Skill level description

[vuroth]

matmat, you're just miffed that there's no slot to rate your kibitzing skills. :)

[Winstonm]

You forgot the I Suck category, which would hold most of us I fear. :)

[ArtK78]

It seems to me that some sort of community based collective intelligence approach to ranking players would be our best shot at ranking players.

Considering that most players think every other player is an idiot, perhaps a "community based collective intelligence approach" isn't the right way to phrase it.

 

How about a "community based collective stupidity approach?"

 

:)

 

Actually, there is quite a bit of merit to this idea, matmat's response notwithstanding.

[Trumpace]

It seems to me that some sort of community based collective intelligence approach to ranking players would be our best shot at ranking players.

 

...

Actually, there is quite a bit of merit to this idea, matmat's response notwithstanding.

I agree. This is probably a good statistical/machine learning problem. It might even be possible to come up with a model which can withstand a few malicious ratings. Maybe helene_t or hrothgar have some ideas about this.

 

But, having a rating system will open a worse can of worms, IMO.

[hotShot]

If there were a working rating system, than some 90% of the bridge player would find out, that their level is much lower than they think.

Once they learn how bad they really are, they might loose the interest to play at all.

 

All the "local heros" would loose their local admirers once they get to know that their local hero is far from good. They would want not play in an environment that exposes their true skill.

 

Real experts would like to have some peace and quiet time on BBO, with their skill exposed they will be bombarded with invitations to team games and the like from total strangers.

 

So how many player would really want to know their rating or would want it published. And is it good for the game or BBO?

 

Who would play with weaker friends, use the "Help me find a game" button or leave the table open for strangers if this would lower the rating?

What will be the climate among the players is someone makes a stupid mistake because his phone rang or what if opps won't allow an undo?

Player are already rude to TD's to get an adjustment, what if the result would ruin their rating. Just imagine someone would get a bad score because of a psyche ....

[CSGibson]

It seems to me that some sort of community based collective intelligence approach to ranking players would be our best shot at ranking players.

Considering that most players think every other player is an idiot, perhaps a "community based collective intelligence approach" isn't the right way to phrase it.

 

How about a "community based collective stupidity approach?"

 

:rolleyes:

 

Actually, there is quite a bit of merit to this idea, matmat's response notwithstanding.

Yeah, but then you get into the "worldclass/expert/advanced when I'm paying attention and 1-3 levels lower when I'm drunk/high/watching the game while playing" thing. Someone might be very much an expert and play very badly in the right (wrong?) circumstances.

[babalu1997]

There are about 200 nations, lets say each has sent 10 teams with 6 members to events to represent them over the last 40 years. So there could be 12000 player that fit the description of WC.

I will soon apply for citizenship in tuvalo, pop. 11,640, then i will be world class.

[cherdano]

The technical solution would be to let players rank other players - kind of like the google page rank algorithm. Here's how it would work -

hmmm...

why wouldn't this work... oh. right.

 

i think Cherdano is an @$$hat. I hate his guts. I don't care that he finished high in the LMP or BRP or whatever. I think he's a RANK BEGINNER.

 

:/

Seems like a good start, maybe the system makes sense after all.

 

Oh, btw, qwery was talking about players ranking players, not kibitzers ranking players...

[matmat]

Oh, btw, qwery was talking about players ranking players, not kibitzers ranking players...

it might be more objective if the kibitzers do the ranking :)

[babalu1997]

There is time maybe BBO to revisit Lehman rating policy, or at least give it a try. If members dont welcome it, its easy to go back to old.

One of the reasons for the success of BBO is the lack of ratings.

 

 

 

Those who care about lehmans`s can play at okbridge or get some king of assessment software some talk about here.

 

I am a member at okbridge. I mistakenly played ONE, ONE rated imps game there, and , my imps rating is 49.87, below 50, i became a pariah. Players will play with you as long as they get the rating, the first 30 percent that get they leave the table, taking your rating points along with them.

 

Want to try an even more torturous experience? Try the rating system at the swans site.

you get rated against the cards dealt to you, and declarer gets the rating so you make a jacoby transfer and partner gets the rating, your partner forgets the transfer and you end up in silly contract.

 

Do you think people are happy with rating there ? No way, they complain of rating raiders who leave game because of bad cards, or ask for redeals evryt time they do not get an opening hand.

