BBO Discussion Forums: Decision Point tournaments - BBO Discussion Forums

Jump to content

  • 2 Pages +
  • 1
  • 2
  • You cannot start a new topic
  • You cannot reply to this topic

Decision Point tournaments Focused skill instead of random luck

#1 User is offline   silvr bull 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 253
  • Joined: 2012-November-11

Posted 2017-February-06, 03:22

I propose a new tournament format called Decision Point (DP). The objective is to provide all players with the same information and circumstances and excellent opponents, and to compare how well each player does at selected DPs. Every player will sit in the same position in pre defined hands that were created for the DP tournament. A player's three "opponents" will not be robots (which can do different things in different situations), but will simply be scripts which present exactly the same bids or plays for every player to work with.

The script requires each player's bids or plays to be the same also. That will be accomplished by having DPs throughout the bid and play of each hand. After a player selects a bid or play, the "correct" or scripted bid or play will be presented on screen instead of the one selected by the player. If that bid or play was a DP, then it will be scored as a maximum plus (if it duplicates the DP script) or a reduced amount (if it is significantly different than the DP script) for the player. Every player will play the same contracts under the same conditions with the same bridge scores, so the cards presented to the player and the opponents will not affect the player's tournament result. The DP tournament winner will be the one who chooses the most correct DPs.

Here is an example hand actually played in a BBO tournament. After two passes, the player has the first bid decision.



Some players might prefer a pass, or 1S, or 3S, or 4S. If the player selects any of those actions, a "Decision Point Adjust" (DPA) alert message will flash on the screen. The player's bid will be changed to 2S (the same bid for all the tournament players), and the player's DP score would be the fractional DP point value assigned to the player's bid. If the player selected 2S, there would be no alert message, but the player's DP score would be increased by one DP point. The script would continue with 4H, P,P back to the player. Any bid other than 4S will result in a DPA message, and the player's tournament score would increase by the fractional DP point value assigned to the player's bid, but a 4S bid will increase the player's tournament score by one DP. The bidding will complete with X, P, P, P.

The opening lead is the club Ace, and the player is declarer with this hand:



The play continues club ace, 2, 6. If the player plays the jack, or ten, or nine, there will be no alert and the player's DP score will increase by one DP point. If the player plays the 5, however, there will be a DPA alert, the player's card will be converted to the jack, and the player's DP score will reduce by one DP point. West next leads the diamond king to the ace. The trump 9 and the 2 from East gives the player another DP opportunity. If the top trump is played, there will be no alert and the player's DP score will increase by one DP point. If the player plays low or the 10, however, there will be a DPA alert, the player's card will be converted to the ace, and the player's DP score will reduce by one DP point. The player will pull trump and can play as many spades as he likes before he switches to clubs. If the player next leads a top club to the king, there will be no alert and the player's DP score will increase by one DP point. If the player leads the 5, however, there will be a DPA alert, the player's card will be converted to the ten, and the player's DP score will reduce by one DP point. This simple hand offers 5 DP points to each player.

The Decision Point tournament will eliminate many forms of cheating online, but there could be one worth noting. A player could enter the tournament under two IDs. Each would see the same information, but if one could play fast to quickly cause a DPA, then the slow player would know what trap is lurking. A way to avoid that is to delay each player's bid or play by a set time (such as 15 seconds) and complete all players' actions at the same time. Alternatively, easy actions, such as pass with little value in distribution or HCP or plays with only low spot cards, could be allowed less time, but more difficult DP actions could be allotted more time. Either way, the action of every player would complete at the same time, so there would be no advantage from using multiple IDs in a DP tournament. If a player does not act quickly enough within the time allotted for a DP, then the system would make the correct decision and the player's tournament score would not increase or decrease from that DP.

Decision Point tournaments would transform bridge results from random luck comparable to backgammon, to more focused skill like chess. BBO would be a great place to begin Decision Point tournaments!
0

#2 User is offline   sfi 

  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 2,576
  • Joined: 2009-May-18
  • Location:Oz

Posted 2017-February-06, 05:28

This is very similar to par contests which have gone out of fashion over the past couple of decades. One major problem with this type of contest is the complexity with setting it up and determining what is the "right" bid or play.

