[cherdano]

 fred, on 2012-April-26, 10:55, said:

Many years ago I wrote a program that basically did this (it made it fairly easy to define various types of end positions and then was able to recognize these positions when they were reached in a "bridge movie").

Maybe one day when I am bored I will try find the code (or write a new version of it) and think about ways that such functionality might be applied to various BBO contexts.

Fred Gitelman
Bridge Base Inc.
www.bridgebase.com

It's nice if it recognized stepping stone double-squeezes. But more importantly it should make a nice congratulatory beep when winning the last trick with the beer card!

[Zelandakh]

 Bbradley62, on 2012-April-26, 07:30, said:

While the rating of "Expert" is rubbish, what you should take from this is that it is time for you to stop playing primarily against B/I opposition.

I don't have much choice about this since I play in the Acol club where there are no experts. A few rate themselves as such but they are mostly even worse than those we see on a regular basis (and we do not allow them to blame their (pick-up) partners for their own mistakes so they typically leave upset after a while).

[diana_eva]

 Mbodell, on 2012-April-26, 17:35, said:

I actually think it is a reasonably useful service to put ratings on for people who like it. And it seems like it is trying to be modeled on more or less the chess elo scale.

Yes, maybe. I was mostly annoyed by the methods used to promote it on BBO. People ask yellows about it, not the creator who did not even put a "contact us" box there. From a BBOer's point of view it's probably fun to have this tool.

This post has been edited by diana_eva: 2012-April-27, 02:52

[hrothgar]

 cherdano, on 2012-April-26, 17:45, said:

It's nice if it recognized stepping stone double-squeezes. But more importantly it should make a nice congratulatory beep when winning the last trick with the beer card!

How could I have forgotten about the beer card...

[jjbrr]

wait, did i just see a leeroy jenkins reference on bbf? that's awesome.

[barmar]

 hrothgar, on 2012-April-26, 10:22, said:

Here's a silly idea:

Imagine if we could design an algorithm that could look at an (arbitrary) hand as it is being played and determine whether declarer had executed a squeeze, or an end play, or a strip and an end play, what have you.

This could also be the answer to people who think bridge scoring is unfair -- when you get fixed you get a bad score even though you did nothing wrong. The algorithm would determine for each pair whether they did something exceptional or wrong, and assign scores appropriately.

This is basically like par tournaments, but it needs to be automated to be useful for large-scale games.

[GMarco24]

Hi all,

As a creator of http://bboskill.com I want to explain some issues...
I can see you already have some info about it: it's not a BBO's tool, it uses BBO's MyHands and it's not perfect. And it would never be. More you think and solve one problem there'll be 2 new problems.

For now service has 3 obvious major problems:

1) Players who play with weak opponents are experts and world class
2) Really strong players who play only with friends, also strong players are advanced, intermediate or even lower than that
3) It gets only 1 month of period

1) and 2)
There is a correction based on partner's and opponent's strength. The only problem is I have really small database for now and I don't have at least one monthly result for each player. I think it would be more than rude to torture MyHands server with some scrap scripts, so I am only sending and storing requests that users make. Well, that takes time.
For correction I came up with this calculation:
(-partner+opp1+opp2)/3 for each board played (I don't think it's perfect and I need to test it). As I don't have many results they are 0,0,0 divided by 3 or 0,0,0.3 divided by 3 is only 0.1 and too many zeros lead to small correction. How would this work, time will show, but with good database correction will be big factor in calculation.
I've tried to protect top class players. They often play under 200 hands monthly and usually with friends and only several different players. If we imagine closed group of world class players, who only play with each other, someone will score under zero. There is no trick I can represent their real level and rating correctly now. Correction will do the job, but only if big part of them play with other, regular players and they score like 2.0 imp/board. So, I am making my list of famous players around the world. Some of them are on the list, some not yet. When I calculate correction against 'protected' players I take their real score if it's bigger than some my, imaginary number, and if it's lower I take that number as a score(like +1.0 and protected world class player can't go under that for calculation of corrections )

3)
I store all requests made by users. After couple of months I'll have some timeline and rating will be more accurate. Myhands won't allow you to take longer period than 1 months and that's it. I am simply limited to work with that.