 

Experts can have privacy at bbo. Shut off the chat!!! Make table invisible !!! have team matches by invitations only!!! disallow kibbitzers!!!

 

Just leave the ratings alone.

 

Leave bbo as is.

[qwery_hi]

There is time maybe BBO to revisit Lehman rating policy, or at least give it a try. If members dont welcome it, its easy to go back to old.

One of the reasons for the success of BBO is the lack of ratings.

 

 

 

Those who care about lehmans`s can play at okbridge or get some king of assessment software some talk about here.

 

I am a member at okbridge. I mistakenly played ONE, ONE rated imps game there, and , my imps rating is 49.87, below 50, i became a pariah. Players will play with you as long as they get the rating, the first 30 percent that get they leave the table, taking your rating points along with them.

 

Want to try an even more torturous experience? Try the rating system at the swans site.

you get rated against the cards dealt to you, and declarer gets the rating so you make a jacoby transfer and partner gets the rating, your partner forgets the transfer and you end up in silly contract.

 

Do you think people are happy with rating there ? No way, they complain of rating raiders who leave game because of bad cards, or ask for redeals evryt time they do not get an opening hand.

 

Experts can have privacy at bbo. Shut off the chat!!! Make table invisible !!! have team matches by invitations only!!! disallow kibbitzers!!!

 

Just leave the ratings alone.

 

Leave bbo as is.

I agree. This is precisely why I let my okb membership expire.

 

However, lack of ratings is a minor factor in the success of BBO. The fact that it is a great software and free to boot are the major factors.

 

For example, even today there are thousands of "experts" on BBO who refuse to let an advanced play at the table. Having ratings would not change their behaviour, although it might change their rating.

[Rossoneri]

Well, the best way I guess would be to find out and make friends with people whom you agree with and play with them. And of course, accept that if you play with randoms people will just come and go.

 

Star novices are nothing...I know of a star who created another account so that no one would bug him!

 

Generally I would say let's not have a rating system, after all, the real experts know who the other real experts are and they play with each other so often. (I can almost always find a real expert/WC table to kib, I am sure matmat would agree with this.)

 

Rating systems always seem to bring out the worst in people IMO, e.g. Yahoo Games.

 

P.S. How many times has someone posted about this on the forums? I've lost count.

[H_KARLUK]

Well I am really sorry. I honestly admit mostly creative ideas.

 

Anyway I wonder do they still say "Do not be in a hurry to tie what you cannot untie" in England for such things ?

[paulhar]

Even though this topic appears from time to time, the comments were interesting. Right now, it seems that the way it works is that everybody knows that the ratings are absolutely meaningless (I played at a table with two experts and a 'beginner' once and the beginner was the best player at the table.) Players who frequent BBO have their list of favorite partners and look for one of them to pop on. When I played, I had about 30 people that I thought were comparable in system and skill and it wouldn't take long for one of them to show up if I didn't mind waiting a few mintues.

 

If everybody did this, it would take a long time for a new player to BBO to break in. Nobody knows him and nobody trusts his rating. How sad. Kind of like the new guy to town wanting to play at the local duplicate. Might be an awesome player but if he isn't known, he will play with the dregs for many weeks, getting the types of results that aren't usually associated with awesome players, and remaining unnoticed.

 

While any kind of rating system (I kind of like the "other players can submit changes to the ratings" approach, which in my opinion would take two votes from the same table to have an effect - and this would have to happen several times before the rating changes) might solve this problem, the negative effects of any rating system would outweigh the positives. Negatives include (but are not limited to:) failure to want to play when you might not be on your best game; failure to want to play with a foreigner because you're afraid some of your bidding might be misunderstood (and therefore be thought to be "awful"), not getting partnered because you are known to have cast a vote against someone else, or not playing with (or against!) someone because he was known to cast a vote against someone else. A Lehman rating system has other drawbacks; people will be less willing to play with an unfamiliar partner for fear that a misunderstanding will cost them Lehman points.

 

So I think we're stuck with what we have, learn to love it. :P

[kgr]

I agree that rating systems could have negative factors.

But for me the main advantage of on-line bridge is that I can play it whenever I want and that I can start play immediatly. This is how I prefer to play on-line bridge.