For instance, you would get a lot of argument on many of your points in the sample hand:

- Opening 2S is not at all obvious. 1S and 3S both have a lot going for them, and you have to cater for different styles when weighting bids.
- Bidding 4S after opening 2S is very poor and should have points deducted IMO.
- Unblocking a club is clearly good technique.
- Winning the DA looks to be inferior to ducking one round, but it's hard to judge by how much when assigning scores.
- The right spade play is not clear, but I think the finesse is more likely to gain than playing the top honours. Again, it's not clear how you determine how much better one play is than another.

Imagine trying to go through all the hands for a tournament and making the scores for each action fair. It is going to be a lot of work for a very good player, which is why I haven't seen any par tournament for over 20 years.
1

#3 User is offline   Vampyr 

  • PipPipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 10,611
  • Joined: 2009-September-15
  • Gender:Female
  • Location:London

Posted 2017-February-06, 10:01

Agree with all of the above.

The "correct" opening bid of 2 is very curious. I do not think that weak-to-intermediate is very popular.
I know not with what weapons World War III will be fought, but World War IV will be fought with sticks and stones -- Albert Einstein
0

#4 User is offline   barmar 

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • Group: Admin
  • Posts: 21,398
  • Joined: 2004-August-21
  • Gender:Male

Posted 2017-February-06, 10:06

Perhaps a better way to run a contest like this is with BridgeMaster-type hands. The players would simply be scored based on the number of hands that they get right.

But this still requires lots of work to create all the hands. We can't use the existing BM corpus, because there are lots of people who have already played them, so someone would have to come up with new sets.

#5 User is online   mycroft 

  • Secretary Bird
  • PipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 7,056
  • Joined: 2003-July-12
  • Gender:Male
  • Location:Calgary, D18; Chapala, D16

Posted 2017-February-06, 10:26

I do like the idea, and I agree with all the objections.

How many times have you seen bidding polls with "I abstain. The stupid decision to make [previous action] has caused this problem, as any reasonable player would have known it would, which is why I would have [done other action]." (well, likely with not quite as much snark, but my point stands).

But it would be fun, even in a "so, the previous player spilled his coffee all over himself when that call got pushed through the screen, and you have to take it from here." way.
When I go to sea, don't fear for me, Fear For The Storm -- Birdie and the Swansong (tSCoSI)
0

#6 User is offline   Vampyr 

  • PipPipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 10,611
  • Joined: 2009-September-15
  • Gender:Female
  • Location:London

Posted 2017-February-06, 10:43

View Postsfi, on 2017-February-06, 05:28, said:

This is very similar to par contests which have gone out of fashion over the past couple of decades. One major problem with this type of contest is the complexity with setting it up and determining what is the "right" bid or play.

For instance, you would get a lot of argument on many of your points in the sample hand:

- Opening 2S is not at all obvious. 1S and 3S both have a lot going for them, and you have to cater for different styles when weighting bids.
- Bidding 4S after opening 2S is very poor and should have points deducted IMO.
- Unblocking a club is clearly good technique.
- Winning the DA looks to be inferior to ducking one round, but it's hard to judge by how much when assigning scores.
- The right spade play is not clear, but I think the finesse is more likely to gain than playing the top honours. Again, it's not clear how you determine how much better one play is than another.

Imagine trying to go through all the hands for a tournament and making the scores for each action fair. It is going to be a lot of work for a very good player, which is why I haven't seen any par tournament for over 20 years.


I enjoyed par contests, and the difference is that with par contests, there was normally just one thing you had to get right.
I know not with what weapons World War III will be fought, but World War IV will be fought with sticks and stones -- Albert Einstein
0

#7 User is offline   silvr bull 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 253
  • Joined: 2012-November-11

Posted 2017-February-10, 11:19

After the above critical comments about my example, I created a thread to ask the key questions (http://www.bridgebas...play-in-part-3/). The comments there were even more critical, so it seems that I could have selected a better example hand for this Decision Point proposal. Needless to say, any additional comments about my example hand should be put into that thread about the hand, and not here where they distract from the Decision Point concept.

