I'm better now/I'm being slagged" requests for comment deletion, never mind the "this is abuse" comments.

I also agree with the people who say "I can tell in 3 hands how good a player is" - with a caveat that "if I can't tell, they're much better than me".

Sure, we can tell if Sachin or Murali or Pieterson is better, given ratings - but I can tell if Meckstroth or Burtens or Korbel are better, too. But you're not asking about them - you're never going to be playing with them anyway. You're wondering if Smith or Jones or Dhaliwal at your national company outing is better based on what random people think of their play against other randoms.

Never mind all the socio-political games and the "any system can be gamed" games, remember the two rules: "in a survey, 90% of bridge players were better than their partners" and "everyone just wants pickup partner to be as good as they are". If you look at it hard enough, you'll see why any visible rating attempt for "get me a partner" is doomed to fail; and why even if it's invisible, it's just better if we don't know it exists.

And I have some ego, too - I don't want to know my rating. I'm a decent enough A player in my area; I try to be a polite partner (but I notice there are things I have to work on there, too); on a good day, with a decent partner, I'll crack the overalls. I'm not winning, and I'm comfortable with that. But there are a number of people I would put in the same "decent enough A player" category I think I'm better than; I'm sure that a rating would say that there are a fair number of those, I'm wrong. Leave me, thank you, to my delusions!

[GreenMan]

barmar, on 2014-December-04, 01:09, said:

There have been studies that show that even though individuals may not be very good at estimating a variety of things, if you get a bunch of them to make these estimates, and then average them, this is quite often very accurate.

http://en.wikipedia....om_of_the_crowd

The so-called wisdom of crowds is only useful in some situations. One situation where it goes awry is when the estimates or judgments are not independent -- that is, you can see what others have said. The suggested rating system here would have that flaw; I suspect it would degrade its usefulness to near zero.

A wise-crowd system also requires diversity in its contributors. I don't know what "diversity" would mean in a BBO context, but we shouldn't assume we have it without some research. (We're already a narrow group -- Internet-connected bridge players with free time. It may be that diversity in country, age, and playing level would be sufficient, but even then we can't be sure the raters represent such a range.)

[GreenMan]

And then there's the changes in behavior that such a rating system would create. Once rating points become valuable, people will do what they can to acquire them. We may expect that the system will only reward and reinforce behavior we want more of, but unintended consequences always lurk.

[mycroft]

"Like me on Facebook, Follow me on Twitter, upvote me on BBO"? I know you have no idea who I am, but it's just one more buttonpress...