As a result I almost never play with someone I know and the rating is therfor very important for me to have an idea of players. Now this is really useless. I try to be very polite, but also without a rating system this is sometimes very difficult if an expert does one stupid thing after the other.

My opinion:

=> If you play mostly with pick-up partners then the advantages of a rating system would be more important then the disavantages.

=> If you don't play with pick-up partners then a rating system is unimportant for you and you shouln't care for a rating system.

To be honest: If I can get something for free then I don't like to pay for it. That is the main reason I play on BBO. If I would play on another site (not that I plan to!!) then the lack of a rating system on BBO would be the reason.

[matmat]

=> If you play mostly with pick-up partners then the advantages of a rating system would be more important then the disavantages.

 

You're likely right, however it would only make sense if the system were accurate; One interesting statistic, that is vaguely a rating, perhaps more of an index, would be the ratio of hands played to the number of distinct partners and the # of online hands played. This wouldn't necessarily indicate the skill of the player, but it would give you a good idea of whether, at the risk of inventing words that don't exist, the player is playable with. (this would need to get adjusted for individual tourneys where the format mandates partner changes every board or two -- and other such corrections that I can't really think of at the moment).

 

=> If you don't play with pick-up partners then a rating system is unimportant for you and you shouln't care for a rating system.

I tend to try to avoid pickup situations. Even when I don't succeed, I don't look at the self rating; I think in the current state it is meaningless. In fact, I try to avoid pretty much anyone labeled as expert/world class whose name I don't recognize from somewhere. They typically come with an ego and attitude to match. not worth it.

[onoway]

It seems that most of the time the rating system, such as it is, works ok ( and this from someone who brought this topic up once in frustration!) Generally as a guideline I have found it often to be pretty close so that when it ISN"T, it brings on shock and dismay. I have found often what the profiles say to be very useful as a counter to what the rating is..if someone says advanced followed immediately by NO TRANSFERS (e.g). they may be a wonderful player but unlikely to be the pard I am wishing to play with. Not that I am a hotshot, by any means! So no profile can be (but not necessarilly is) a warning sign.

 

If it is truly distressing it would be possible to track who goes into the PAID tourneys such as ACBL and BBO_PL or whatever as people who pay for tourneys on a regular basis seem to tend to have a better idea where their skills sit on the scale. Or simply play in them yourself. Supporting BBO is a GOOD thing :) Bewildered players do sometimes show up there as well, but not so often. And anyone can have a bad day.

 

Ratings would be a disaster for tourneys..subs can already be a problem and if you were rated according to your results they would likely be impossible to get.This morning I subbed for one (the final) hand of a tourney. It was worth +7.5 imps but my result for the TOURNEY showed as -20something..who would sub if they thought that would have an effect on who would play with them?

 

Also, hardly anyone ever complains about the people who are true experts who wander around calling themselves intermediates or even novices, it is only when blindsided by people in the other direction. I think that there are too many negatives (already noted by others, above) associated with some sort of more "official" ratings. Perhaps something might be done for somebody's personal interest, such as using Bridgemaster, but to make it something "official" which would be attached whether you wanted or not would imo be a big mistake.

 

Perhaps it is more reasonable to assume that the occassional anguish of being trapped is like getting your car dinged in the parking lot, just part of the way life is sometimes. And it helps having the forums to moan in :)

[benlessard]

Ive played on about 20 game servers from money backgammon, GO,shogi and dumbass high luck games, I can assure you that ratings rarely pose problems and bring significant benefits.

 

In low/NO chance games rating are a most, slaughtering beginners isnt fun and intermediate players simply dont like getting crushed vs top players. Fighting vs a player slightly above your strenght seems ideal if you want to imporve (ladder) In games with high luck factor rating are cool because the rating tend to be close anyway and its a still a fairly useful scale.

 

Anyway numerical ratings can be put in fairly large categories so if we got 5 categories people wouldnt care that much about their ratings.

 

Ratings could also be hidden (can be used for automatic matchup etc)

 

 

Its possible that bridge is an exception because its a partnership game but im pretty sure ratings would be a nice plus to BBO.

 

 

What is way more important however is setting up a partneship room a room for already formed partnership. Starting a TM is a real pain in the ass on BBO, i dont know any game server where you have to write opponents name to play against them.

[hotShot]

Ive played on about 20 game servers from money backgammon, GO,shogi and dumbass high luck games, I can assure you that ratings rarely pose problems and bring significant benefits.