The basic premise of Decision Point is valid, and the idea of giving every player identical information throughout the bidding and play would substantially reduce the amount of luck (good or bad) that is an inevitable part of current tournament formats. BBO would be an excellent venue for hosting Decision Point tournaments, and I hope BBO will give serious consideration to the Decision Point approach.
0

#8 User is offline   Vampyr 

  • PipPipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 10,611
  • Joined: 2009-September-15
  • Gender:Female
  • Location:London

Posted 2017-February-10, 11:50

A lot of the comments on your example hand highlighted the difficulties with your idea. Who would make up these hands and ensure that the decisions were "right"? I seem to remember that Hugh Kelsey wrote books like this. I think that they were a lot of work, and the books were just one match.
I know not with what weapons World War III will be fought, but World War IV will be fought with sticks and stones -- Albert Einstein
0

#9 User is offline   silvr bull 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 253
  • Joined: 2012-November-11

Posted 2017-February-11, 01:36

Hands would be easy to construct for Decision Point. The free daylong tournaments are full of hands that are double dummy to bid and play (but with single dummy vision), and many other hands on BBO are similarly exacting. Any good player with double dummy vision can identify the key spots in bid and play that would be decision points (systemic bids comparable to the robots now and critical plays like hold ups, avoidance, safety plays, squeezes, end plays, etc.), and assign values for each (with close decisions getting close Decision Point scores). In my example hand, the opening bids of 2S or 4S would be very close in value, and it would not make much difference which was assigned a fractionally higher value. The "point" of Decision Point is not to identify perfection, which would be impossible since many players will disagree about the fine point details of almost everything. Instead of bridge perfection, Decision Point will reward good bridge technique, and (most importantly) eliminate luck. The winners of any current bridge event will be good players who were also very lucky, with opps who did strange things or with guesses that worked out well. Decision Point, in contrast, will eliminate variations in opps or partner (since the script for those will be identical), and focus completely on the actions of the player when presented with exactly the same information as all the other players his Decision Point score will be compared against.

The only difficulty I see in creating Decision Point tournaments is for the software to be set up to enable them. My guess is that BBO could create that software without much difficulty. I hope BBO will rise to that challenge and opportunity.
0

#10 User is offline   manudude03 

  • - - A AKQJT9876543
  • PipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 2,610
  • Joined: 2007-October-02
  • Gender:Male

Posted 2017-February-15, 05:30

While it would be interesting to see, I think it would just be too impractical to set up. Who decides what the best action at a particular point in time is or how to score each decision? If it's the GIBs, then you may not be eliminating the random luck factor and if it's a human, it will take too long to set up. That's not to mention style issues.
Wayne Somerville
0

#11 User is offline   hrothgar 

  • PipPipPipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 15,372
  • Joined: 2003-February-13
  • Gender:Male
  • Location:Natick, MA
  • Interests:Travel
    Cooking
    Brewing
    Hiking

Posted 2017-February-15, 05:42

View Postsilvr bull, on 2017-February-11, 01:36, said:

Hands would be easy to construct for Decision Point.


It is great to hear this.

Given that this is all so simple, why don't you sit down and generate 36 hands that you feel would be good candidates for decision point tournaments.
Post them here. (Also let us know how long it took to generate the set of hands)

I guarantee that folks on the forums will take a gander and let you know what they think...

If it really turns out to be simple to create obvious, non controversial hands then it would make sense to have follow on discussions...
Alderaan delenda est
1

#12 User is offline   silvr bull 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 253
  • Joined: 2012-November-11

Posted 2017-February-16, 01:08

Daylong tournaments I played are 8 boards that I completed in little more than 20 minutes. (No, I did not score well, but thank you for asking.) In 20 minutes, a player better than me could review those 8 boards in double dummy, identify several decision points in each hand, and script the "correct" bid and play sequence. If 50 decision points are identified for the 8 boards, and values of 0-2 were assigned to the key possibilities, then a half hour would be enough time to convert the hands in a Daylong into a Decision Point tournament that would be ready to play. That requires the software to be developed to enable Decision Point, of course, but encouraging BBO to develop that software is precisely the point of this thread.

