[Hanoi5]

I have often wondered why isn't there an ELO or ELO like rating of bridge players. OkBridge has the Lehman, there is none at bridgebase (except for the card that appears next to the name which shows how good you've done in certain tourneys).

So, does anyone have the Lehman formula? and what aspects do you think should be taken into account to create such a rating in real life events? Butler's in team tournaments use to be very good, but I'm sure something better can be done.

Your ideas?

[matmat]

ah. this again.


this topic creeps up with stunning regularity. stunning!

[mtvesuvius]

Didn't we just finish discussing this a few weeks ago?

The most recent "Ratings" thread

[jdonn]

mtvesuvius, on Feb 21 2009, 06:58 PM, said:

Didn't we just finish discussing this a few weeks ago?

And a few weeks, months, and years before that.

[Hanoi5]

Link me to it, please.

[jdonn]

I'm bad at searching on the forums, but I'll bet a savvy person could find at least 5 such threads and maybe even 10 if they went back far enough.

[hotShot]

1) I think it's possible to create an ELO equivalent rating for partnerships. But how will you combine 2 pair ELO's to a team ELO or split a pair ELO into 2 player ELO's?
(This problem is inherent in the Lehman rating as well, that's why playing with a weak partner will ruin your rating.)

2) A rating system should not reward "bunny bashing". (The Lehman rating does that.)

3) Prior to investing time into that subject, ask yourself: who would want to have a true rating?
If my ELO said I suck in chess, I might find other ways to waste my time, having more fun.
So maybe a working rating would drive people away from bridge.
If I were a "BBO Expert" or a star from "a country where its easy to be nominated to international tournaments because nobody else is interested/willing/able do go" I would not want to know my rating or that anybody else know my rating.

4) We have discussed the social implications of a rating system on BBO in ### threads over the years, and they are not welcome.

5) I have experimented with a rating system based on the card play abilities of players, this could be a way to rate individual player.

[matmat]

http://forums.bridgebase.com/index.php?act...highlite=rating

for a start

[mtvesuvius]

Hanoi5, on Feb 21 2009, 07:13 PM, said:

Link me to it, please.

Not all these may be 100% relavant... But they are what I found so far:
http://forums.bridgebase.com/index.php?sho...=28339&st=0&hl=
http://forums.bridgebase.com/index.php?sho...=27945&st=0&hl=
http://forums.bridge...topic=23673&hl=
http://forums.bridgebase.com/index.php?sho...=22709&st=0&hl=
http://forums.bridge...topic=22775&hl=
http://forums.bridgebase.com/index.php?sho...27445&hl=rating

These... In addition to Matmat's links

[H_KARLUK]

http://wiki.classicalfencing.us/index.php5...c:Rating_system

http://www.nasamw.or...ko_overview.htm

After our offer we are promised by admins on another web based bridge site in 1998 that they will contact with that notable Dr.

Then a year passed and unfortunately nothing changed. So I visited Science of Maths branch in my city's biggest University.

Coincidentially there were a small group of bridge players working as Academicians. After a month I am told it's also suitable to apply a new ranking system for Bridge scoring like applied in Chess tournaments.

But it seems to me still mostly American methods are on. Maybe th reason that they are skilled to use servers. ( Tho I think naginx was a Russian's invention).

ps. Most Chess federations changed their rating sys to Glicko-2 etc. I m pretty sure they r happy.

[shevek]

Try

http://www.australia.../login_gsdb.php

& click on the two ratings links bottom right

[hrothgar]

This is probably controversial, however...

If anyoe wants to revise the ratings system, I'd strongly advise them to start by coming up with a ratings systems to evaluate the performance of pairs. Once there is a system in place to evaluate pairs, this system can be extended to evaluate individuals.

In my mind, this implementation scheme has a couple major advantages (as well as one obvious problem)

1. The first problem is that there isn't any kind of widely deployed rating system for pairs. You don't need to worry about replacing the masterpoint system. You're supplementing it... This sidesteps a number of political / marketing issues.

2. Conceptually, its much simplier to rate pairs than individuals.

The problem is one of combinatorics. There 's a lot of bridge players out there. The number of possible pairs of bridge players is MUCH bigger. However, lots of bridge players tend to play in well established partnerships which should make this more manageable...

[hotShot]

hrothgar, on Feb 22 2009, 02:02 PM, said:

1. The first problem is
......

what kind of results carry enough meaning to be used.

Casual play in the MBC with randomly changing partners and opps, 4-12 board (individual) tourneys or # board team matches