Basically I am aware of all problems, and more and deeper you think even more problems you'll find. Just think about: GIB hands, TM ("Why is my fault when my team mates go down on 6♠ where they should claim easy 12?"), "I play only MP's. Is that transformation you make to imp correct?" and many others.
I look at this site this way:
-It's made for several days with very limited resources
-It's much more popular than I expected, so it's not complete "rubbish" as many love to say
-It's great for some purposes I thought it would be mostly used: e.g. discover "world class" player in individual tournament in front of you with -2.0 imp/board or make fun of your friend because he is only advanced and you are expert (Some of my friends already told me they stopped fooling around and they are trying to play more seriously. Some of them told me they feel more competitive now. For me they all look much more like 7♦ last days )

As long as people don't get it too seriously, there is no danger.

Marko Gligorijevic

[Leo LaSota]

 JLOGIC, on 2012-April-26, 11:00, said:

http://bboskill.com/user/rogerclee

clee is beating him over 736 hands. I am close but have only played 210.

Anyways, there is no doubt that leo lasota could beat his own number imo, he could just play slower I would guess. I am 100 % sure I could beat 62 % playing seriously, I would even wager on it and lay some odds if anyone wants! I kept track for a while when I played non drug affected ones, but I would like to try it officially (and possibly blog about it).

Also, Leo Lasota plays to win always I'm sure, which means swinging at the end. That alone hurts your average, not to mention tilting Playing for average in a bet would change those things.

Clee, if you're reading this just take 3 % and lets play 100 of them, don't be such a wimp!

Justin,

I promise you that I do not use strategies to maximize my % in the robot games. I just like to get more and more declarer hands in in a short period of time. If my goal was just to max my average %, I have no doubt that I could average 65%. I am naturally competitive at alot of stuff as you are, but having the best % average is not a competitve itch of mine. I have never even visited this rating site that everyone has talked about. For all I know, I may be considered a novice.

[JLOGIC]

I believe you! I think we should try some competition where we try to avg over 65, I have heard uday doubts its possible.

[Leo LaSota]

 JLOGIC, on 2012-May-01, 21:26, said:

I believe you! I think we should try some competition where we try to avg over 65, I have heard uday doubts its possible.

Not sure why Uday would doubt that it is possible. Many people (including Uday) find it hard to believe I win about 50% of the 18 board tourneys I enter as it is. If my goal was just to win the things, I would probably win about 80% of them.

[Cthulhu D]

 hrothgar, on 2012-April-26, 14:15, said:

All that I ask is that you have a "squoozen" badge...
Maybe something for taking too many dummy finesses

(I am a firm believer that silliness is a powerful motivating force)

Leeroy JENKINS!!!!!!!!!!!!!!

Shouldn't that be for going for 1100 in a phantom sacrifice?

Edit: Incidently, I would love Bridge Achievements.. and BBO needs steam integration. (Only slightly Joking about the second part).

[nathan2008]

 JLOGIC, on 2012-May-01, 21:26, said:

I believe you! I think we should try some competition where we try to avg over 65, I have heard uday doubts its possible.

That is why he is Ace. I heard that he was a star before, thanks bbo to remove his star and put Ace on him! Then we will put more attention when playing against him.

To Leo LaSota:

Although i am a fan of yours and u impress me a lot, i don't doubt u can get 65%+ if you want, but i doubt u can win 80% game . get 65%+ doesn't mean "win the game". Others are growing .

[chasetb]

First, after reading all 52 posts before mine, I think I am willing to challenge Justin. The stipulations would be that JLall would play 50 games of Robot Duplicate, the ACBL matchpoint ones. The AVERAGE of the 50 games would need to be 65.00% or better for him to win - the percentage might be negotiable. This wouldn't be his best 50, but rather 50 in a row, from a starting point to later be determined if this 'bet' goes through.

If Justin wins, then I pay him $50 to cover his expenses. If he fails, then what I really want is that at Philly, he and I would play in a single session event, as long as it didn't conflict with Spingold or another event he is getting paid to play in. I would prefer pairs, but if he wanted to get a good Swiss team, fine. I play Precision with 4-card Majors. Just kidding about the 4-card Major in 1st or 2nd seat, and I rarely do in 3rd or 4th. I would pay the entry fee for Justin, and would buy him a beer later that day just to have a conversation with someone who will be a force to reckon with for a LONG time (he still has untapped potential if you ask me). If that doesn't work out, it'll be something that won't be very expensive or painful for him.