So, why not just play the Daylong with robots instead of Decision Point? Several reasons:

(1) There are many guesses that one must make (I would call them decision points) to do well in each hand. One unlucky guess could doom a hand (which is 12.5% of the tournament) to average or worse, and would drag one's score down substantially. In contrast, a miss guess in Decision Point would only affect the score in one of fifty decision points, and the "correct" bid or play would be presented to all players for the remaining challenges in that hand.

(2) Every player would be presented with exactly the same information throughout the bid and play for every hand. Every player would have an equal opportunity to do well at the next decision point regardless of earlier misjudgments.

(3) Robots are not as good as one would like for tournament play. A small and possibly irrelevant change in a human's bid or play can produce significantly different reactions in the robots. Robot variations are an added luck factor that would not cause a problem with Decision Point, where every bid and play will be exactly the same for every player.

(4) Wild bids would not be rewarded in Decision Point. A Daylong (IMPs) I played, North (robot) opened 1D and the human South's only bid was 6NT with AKQTx Kx ATx xxx! That wild bid did produce a top score, in much the same way that a psych bid can result in a top sometimes, but I prefer to play in events that reward skill and minimize the effects of luck.

(5) Decision Point tournament results can easily be compared with other players, because every player would play exactly the same hands in exactly the same way (with score differences only in the number of correct decision points) and at the same time. Daylong, in contrast, has different sets of hands for different players. No doubt, that is done to prevent a player playing the hands in one ID and then replaying them double dummy in a different ID, but that would not be a problem with Decision Point. The hands would be exactly the same for all participants, so the decisions each made are directly comparable.

My signature line is a good closing comment:
0

#13 User is offline   hrothgar 

  • PipPipPipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 15,372
  • Joined: 2003-February-13
  • Gender:Male
  • Location:Natick, MA
  • Interests:Travel
    Cooking
    Brewing
    Hiking

Posted 2017-February-16, 03:21

View Postsilvr bull, on 2017-February-16, 01:08, said:

Daylong tournaments I played are 8 boards that I completed in little more than 20 minutes. (No, I did not score well, but thank you for asking.) In 20 minutes, a player better than me could review those 8 boards in double dummy, identify several decision points in each hand, and script the "correct" bid and play sequence. If 50 decision points are identified for the 8 boards, and values of 0-2 were assigned to the key possibilities, then a half hour would be enough time to convert the hands in a Daylong into a Decision Point tournament that would be ready to play. That requires the software to be developed to enable Decision Point, of course, but encouraging BBO to develop that software is precisely the point of this thread.


Even if you aren't qualified to do this work yourself, it should be simple enough to find someone who you can pay to create a set of 36 hands.
(That is, if this is all as easy as you say it is)

Right now, you are insisting that other people should go and do a bunch of work to support your pet project.

It is rare that this happens without your first putting some skin into the game.

If you are unwilling to take the most rudimentary efforts to bring your scheme to fruition, why should other people invest their own time and effort?
Alderaan delenda est
0

#14 User is offline   diana_eva 

  • PipPipPipPipPipPipPip
  • Group: Admin
  • Posts: 4,854
  • Joined: 2009-July-26
  • Gender:Female
  • Location:bucharest / romania

Posted 2017-February-16, 03:37

I still don't understand how this is supposed to work. What if a player opens 1D instead of 1C, and the decision point thing says responder is supposed to respond 1D? It just can't be identical, each time the human player does something different the bidding and play will have to change accordingly. Or is everyone given the same bidding and they just play like in Bridge Master?

#15 User is offline   Vampyr 

  • PipPipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 10,611
  • Joined: 2009-September-15
  • Gender:Female
  • Location:London

Posted 2017-February-16, 04:44

Most people do not play eight boards in 20 minutes, and analysing eight boards, then creating a quiz for them would take a considerably longer time. Also boards would first have to be identified which had a clear eg bidding and declarer play or defence problem. The ACBL bulletin has a relative wealth of quiz-type problems, but these are created on a monthly basis. How many play or defence problems does Eddie Kantar produce per month?

