Real Experts?

[uday]

Use self-ratings, combined with some form of subjective ratings by partners and opponents.

 

Is the thinking that you'd accumulate subjective ratings w/o worrying about the source of the rating (ie., the rating of the person making the rating ) ?

 

How could we express the result simply?

 

 

For example: 10 people mark XYZ as an expert, 100 as advanced, 200 as intermediate. Now someone checks up on XYZ. Would you expect him to see

 

expert: 10

adv: 100

int: 200

 

and draw his own conclusions?

[awm]

Use self-ratings, combined with some form of subjective ratings by partners and opponents.

 

Is the thinking that you'd accumulate subjective ratings w/o worrying about the source of the rating (ie., the rating of the person making the rating ) ?

 

How could we express the result simply?

 

 

For example: 10 people mark XYZ as an expert, 100 as advanced, 200 as intermediate. Now someone checks up on XYZ. Would you expect him to see

 

expert: 10

adv: 100

int: 200

 

and draw his own conclusions?

There are established ways to do this sort of thing, via reputation systems and so forth. It's a complicated topic so I won't go into a lot of details here (nor do I have more than a basic familiarity with the relevant literature).

 

I believe Dr Todd implemented such a thing for BBO a while back; you might want to contact him if you're interested in this stuff (or talk to someone at netflix.com, amazon.com, ebay.com, etc).

[hotShot]

Suppose I take a set of boards played by top-WC players.

I give them to a bunch of beginners.

After that I analyze the boards where the suit played is the same.

Suppose I have 20 boards to compare.

The WC-Player made not a single mistake.

Pair A made a trick loosing error in 19 of 20 boards, pair B made a trick loosing error in every board. When they play pair B gives away an extra trick each board, but pair A gives it back 19 of 20 times.

 

So pair A has 19 times the same result as the WC players and one score is better. Pair B has 19 times the same result as the WC players and one that is worse.

 

Looking at the results only obviously pair A is of galactic strength worthy to represent earth in intergalactic championships, while pair B is a little weaker than WC probably expert level.

 

(Self)ranking is hard ......

[sceptic]

actually E Bay have a ratings system. why can Bridgebase not adopt such a system, it would not be hard to work out a method of rating someone and also the ratings have some bearing as to the strength of the person ranking them also, so the higher standard someone is, the more emphasis put on their vote. it would probably take a few months for the rankings to settle down to mean anything

 

it may be possible to only see the rankings your friends or partners have given to a specific player

[mike777]

So far I have not seen anything easier, that works than what I suggested ages ago.

 

The goal is not to make Zia the best rated player on bbo.

 

The goal is not to rate the best double dummy players.

 

If you win a Nat open tourney or an open WC tourney you are good player on bbo.

 

If you win a Reg open tourney player you better than most.

 

that is good enough for bbo.....

 

We are not making a grandmaster chess list.

[hotShot]

Assume a board where a contract can be made by taking the only possible finesse. Now assume that a squeeze will have a 90% chance while the finesse has only 50%. So in 10% of the cases players who don't know a squeeze will rate this expert to be a fake.

What if a player picks a tricky 55% line and fails?

 

It is much easier to rate an ebay seller/buyer, than to rate a bridge player.

[helene_t]

If someone sells crap or stolen goods on ebay, or uses a stolen credit card for purchase, it is a serious problem. It is vital to ebay to have a mechanism to prevent, detect and stop such behaviour effectively, even if it means that some ebay users may feel overly policed and ebay staff spend a significant amount of resources implementing and running the rating system.

 

If a self-rated expert on BBO makes a vulnerable 2-level overcall on Jxxxx and then yells at his p for not rescuing him (or calls the TD because opps did not alert the double) then, well, it's nasty but the players involved (and the TD) can just mark the jerk as an enemy, and in the more severe cases one can report the insident to abuse@. The system may not provide the same security level as that on ebay but it's adequate. We are not going to pay too high a price for maximum safety against such incidents.

 

What if a player picks a tricky 55% line and fails?

 

It is much easier to rate an ebay seller/buyer, than to rate a bridge player.

Not sure how concerned one should be about this. The noise will even out in the long run. There may be a bias against good players who rely on subtle plays and scientific bidding, but the latter is not worth much in a pick-up partnership anyway. Simon's "Unlucky Expert" may feel underrated but then again, what rating should he really have?