 

In low/NO chance games rating are a most, slaughtering beginners isnt fun and intermediate players simply dont like getting crushed vs top players. Fighting vs a player slightly above your strenght seems ideal if you want to imporve (ladder) In games with high luck factor rating are cool because the rating tend to be close anyway and its a still a fairly useful scale.

 

Anyway numerical ratings can be put in fairly large categories so if we got 5 categories people wouldnt care that much about their ratings.

There are games and sports where it is easy to do an individual rating, but in team games / sports it gets a lot more complicated.

 

First of all you need a scoring system that is capable of representing the playing strength.

Bridge does not have such a scoring system. Lets take a look at MP-Scoring, if you pick a field of really good world class player, someone will have to loose that tourney and in an MP tourney they will get a score lower than 50%. Playing in the club of a nearby retirement home some friends of mine scored more than 80% recently.

IMP scoring isn't helpful too. If 2 WC teams fight over 100 boards and one wins with 4 IMP's, they have a 0.04 IMP/board average. Bunny bashing should give you a score of (much) more than 1 IMP/board.

 

Now how do you pinpoint a good or bad score to an individual player? Who's at fault, if you follow partners lead directing dbl or lead the suit he signaled and it costs a trick?

Who's at fault if a Texas transfer or a super accept gets you to high?

What score do you get when you're dummy? What if you misbid or overbid, but get to be dummy?

 

What kind of boards to take into account? How many boards should a new partnership be allowed to play, before it affects their rating? What if a player runs before the board is finished, when the substitute player could just concede the remaining tricks.

 

How do you get the strength of a typical BBO tourney field, if it contains WC-player and novices and each pair is only playing a very small subset of the contestants.

[bid_em_up]

You know... I'm interested to know if we actually have numbers on how the self-ratings are distributed on BBO. What % of people list themselves in each category? Sure, there are admittedly people inflating their status, but how big of an issue is it REALLY?

Uday posted these numbers several months back. I can't seem to locate the thread though, maybe he would be kind enough to do it again?

[RichMor]

I agree that rating systems could have negative factors.

But for me the main advantage of on-line bridge is that I can play it whenever I want and that I can start play immediatly. This is how I prefer to play on-line bridge.

As a result I almost never play with someone I know and the rating is therfor very important for me to have an idea of players. Now this is really useless. I try to be very polite, but also without a rating system this is sometimes very difficult if an expert does one stupid thing after the other.

My opinion:

=> If you play mostly with pick-up partners then the advantages of a rating system would be more important then the disavantages.

=> If you don't play with pick-up partners then a rating system is unimportant for you and you shouln't care for a rating system.

To be honest: If I can get something for free then I don't like to pay for it. That is the main reason I play on BBO. If I would play on another site (not that I plan to!!) then the lack of a rating system on BBO would be the reason.

This is my profile too. I am a frequent drop-in player but I don't want to plan and schedule games in advance.

 

I don't need a graded rating system for pards or opps. I would like to be able to see some kind of 'minimum competence' flag in the profile of the other players. Something I could see and believe.

 

It should be possible to set up a voluntary minimum competence test, something based on an intermediate online quiz.

 

Q1: Some of the red suits are:

A. Hearts

B. Spades

C. Hugo Boss

D. All of the above

 

Nobody has to take the test. Anyone who doesn't want to take the test can still play. Once you take and pass the test you get a little 'seal of approval' in your profile. It doesn't go away and is not dynamically recalculated like a Lehman rating.

 

But those who take and fail an intermediate-level test and still rate themselves as Expert could be avoided.

 

Fred, does this make any sense ?

[matmat]

Q1: Some of the red suits are:

A. Hearts

B. Spades

C. Hugo Boss

D. All of the above

 

Nobody has to take the test. Anyone who doesn't want to take the test can still play. Once you take and pass the test you get a little 'seal of approval' in your profile. It doesn't go away and is not dynamically recalculated like a Lehman rating.

 

But those who take and fail an intermediate-level test and still rate themselves as Expert could be avoided.

 

Fred, does this make any sense ?

how many times can i take the test?