It seems these hands would also suffer the same problem as Autobridge -- you would have to do things in the right "order" even if the order is irrelevant.

For many reasons, your idea would work better in a simple quiz format. This still does not mean that it would be simple to create.
I know not with what weapons World War III will be fought, but World War IV will be fought with sticks and stones -- Albert Einstein
0

#16 User is offline   nige1 

  • 5-level belongs to me
  • PipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 9,128
  • Joined: 2004-August-30
  • Gender:Male
  • Location:Glasgow Scotland
  • Interests:Poems Computers

Posted 2017-February-16, 05:14

View PostVampyr, on 2017-February-06, 10:43, said:

I enjoyed par contests, and the difference is that with par contests, there was normally just one thing you had to get right.

Like vampyr, I enjoyed Par Contests. Great idea silvr bull! :)

View PostVampyr, on 2017-February-10, 11:50, said:

A lot of the comments on your example hand highlighted the difficulties with your idea. Who would make up these hands and ensure that the decisions were "right"? I seem to remember that Hugh Kelsey wrote books like this. I think that they were a lot of work, and the books were just one match.

BBO programmers have already done most of the work :) The Bridge World, other magazines, books and BBO itself, all feature Bridge Movies which embody this idea. Sometimes the right action is a matter of judgement and opinion. But it would still be fun. :) Critics would not be forced to participate :(
0

#17 User is offline   silvr bull 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 253
  • Joined: 2012-November-11

Posted 2017-February-16, 08:00

View Postdiana_eva, on 2017-February-16, 03:37, said:

I still don't understand how this is supposed to work. What if a player opens 1D instead of 1C, and the decision point thing says responder is supposed to respond 1D? It just can't be identical, each time the human player does something different the bidding and play will have to change accordingly. Or is everyone given the same bidding and they just play like in Bridge Master?

The fundamental idea of Decision Point is that at every turn to bid or play, the player will have seen exactly the same information as every other player at other tables. A script is prepared in advance for each hand, and that script is the exact bid and play sequence for that hand. The player can try to make a different bid or play, but the tournament software will change the player's bid or card to match the script. For bids or plays the hand developer selected as decision points, the player will receive the maximum points if the bid or play he selected is the same as the script. If the players choice is different than the script, then the player may receive a smaller decision point value than the maximum, and the software will replace the player's choice with the one in the script, so the bids and plays will be identical at every table. The only difference will be the decision point scores that the player accumulates during the hands

Maybe another (and very simple) example will help to clarify the Decision Point approach. Suppose E-W bid to 6S. After the scripted opps pull trump and lead toward a dummy that still has only trump and KQJ in the suit just led. The South player will have to decide if he will play low or the A of that suit. If he plays the ace, then play will continue as planned in the script. If he tries to play low instead, the software will flash a Decision Point Alert (DPA) message, and then the software will play the ace, just the same way it is played at every other table. The player will be credited with the maximum decision point value assigned to that play (if he plays the ace), or to zero for that simple decision point (if he plays low).

Your question is what if a player decides to open anything other than the bid (1C in your example) in the script. Then the software would change the player's bid to the one in the script, so that the situation at that table will be exactly the same as it is at all the other tables. Suppose that the player's hand is xx xx AKJx AKJxx and he is first to bid. If I set up the Decision Point script for this hand, I would select an opening bid of 1C as the bid that the software will actually make at all the tables. I view a 1NT opening with nothing in both majors as less than best, and a player who tried to open with 1NT would receive a DPA. The software would then replace the 1NT the player tried to bid with 1C, and later he would receive less than the maximum decision points for trying to open 1NT. If the player tried to open 1D, his bid would also be replaced with 1C so that all bids and plays are the same at all tables. However, there has been a long running disagreement about whether it is better to open 1C (and possibly be forced to rebid clubs) or to open 1D with the flexibility to rebid Cs later. I would assign the same decision point value to either a 1C or 1D open, but in any case, the software would replace a different bid with the 1C open that is in the script, so that all tables do the same thing at all times.