Second, I really like the Achievement idea. I have 5 I would like to propose:
1.) Bid and make 7NT.
2.) Sacrifice with a 7-level contract and lose less points than a slam made by the other direction.
3.) Successfully pull off a squeeze (maybe should be more to it than that).
4.) In a contract that you or partner declare, get the same score (negative) that a pair sitting in the other direction and declaring a contract would have given you. An easy example is if you as North are in 1NT down 3 for -150, and an East or West player is making +150 in a contract that they declared.
5.) Set a contract declared by a star through some ingenious play, this might need to go through a committee though.

[fuburules3]

 chasetb, on 2012-May-06, 00:37, said:

If he fails, then what I really want is that at Philly, he and I would play in a single session event, as long as it didn't conflict with Spingold or another event he is getting paid to play in.

I think perhaps you should settle for a Midnight Knockout.

[babalu1997]

somehow i have greatly devalued the word expert since loitering at bbo.

last friday i was having an argument completely unrelated to bridge. my interloper had some weak argument to support her cause.

The matter did come to an end when she pronounced ( and it sounded rehearsed):

Having a master's degree in ********* and being an expert in ******, I determined that to be the correct course of action.


Ah. That explains it, it is like trying to explain to someone that 1nt-4hnt is not blackwood. I walked out, lol.

[Vampyr]

 babalu1997, on 2012-May-06, 12:46, said:

my interloper had some weak argument to support her cause.

Since the original opponent had needed to appeal to authority, she must have known that her own arguments were even weaker. So she probably knew and knows that she was wrong.

[dwar0123]

 GMarco24, on 2012-April-27, 10:52, said:

Hi all,

As a creator of http://bboskill.com I want to explain some issues...
I can see you already have some info about it: it's not a BBO's tool, it uses BBO's MyHands and it's not perfect. And it would never be. More you think and solve one problem there'll be 2 new problems.

Edit: Read faq, apparently there is some correction based on strength of partner/opponents.

Rogerclee's correction is pretty massive relative to a few others I looked at, which sounds right.

What about using robo tourny results as a way to rate people, not really exactly bridge, but certainty there is a direct relationship between the two and it would be pretty easy to call out that it is a robot tourny rating not a 'bridge playing' rating.

[barmar]

I'm a decent player, but not an expert in my opinion. But I think my rating in robot tourneys would significantly overstate me.

[dwar0123]

 Phil, on 2012-April-26, 08:38, said:

If you look at the stats of say, "Norby" for the last month, you'll see negative IMPs. For anyone else, this indicates a beginner. Even worse, I looked at the record for a friend of mine, and her classification was "like a babe lost in the woods" (wtf..?).

So I think that the designation 'professional' is plugged in to make the site appear more credible.

The bbo rating site does have norby with a very high rating, at the expert/world class level even without the artificial professional designation. I am sure there are flaws but just checking random people I've played against in the main club it actually seemed about right.

On the other hand, it has jec with a very low score, mostly due to a massive negative correction. He appears to have an artificial world class designation too, so I guess that is possible with both professional and world class Wonder which is higher in the creators opinion.

And really, given the number of hands that jec has played, it really should be accurate for him, so something is not right with the system yet.

[barmar]

I think rating systems like this have an inherent problem when you have subsets of players that generally play amongst themselves, but are mostly disconnected from other groups. With little overlap, you can't determine the relative strength of each group. And then you can't determine how the players from one group stack up against players in the other groups.

As an extreme example, suppose you have a group of experts who all play with each other, and a group of novices who do the same. Within the novice group, there may be a standout who regularly wins, so he would be given a high rating. Conversely, there must be an expert who isn't as good as his peers, so he would be given a low rating. But these two never play against each other, so the worst expert might be given a lower rating than the best novice, even though he's actually far better.

And for many players on BBO, this is a realistic situation. I don't think JEC enters many tourneys or plays with randoms in the MBC, he just plays his two JEC vs XXX team matches every day. He's a very good player, but he partners with some of the best, so the algorithm will assume that they're contributing more to the results.