[RichMor]

Q1: Some of the red suits are:

A. Hearts

B. Spades

C. Hugo Boss

D. All of the above

 

Nobody has to take the test. Anyone who doesn't want to take the test can still play. Once you take and pass the test you get a little 'seal of approval' in your profile. It doesn't go away and is not dynamically recalculated like a Lehman rating.

 

But those who take and fail an intermediate-level test and still rate themselves as Expert could be avoided.

 

Fred, does this make any sense ?

how many times can i take the test?

I think you could pass in three tries.

 

Some of the questions are a little harder.

[barmar]

There are games and sports where it is easy to do an individual rating, but in team games / sports it gets a lot more complicated.

 

First of all you need a scoring system that is capable of representing the playing strength.

Bridge does not have such a scoring system. Lets take a look at MP-Scoring, if you pick a field of really good world class player, someone will have to loose that tourney and in an MP tourney they will get a score lower than 50%. Playing in the club of a nearby retirement home some friends of mine scored more than 80% recently.

IMP scoring isn't helpful too. If 2 WC teams fight over 100 boards and one wins with 4 IMP's, they have a 0.04 IMP/board average. Bunny bashing should give you a score of (much) more than 1 IMP/board.

 

Now how do you pinpoint a good or bad score to an individual player? Who's at fault, if you follow partners lead directing dbl or lead the suit he signaled and it costs a trick?

Who's at fault if a Texas transfer or a super accept gets you to high?

What score do you get when you're dummy? What if you misbid or overbid, but get to be dummy?

 

What kind of boards to take into account? How many boards should a new partnership be allowed to play, before it affects their rating? What if a player runs before the board is finished, when the substitute player could just concede the remaining tricks.

 

How do you get the strength of a typical BBO tourney field, if it contains WC-player and novices and each pair is only playing a very small subset of the contestants.

Have you read the description of Lehman ratings? It tries to take most of these issues into account.

 

If you're playing against opponents who have lower ratings than you, you're expected to do well. If you do well, your rating doesn't change, but if you do poorly your rating goes down. Conversely, if you're playing up, you aren't penalized if you do poorly (that's expected), but your rating goes up if you do well.

 

If you're playing with a worse partner, but manage to do well, your rating goes up because you managed to overcome the handicap; his doesn't go up as much because he's riding your coattails. But if you do poorly, you're not penalized too much -- the assumption is that the novice partner is messing up most of the time.

 

If you get too high on some board, it's just one board. Ratings should be based on large numbers of boards, so that occasional bad (and good) luck is lost in the noise. Long-term results can reasonably be assumed to reflect your actual skill level.

 

Your objections could just as easily be applied to any bridge event. A typical club game is about 24 boards, is that really enough to assess how well someone plays? Yet we call the pair that comes in first the "winners" and award them masterpoints. Most BBO tourneys are even shorter, just 12 boards. If this is enough to award masterpoints, surely a few hundred boards would be enough to establish ratings.

[hotShot]

I'm familiar bis the Lehman rating.

 

Are you familiar with movement theory?

 

The best movement is a 7 table 13 round Howell. Each pair plays each other pair, every board is played by all participants.

 

We will score a MP-tourney. Now take 13 pairs of identical strength and one superior pair. Pairs of the same strength get the same result, the superior pair always gets the top, their opponents the zero.

Since there are 13 rounds some pairs will play 7 times on the same side as the Superior Pair (SP) and 6 times on the other side, others will play six rounds on the same side and 7 rounds on the other.

Since the SP gets the top (6MP) the other pairs on the same side score (0+1+2+3+4+5)/6 = 2.5 MP. On the other side SP-opps get 0 MP and the other pairs on this side get (1+2+3+4+5+6)/6= 3.5 MP.

 

The top pair scores 2*13* * 6 = 156 MP = 100%.

The lucky pairs play 6 rounds on the same side ( 6*2*2.5 ) and 7 times on the other side one of these they play the SP (2*0 + 6*2*3.5) => 72 MP = 46.15 %.

The unlucky pairs play 7 rounds on the same side ( 7*2*2.5 ) and 6 times on the other side one of these they play the SP (2*0 + 5*2*3.5) => 70 MP = 44.87 %.

 

So the best possible movement separates pairs of equal strength by 1.28%-points, just because of the seating.

 

In a typical BBO tourney you don't play most of the field, the effect of the movement are much bigger. Especially the middle positions of a swiss movement a lottery.

These results are useless for a rating.