Join Bridge Winners
All comments by Ping Hu
You are ignoring the author of this comment. Click to temporarily show the comment.
Peg, I agree with you player's character and behavior are all important for a good partnership. I would also like to add player's style and philosophy also matter a lot. For example I like to twist bidding system to make improvement, some of my partners don't like that. Some players might be too aggressive in bidding. I once had a partner who is a good player. However he sometimes like to make a risky bid. If this only happens in early round of Swiss team or a KO where we falls behind I would not mind. He did a few times at the point when we had a good round in team game, including last round of Swiss team and last board of a session in a Championship match. When I pointed out this is not the correct strategy and he disagreed, we could not keep our partnership.

However I agree with you completely there are a lot of intangible that makes a partnership working.
Dec. 30, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Peg,

You are absolutely correct that how good a fit of a partnership is most important. That is also my point #1.

If you started playing at a young age and has a lot of time to go to tournament, it is not a problem since you get to know a lot of players. For players like Richard and me, we don't have a lot of time to play many tournament. So we do have a lot of choices. In addition if we try to develop a partnership, we want to have a good chance to succeed. We don't have time to try with 10 different partners and find most of them would not work.

Let me give you an example. I looked for a partner for Providence NABC last year. So I contacted a player I met in summer Las Vegas NABC, we spent almost two months on BBO to practice our system and tried some online tournament. Eventually I realized we just could not play at the level I expected and had to call it off.

With a rating system it would be different. Even though I only had limited data now, this player (in my system) has played with 18 different partners and most of them rated between 1300 and 1600. I would not pick him as a candidate if I'm looking to get 1800+. While I spent time to practice with him, I missed the chance to try out a couple other partnership that might work out better.

Ping
Dec. 30, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Richard,

Chris's Power Rating is the best bridge rating data available in public currently. It could give you a good guidance in selecting partners. In fact I used it when I need to find a temporary partner at NABC.

The rating system I described in my thread is slightly different. The rating is based on pair not individual players. In addition my system is based on per board results so it gets much more granularity than other rating system. On the other hand, Chris has run his system many years and has larger data set so I think his data is good as a general guidance.

I absolutely agree with you that ACBL should implement a rating system. I'm planning to make a proposal next year to have a rating system on a volunteering basis. It is going to take a while to make it reality. ACBL needs to collect game level data even I could give them my program. However if a lot of players want it, it could happen.

Ping
Dec. 30, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Richard,

I share your experience and frustration. ACBL and most of its tournament has a partnership desk. However it mostly serves the purpose of getting a temporary partner. Although occasionally you get someone who could be a long term partner (I had one from my first NABC) but that is not norm.

What you look for is a long term partner and you want to have some success at national level. This would be much harder because you need a player who
1) share the same goal as you (success at national level) and willing to put in the work to develop the partnership;
2) has adequate playing skill to perform at the level you aim at.

Item 1 needs commitment not every player is willing to make. Item 2 is even harder to get. ACBL masterpoints is not an accurate measure of player's playing skill. Then you have to depend on others recommendation or your personal experience to find out the right candidates. This is not an easy task. The best way is to have an rating system I discussed in another thread. Based on my study, a pair winning a limited national championship (like Mini-Blue Ribbon) needs to have a rating of 1900-2100. Winning an open championship would need a rating of 2200+ with a performance of 2500+.

Ping
Dec. 30, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Auction part is best to go with an electronic solution like BBO, or an enhanced BridgeMate/BridgePad.
Dec. 15, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
In another word playing of a trick does NOT end when everyone turn their cards over.
Dec. 13, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
The definition of Established Revoke is important here. There could be situations under current rule offending site could gain advantage without being punished.
For example on a play, defender may “forget” his card that could win a trick and played another card. When declarer plays next card, he could find his and correct his revoke. Then his other card becomes a major penalty card and he has to play it. The net result is defending side get to see what declarer wants to play without suffering any penalties.
Dec. 13, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
What would happen if Established Revoke is defined as either side plays next trick or makes a claim?
Dec. 13, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
In Denver I saw a demo from one of manufacturers that used Smart Phone camera as card reader and gave direction where it should be dealt. They said it is a concept they are developing for home play or small club. It does not actually dealt the card but human need to deal the card without seeing what the card is.