Similarly during the play, some spots will be decision points, but other spots may be irrelevant. Suppose a player on defense holds the 432 of a suit that declarer leads, in a situation where it makes no difference which card the player plays. Then the script might call for the 2, and a different card the player tried to play would be changed to the 2 (so that all tables are exactly the same), but there would be no decision point and no affect on the player's Decision Point score regardless of which spot card he tried to play.

Current tournament formats allow different information to be presented to a player, and those differences compound through the play of a hand so that the final results are often skewed as much by luck as by skill. The objective of Decision Point is to make all of the bid and the play information exactly the same, so that the player at every table will have the same decisions to make under the identical circumstances. Luck will be kept to a minimum, and skill will be reflected in the players' accumulated decision points. The winner will be the player who best selects the bids or plays at each of the decision points throughout the hand.
0

#18 User is offline   hrothgar 

  • PipPipPipPipPipPipPipPipPipPipPip
  • Group: Advanced Members
  • Posts: 15,372
  • Joined: 2003-February-13
  • Gender:Male
  • Location:Natick, MA
  • Interests:Travel
    Cooking
    Brewing
    Hiking

Posted 2017-February-16, 08:31

Sounds very much like Bridge Master...
Alderaan delenda est
0

#19 User is offline   barmar 

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • Group: Admin
  • Posts: 21,398
  • Joined: 2004-August-21
  • Gender:Male

Posted 2017-February-16, 09:29

View Postsilvr bull, on 2017-February-16, 01:08, said:

Daylong tournaments I played are 8 boards that I completed in little more than 20 minutes. (No, I did not score well, but thank you for asking.) In 20 minutes, a player better than me could review those 8 boards in double dummy, identify several decision points in each hand, and script the "correct" bid and play sequence. If 50 decision points are identified for the 8 boards, and values of 0-2 were assigned to the key possibilities, then a half hour would be enough time to convert the hands in a Daylong into a Decision Point tournament that would be ready to play. That requires the software to be developed to enable Decision Point, of course, but encouraging BBO to develop that software is precisely the point of this thread.

The problem with this idea is that most hands are not really appropriate for DP tournaments. Either they're too easy -- the right bid or play is obvious to just about everyone, or they're too hard because there's no definitively "correct" actions, there's lots of judgement calls or stylistic differences (if you're an aggressive bidder you'll consider a different bid to be the most correct).

The hands that appear in bridge columns have been carefully designed or selected. The same with the hands in Bridge Master. Expert bridge players choose them to illistrate particular points.

#20 User is offline   silvr bull 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 253
  • Joined: 2012-November-11

Posted 2017-February-16, 11:11

View Postbarmar, on 2017-February-16, 09:29, said:

The problem with this idea is that most hands are not really appropriate for DP tournaments. Either they're too easy -- the right bid or play is obvious to just about everyone, or they're too hard because there's no definitively "correct" actions, there's lots of judgement calls or stylistic differences (if you're an aggressive bidder you'll consider a different bid to be the most correct).

The hands that appear in bridge columns have been carefully designed or selected. The same with the hands in Bridge Master. Expert bridge players choose them to illistrate particular points.

Thanks for a helpful reply. I must admit that I underestimated the difficulty with establishing a "correct" sequence for bid and play. The intense reaction to my simple example hand should have been sufficient warning to me, but I am a slow learner and an incurable optimist. Your mention of Bridge Master at the top of this thread was the first time I heard of it. I will have to look into it. Meanwhile, I continue to think Decision Point has a useful role in BBO individual tournaments, but maybe as a double dummy (instead of most skillful, that is so problematic to define) replacement for, or alternative to, Daylong tournaments. Thanks.
0

  • 2 Pages +
  • 1
  • 2
  • You cannot start a new topic
  • You cannot reply to this topic

1 User(s) are reading this topic
0 members, 1 guests, 0 anonymous users