 

I think most of us can tell a low-intermediate from a good intermediate on the basis of his frequency of obvious blunders.

 

I think a more serious problem with peer-rating is that friendly but bad player would get good marks by some and bad marks by others. It won't be possible to give different marks for different virtues since people will not be able to separate the two things.

 

A simple proposal: compute, for each other player, the number of friend-ratings he has received from my friends, and the number of enemy-ratings he has received from my friends.

 

Not sure if that would be useful. Maybe it would turn out disastrously.

[Gpm_bg]

there is a rating which is undisputable, and that is the star symbol. That is not a self-rating, because that person has been designated as a class player .

All of the rest self-rating of the players is a matter of private honor and trust. Who if not I can confess my own level compared with what i see every day live or in Internet, reading as well. If i think that i'm a genius, who can told me that i'm not. But if i'm not honest, and declare that i think i have world class skills, that i will be at least "funny" for the kibitzers watching me if it's not even 5% true.

So to declare yourself as WorldClass, Expert and etc. first think about expectations about that.

Looking the people from the live grand tournaments, even only at BBO Vugraph broadcasts, do you say often WOW, How he/she made it, i couldn't even imagine that.

Even you can select your level freely, i think that it will be like to wear other person cloths.

Do you will feel comfortable if you pretend that you are someone else ?

I wouldn't.

[bid_em_up]

Adam,

 

I like your proposed definitions a lot better than the current BBO ones. I've always felt like the current ones were rather vague, and especially left too little distinction between "expert" & "world class", since to me the people who "have success in national tournaments", at least in the U.S., are the same usual suspects who I would consider "world class".

So do I.

[bid_em_up]

Ben/Uday, I'm curious.

 

Do you have any way of coming up with the numbers for each category that people have rated themselves?

[ArcLight]

The problem with rating other players is:

 

- whats to prevent me from making some BBO nics and rating my pal Expert and him reciprocating

 

- whats to prevent me from rating people I don't like 1-2 levels below what I perceive their skill to be. If I ask what a bid means and you don't answer (maybe you don't speak English) you must be Intermediate or lower.

 

- how do you know what an expert / world class player is unless you are very good? Suppose you balance 3♣ with 5 mediocre clubs and a balanced hand and it works? Are you Advanced because this one time you got lucky? You pard might think so. A few months back I was playing a pick up game and a nationally recognized player happened to sit as my pard B) Against 3NT I lead a suit and had pard won and returned it we would have set them. But the correct play was to duck, and he did, and they made the hand. He apologized, but later said it wasn't clear to have taken that trick. Would many players have rated him a Beginner?

 

What I would like out of a rating system is the ability to play in games at (or above) my level, and not be fooled into wasting my time (and becoming frustrated) with much weaker players. I don't need the system to be very accurate just accurate enough that the players I play with aren't too far below my level.

[keylime]

Adam,

 

I also like your ratings because there's some tightly defined criteria that helps provide definition to the rating system. It also gives benchmarks to achieve for those who are improving.

[uday]

Do you have any way of coming up with the numbers for each category

 

Sure. Using people who have logged in at least once since Nov 1, and using the self-rating from the profile

 

 

private      22%
Novice          4%
Beginner       6%
Intermediate 25%
Advanced      27%
Expert        15%
WorldClass     1%

 

This includes stars. Stars made up 0.65% of the total.

[keylime]

Uday,

 

That's an interesting spread of numbers. I'd be curious to see how the large populations (like US, Canada, Australia, etc et al) dispersed versus this baseline.

[uday]

messed up my 1st pass, will generate some other countries in a cpl hrs

[Fluffy]

A true rating should reflect how good players are.

It would reflect how much they cheat, how unpolitelly they reject undos, how eagerly they bore everyone to try to make impossible overtricks, and to a lesser extent how much concentration they have on bridge while they are watching a movie, or posting in some stupid forums.

 

Oh did I mention sudden disconects that ruind fun games?

[matmat]

you left out the mandatory count of the total numbers of '?' '!' and 'z' typed in. for example

 

LHO: ???????????????????????????????????????????????!

[finally17]

BUT ALL CAPS IS FINE, FEEL FREE TO TALK IN ALL CAPS.

