Carry over in Bermuda Bowl

easy · November 8, 2003

Ok this is a 3 part thread.

1. Where in the conditions of contest (or anywhere else) can i find the formula for calculating carryovers? I downloaded the coc from wbf web site but found NO references to carryover.

2. When did the concept of carryover get introduced into High level bridge and why?

3. Is this concept a reasonable approach to determining the world championship?

Cascade · November 14, 2003

Ok this is a 3 part thread.

1. Where in the conditions of contest (or anywhere else) can i find the formula for calculating carryovers? I downloaded the coc from wbf web site but found NO references to carryover.

In the supplemental conditions of contest:

http://www.worldbridge.org/tourn/MonteCarl...nsOfContest.pdf

Table of contents entry 8 is Carry Over.

Wayne

McBruce · November 16, 2003

I don't know the answer to #1 or #2 but I have strong opinions about #3.

Italy won the round-robin. The round-robin is the stage of the event where all of the teams play exactly the same number of boards against each team. No team has an advantage other than the luck of the deals in each match--swingy or not. This is random and mostly cancels itself out by the end of the round-robin. The WBF awarded USA1 a carryover because in the Italy-USA1 match, USA1 won by 24-6 (in Victory Points). Over the course of the rest of the round-robin, constituting TWENTY-ONE TIMES AS MUCH BRIDGE, against THE EXACT SAME OPPONENTS, Italy was better than the USA to the tune of 1.875 Victory Points per match. Over 336 deals against the same opponents Italy was clearly the better team. But because of what happened in the 16 boards they played against one another, USA1 got a carryover.

I know of no sport that does things this way. The WBF's bizarre carry-over scheme is combined with this farce of the top half of the qualifying teams choosing their opponent from those in the bottom half. Among the stratgies this makes possible: let's pick Country A--they're not our biggest carryover, but it makes it impossible for the next team to play a match in which they get a carryover.

Why not be logical? The carryover should be related to the final VP difference after the round-robin, pro-rated so that the top-ranked round-robin survivor vs the low-ranked round-robin survivor, if they meet, gets the maximum carryover for each round. The difference between #1 Italy and #8 USA2 was 66VPs and the maximum carryover for the final was 24 IMPs. 66/24 is 2.75, and the difference between Italy and USA1 was 19.5 VPs. Divide that by 2.75 and you get a FAIR carryover for Italy of 7.09 IMPs.

Do it this way and the #1 team will have a fair advantage, well-deserved from its total accomplishment over 352 deals. The way it is set up now, a team could squeak into the knockout phase in the last spot having lost maybe 11-19 on average to the lowest fourteen teams, with 25-5 blitzes over the surviving teams. This team would get a carryover no matter who they played, despite being the lowest-ranked team in the knockout phase. That's just silly.

Italy finished first in the round-robin. First by a large margin. They shouldn't have had to be paying a carryover to a team they had already proven they were superior to. Without the carryover they would have won by 12. Most thinking people would say a win by 12 over 128 boards is more statistically significant than a loss by 39 over 16 boards (in a final round match with first-place clinched). The WBF carryover scheme is just nuts.

eyhung · November 16, 2003

I think it would help if some other views were added to this highly

charged issue.

1. For the most part, bridge is a game of making fewer mistakes than

your opponents. It is unsound to compare two strong teams by how they

play a set number of boards against weak teams, because the weak teams

may make fewer errors against one team by getting more straightforward

deals against them.

2. The current carryover scheme, while not perfect, does an excellent

job of ensuring that everyone in contention will always play to win

against another contender. Otherwise, in the late stages of a round

robin, it may be to the leaders' advantage to "dump" matches to an

inferior team on the bubble so as to eliminate a superior team on the

bubble. The current carryover scheme enforces that it will always be in

a team's best interest to play their hardest in matches against other

potential qualifiers, thus avoiding the appearance of impropriety.

3. The current carryover scheme also minimizes the problem of apathy by

teams out of the race affecting the later rounds of play. Say there are

3 teams, A, B, and Z. Going into the last round, Z is in last place,

while A and B are both certain to qualify but interested in maximizing

their carryover. If A plays Z in the first round and B plays Z in the

last round, it would not be surprising if Z played harder against A than

B, changing the "field" faced by these two teams and giving B an

advantage.

3. Tying carryover to overall VPs would also increase the opportunity

for "kingmaking". If Z likes B better than A, it could influence the

carryovers by dumping to B on the last day. While most people would not

stoop to such tactics, it is best to reduce even the potential for

it so as to avoid comment and/or hard feelings when a team out of

contention loses badly to a team in contention on the final day.

4. If a team picks a country with a lower carryover margin _only_

because it saddles another team with a worse carryover margin, this team

is worrying too much about other teams and not enough about their own

opponents. I think most teams (properly) pick based on personal

evaluation of the quality of the opponent and their own carryover, not

on how their pick affects other teams.

5. Although your example is contrived, I think it is more impressive to

go 25-5 against the surviving teams and 11-19 against the field than

vice versa. While in general, a large sample size is usually better

than a small sample size, in bridge, the quality of the opposition is

far more important. If you told me a pair did very well in the last 5

National Open Pairs but has also scored below 50% for all the sessions

played in the last 5 weeks at the club, I would conclude they were

playing carelessly against the lackluster competition (or

experimenting). If you told me a pair has done well in the past 5 weeks

of club games but has never done well in the 5 National Open Pairs

events they have entered, I would conclude that they were much better

than the members in their club, but not of national-level caliber. In bridge, winning against excellent opponents is more meaningful than losing against poor opponents.

Eugene Hung

c2 · November 16, 2003

I think the current schema is sound.

paulg · November 16, 2003

Whatever concerns anyone may have about carry over, in this case it's hard to argue about it in this instance.

Italy played USA1 in the final match of the round robin. Italy were already assured of first place and USA1 were lying second.

Both teams knew there was a very high probability that they would play each in the later stages and that carry over may be important. Both teams played their strongest line ups.

If Italy could have beaten USA1 in this match, then this would give the Americans a more difficult quarter final.

Perhaps in these circumstances you could ask why the 16 board match should not count in its entirety :) (although this may be difficult to get into the CoC).

Paul

(actually an Italian supporter)

McBruce · November 20, 2003

I think it would help if some other views were added to this highly charged issue.

Let me state (as I have on rgb) that my motivation in this is to make things better for next time, not to change the result of last time or question the result's validity.

1. For the most part, bridge is a game of making fewer mistakes than

your opponents. It is unsound to compare two strong teams by how they

play a set number of boards against weak teams, because the weak teams

may make fewer errors against one team by getting more straightforward

deals against them.

These are not weak teams! These are teams which have:

a) won their national championship, AND

:) finished high enough in their zone to earn a spot in the BB!

The 22 teams may not be the best 22 in the world because of the limitation of one (or two for USA) team per country, but none of them are hopelessly weak (although the Bermudans had a bad time of it).