[Dinarius]

How often is BBO Skill - you'll find it here http://bboskill.com/ - updated?

I'm aware it has nothing to do with BBO, but I thought someone here might know anyway.

Thanks.

D.

[1eyedjack]

Dinarius, on 2015-December-27, 12:38, said:

How often is BBO Skill - you'll find it here http://bboskill.com/ - updated?

I'm aware it has nothing to do with BBO, but I thought someone here might know anyway.

Thanks.

D.

The answer to your question is never. At least not any more. I cannot say when it was last updated but I think that may have been a couple of years ago. It was blocked well before you had to log into myhands (due to the traffic stress it was putting on BBO servers, not because it was statistically flawed, which I believe it was), but the login issue would have put a nail in the coffin even if it hadn't before.

To be frank I find it very surprising that the site is still up there. It is not generating any revenue and I suspect is costing someone. Just sews confusion now. I still see players online on BBO who extol it.

[Dinarius]

1eyedjack, on 2015-December-27, 12:47, said:

The answer to your question is never. At least not any more. I cannot say when it was last updated but I think that may have been a couple of years ago. It was blocked well before you had to log into myhands (due to the traffic stress it was putting on BBO servers, not because it was statistically flawed, which I believe it was), but the login issue would have put a nail in the coffin even if it hadn't before.

To be frank I find it very surprising that the site is still up there. It is not generating any revenue and I suspect is costing someone. Just sews confusion now. I still see players online on BBO who extol it.

Thanks for the prompt reply. I hadn't realized it was defunct.

So, is there any way of finding out how one is doing relative to one's opponents on BBO?

Thanks again.

D.

[1eyedjack]

With considerable difficulty. But then that is no different than were you to rely without justification on some spurious rating provided by BBOSkill.

There are some private clubs that keep their own stats.
If the individual is of the type to play in paid tournaments then you can view their BBO masterpoint history.
If you install the third party application Double Dummy Solver from Bridge Captain you can get some statistics about an individual's performance over the last 2-3 months (but excluding results from certain events to which you are denied access on MyHands website).

All of this is quite time consuming (but then BBO Skill was not instantaneous).

Best bet I think is to play a few hands with said individual and form your own conclusion based on personal experience.

[FrancesHinden]

1eyedjack, on 2015-December-27, 12:47, said:

. Just sews confusion now.

I thought that one sowed confusion in he sense of planting seeds. Sewing uses a needle and thread.

[1eyedjack]

FrancesHinden, on 2015-December-27, 16:05, said:

I thought that one sowed confusion in he sense of planting seeds. Sewing uses a needle and thread.

yup sow, schmow. It didn't look right, but you know, Port, snort.

[billw55]

Definitely ignore BBOskill. And preferably anyone who advocates it as well.

[Caitlynne]

It is no longer updated. Why, I don't know, but rumor has it (and this is a rumor that may or may not be based in reality) that BBO shut off the data.

[johnu]

billw55, on 2015-December-28, 07:21, said:

Definitely ignore BBOskill. And preferably anyone who advocates it as well.

BBOSkill may have a lot of theoretical issues, but it's certainly a better judge of skill than the BBO self assessment ratings where players can call themselves world class when they are barely intermediates. You would have to try actively game the system for an intermediate to get a world class rating in BBOSkill.

When BBOSkill was still active, I looked at some ratings for some of the top masterpoint winners in the ACBL (robot) tournaments, and IIRC, they ranged from the 2200's to 2500's which is in the expert to world class level. I also looked at the ratings for some players who didn't appear to be experts, and their scores were mostly in the 1200-1800 range, and some of them self-rated as experts or even world class. I will admit that I saw a player who said he wasn't an expert with a score in the 2800's !!! (I never saw anybody in the 2600's or 2700's) but it looked like he only played against robots so his competition was players who played against robots who presumably are almost all beginners or low intermediates so he was killing the competition, but the competition wasn't any good.

[PhantomSac]

johnu, on 2015-December-28, 13:02, said:

(I never saw anybody in the 2600's or 2700's)

clee was over 2600 with like a 10k hand sample iirc (I am more amazed he played so many hands on BBO than with his score honestly lol).

Anyways I agree with you that it is far more accurate than self ratings.

[1eyedjack]

Nobody pays any credence to the self-rating, and as such there is no danger of being misled.

But there were those who regarded BBOSkill as sacrosanct. So even if BBOSkill algorithm had some merit it presented the greater danger of misleading.

I don't think that its accuracy or inaccuracy played a part in its being shut down. My understanding is that it was purely down to the load placed on the servers. Maybe someone who really knows can comment? If so, I would have thought that some sort of limited access to the server would be achievable. Ie only one request permitted by a specified ID per month. If the will were there. Perhaps the will isn't there, after looking at the way that the OKB rating system went.

[Zelandakh]

johnu, on 2015-December-28, 13:02, said:

BBOSkill may have a lot of theoretical issues, but it's certainly a better judge of skill than the BBO self assessment ratings where players can call themselves world class when they are barely intermediates. You would have to try actively game the system for an intermediate to get a world class rating in BBOSkill.