It sounds like a very interesting idea and could make it very cheap because it is just an app for cell phone.
Dec. 10, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Chris, the director's ruling was 4 by East making 4 as I described in initial statement.

The score I listed here were from other tables (edited my statement to make it clear).
Dec. 7, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
FYI, the scores on this board from other tables were the following:
NS 4 making - NS 81%, EW 19%;
EW 4 making - NS 10%, EW 90%.

The director did not consider 3 by South in her ruling.
Dec. 7, 2015
Ping Hu edited this comment Dec. 7, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Yuan, I agree with you there was MI, but I don't completely agree with you there was damage.

If North remembered their agreement, he would bid 4 instead of pass. East would face 4 and not 3 passed to her. She probably would pass 4. With North forget their agreement, EW had a change to get a top score they would not have otherwise on this board but they did not capitalize it. So I'm not sure there was clearly a damage.
Dec. 7, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
It has some limited benefits. The better solution is to make the table like those game tables in Casino that has the mark where board should be placed and where card should be placed. So no player could use these positions and orientation to pass a message. The bidding tray also could have marks for where the bid should be placed so BZ's spacing could not work.
Nov. 12, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Why was 2000 boards chosen? We're there any studies to validate this number was better than others?
Nov. 5, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Polish 1 is considered as a artificial bid, not natural. The response 1 could be anything but other responses should be natural unless it shows GF value.

Because 1 is conventional, it is not allowed to psychic Polish 1 bid. I encountered such a case in Chicago NABC and filed a player's memo (I did not realized opponent psychic it until I looked at the hand record after the game).
Oct. 29, 2015
Ping Hu edited this comment Oct. 29, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
It looks like location matters most. In 1998 Spring NABC in Reno had more table counts than Summer (Chicago) and Fall (Orlando). The similar pattern repeated in 2004 (except New York replaced Chicago) and 2011 (summer in New Orleans).
Oct. 26, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
I filed a player memo during NABC but has not heard anything about it. I think the filer should be notified about status change or investigation result at least.
Oct. 14, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Certain information may still be useful. For example, IF the percentage for 0, 1, 2 and 4 matches hold for any distribution, their relative ratio (7%, 23%, 40%, 30%) will be same, assuming you don't have any information from bidding.

For example, in above cases you quoted, once you see your dummy's card, it has ruled out 4 match case. It still have 0, 1, 2 left and their relative ratio should be same.

However I doubt the general probability Barry gained from his simulation applies to a specific hand(s) like 3532 and 4333. I believe it would vary a lot for specific hand(s).
Oct. 14, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
We are constrained by each hand must have 13 cards (see my comment of Sudoku analogy). If your hand is 3532 and dummy is 4333, then Spade could not be 5332 since you get 4 cards. The other 3 suits could still be 5332. However the possibility only exist for 2, 1 and 0 suit having 5332.

You could apply the same logic from dummy point of view and conclude H and C could not be 4333.

However I think Barry is just simulating with random distribution, not a hand started with a particular distribution. It is the probability with specific distribution more useful.
Oct. 14, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Jean-Pierre,

I don't think his simulation is doing what you said. Look at my post right above this section about the analogy to Sudoku. I think he just ran a general simulation for random distributed hands. Then he count for a given set of row values (like 5-3-3-2) how many columns matches the same values. This is more academic.

I was trying to ask him what the result would be if it starts with a specific distribution like 4333, 4432 or 5530. These would be more practical. I doubt they are same as the general answer (7%, 23%, 40%, 30%).
Oct. 14, 2015
.

Bottom Home Top