[uday]

Breakdown for a couple of countries

 

USA

 

+-------+------------+

| level | percentage |

+-------+------------+

| priv | 10 |

| nov | 3 |

| beg | 6 |

| int | 36 |

| adv | 31 |

| exp | 13 |

| wc | 1 |

 

 

 

Italy

 

| level | percentage |

+-------+------------+

| priv | 19 |

| nov | 6 |

| beg | 6 |

| int | 23 |

| adv | 25 |

| exp | 20 |

| wc | 1 |

+-------+-----------+

 

 

 

Turkiye

 

+-------+------------+

| level | percentage |

+-------+------------+

| priv | 22 |

| nov | 5 |

| beg | 4 |

| int | 18 |

| adv | 29 |

| exp | 20 |

| wc | 1 |

+-------+------------+

[Al_U_Card]

What we really need is more expert "categories"....

 

wannabe

 

couldneverbe

 

mightsomedaybe

 

;)

[Rossoneri]

Glad to know I am one of the 4% who selected novice =)

 

I think most of my friends gave up on the rating system and just selected novice.

[jikl]

Uday,

 

That's an interesting spread of numbers. I'd be curious to see how the large populations (like US, Canada, Australia, etc et al) dispersed versus this baseline.

 

Since when did Australia qualify as a large population? :P

 

Sean

[foo]

I would rate expertise by measuring the average nr. of errors per board. In 20 boards, for instance,

 

<1 error: master (worthy of national team)

1: expert

2-3: adv

4-6: int

7+: newbie

Hmm, how about me then :) ?

 

I make around 30 mistakes / 20 boards, what is my ranking :)?

7+: newbie <_<

 

Not sure why we select you to represent Denmark <_<

 

Roland

Don't know why either, Roland.

 

But seriously, IMO is bridge a game of oodles of mistakes.

Part of the problem is that the definition of "mistake" is dependent on the expected caliber of player.

 

The novice forgets the meanings of bids, but sometimes so does the expert if it's a very rarely used convention. (But the expert will never forget an auction during the play.)

 

The novice doesn't signal correctly. Sometimes so does the expert, but the bad signal is far more subtle in the case of the expert's mistake.

 

The novice doesn't play a card combination correctly. The expert takes a line of play that turns out to be based on a logic mistake in assuming the layout of the board.

 

The novice doesn't play 2nd hand low or 3rd hand high. The expert makes a mistake and doesn't play 2nd high high or 3rd hand low when the situation called for it.

 

Etc. It isn't just how =many= mistake you make that determines your skill level, it's what =kind= of mistakes you make. I suspect most of the mistakes made by players representing their countries are subtle enough that the vast majority of players could not figure out they were mistakes it they were kibbing ATT ITRW (rather than Vugraph where everything is usually double dummy.)

 

Side note: A pet peeve of mine is "analysts" who think that because something is obvious double dummy on Vugraph it should be obvious single dummy ATT. Even worse are the "analysts" who Get It Wrong even though they are looking at it Double Dummy. *sigh*

[mikeh]

I posted a hand earlier today in the B/I section: the play in 4♠, after RHO opened 1♦.

 

North

 

K   A  Q  x

10 J  10  x

     9  9

     x  x

     x

 

South

A  10  -  A

J   x   -  K

9  x   -  x

8       -  x

7

6

 

After a low diamond, the 10 forces the K and you ruff. The best line, as posted by jdonn, is to cash one top club. cross to dummy, play a club to hand and then ruff a club with the trump K, then a diamond ruff and ruff the last club with the 10, thus guarding against RHO holding a stiff club (leading the 2nd club from dummy so he ruffs air if he ruffs) and also guarding against RHO holding 2 clubs and overruffing the 10 on the first round.

 

I neglected to lead the second club from dummy. So I made an error, but an error that few players would spot. In my defence, had RHO played a club honour on the 1st round, I am sure I would have woken up, and the odds of LHO not leading a club from QJ10xxx seems low.

 

But this goes to show why, in my view, experts tend to state that there are a lot of mistakes being made, while less-skilled players think that experts make almost no errors.

 

I think many players would not even see the need to ruff clubs in dummy, preferring to try to establish hearts, while those who do play for ruffs will often ruff with the 10 on the first ruff (altho this is almost certainly risk-free even if rho can overruff and return a trump... I leave the analysis to those interested)

[foo]

Thanks mikeh for illustrating my point so well.

 

Nice hand BTW! B)