I spent most of the time rated as an Expert by BBOSkill and am pretty sure the rating would have been World Class when I was playing my strong club system regularly. I am most certainly an intermediate and did not game the system at all. On the other hand, Han was typically rated between Beginner and Intermediate despite being a much stronger player. Without any real adjustment being made for quality of opponents, BBOSkill was pretty useless. Most of the time, a simple check of the last month's results give just as accurate an assessment with the additional advantage of being able to check who the opponents were.

[nullve]

Zelandakh, on 2015-December-28, 16:17, said:

Most of the time, a simple check of the last month's results give just as accurate an assessment with the additional advantage of being able to check who the opponents were.

And how do you determine their skill level without a rating system?

[billw55]

nullve, on 2015-December-29, 09:57, said:

And how do you determine their skill level without a rating system?

The bottom line is that individual ratings just don't work very well for bridge. The smallest unit of contestant is a pair, and hence that is the smallest unit that can be effectively rated.

[blackshoe]

I don't know what "unit of contestant" means, but the contestant is, in an individual, one player, in a pairs game, one pair, and in a teams game two or more pairs, possibly including a non-playing captain.

[nullve]

blackshoe, on 2015-December-29, 17:40, said:

I don't know what "unit of contestant" means, but the contestant is, in an individual, one player, in a pairs game, one pair, and in a teams game two or more pairs, possibly including a non-playing captain.

So why not keep separate rankings for each category, if necessary? Look at chess: there are separate rankings for standard, rapid and blitz. (See e.g. https://ratings.fide.com/toplist.phtml)

This post has been edited by nullve: 2015-December-29, 20:16

[helene_t]

Why would you want ratings of pairs? For selecting compentent opps for a training match? If you are that serious about it you'd probably play against opps that you know, and then you don't need their objective rating. For all you know, opps with high rating may be cheaters, or just rude or slow.

Actually, if a pair has a much higher rating than the sum of their individuals ratings suggest, they may well be cheaters.

The whole idea of calculating ratings based on human-vs-human online play is silly. I can understand why the idea appeals to some people but I also think that most of the bboskill users would acknowledge how silly it is if they thought deeply enough about it.

For those who want to monitor their own progress (without necesarilly making the information available to others), there is always the possibility to play against robots and calculate your roling average IMP vs PAR or some such. Maybe BBO should facilitate that. But I can understand that they don't since some players would become disillusioned when they see how bad they are (or, alternatively, become convinced that the robots cheat or the dealer is biased against them) and then leave BBO.

[nullve]

helene_t, on 2016-January-05, 05:38, said:

Why would you want ratings of pairs?

Partly because bridge can naturally be viewed as a two-player game with each pair being one player. In the words of von Neumann & Morgenstern:

Quote

Although Bridge is played by 4 persons, to be denoted by A,B,C,D, it should be classified as a two-person game. Indeed, A and C form a combination which is more than a voluntary coalition, and so do B and D. For A to cooperate with B (or D) instead of with C would be "cheating," in the same sense in which it would be "cheating" to look into B's cards or failing to follow suit during the play. I.e. it would be a violation of the rules of the game. If three (or more) persons play poker, then it is perfectly permissible for two (or more) of them to cooperate against another player when their interests are parallell - but in Bridge A and C (and similarly B and D) must cooperate, while A and B are forbidden to cooperate. The natural way to describe this conists in declaring that A and C are really one player 1, and that B and D are really one player 2. Or, equivalently: Bridge is a two-person game, but the two players 1 and 2 do not play it themselves. 1 acts through two representatives A and C and 2 through representatives B and D. [von Neumann & Morgenstern, "Theory of Games and Economic Behaviour", p. 53]

And why shouldn't pairs be interested in finding out how good they are as single players at this 2-player game? Why does bridge have to be different from chess and other ELO-rated 2-player games in this respect?

[helene_t]

nullve, on 2016-January-05, 06:51, said:

And why shouldn't pairs be interested in finding out how good they are as single players at this 2-player game? Why does bridge have to be different from chess and other ELO-rated 2-player games in this respect?

ELO compares your skill level to the bulk of human IRL chess players. This is meaningful, and I suppose it is useful for chess competition since it is not so funny to play chess against someone with a very different skill level. And ELO is based on realistic competion where you don't have players who leave halfway through the board, who start with a handicap, who request dubious undos etc etc.

I don't see how these criteria apply to online bridge.

But by all means, if you want to monitor your progress as a pair, just play against robots and compare to PAR.

[1eyedjack]

helene_t, on 2016-January-05, 06:59, said:

But by all means, if you want to monitor your progress as a pair, just play against robots and compare to PAR.

It may not be necessary to play against robots. If you download your monthly playing history into Double Dummy Solver, run the DD computation on every hand, it will provide a summary of par comparisons as well as (perhaps less meaningful) actual comparisons.

Keeping track of your own performance is not so hard, but requires some effort and is not instantaneous. What is harder is getting an instant assessment of the value of another player at your table. That is where most players who valued BBOSkill highly placed its value. Justifiably or otherwise, and without commenting on whether the BBO society would benefit from having that information publicly available.