The Vagaries of Scoring (latest posting March 23, 2016)Duplicate bridge is scored like a round robin sports tournament. For each board we compare the scores of all N/S teams with each other and the scores of all E/W teams with each other. For any given pair of N/S teams the one with the higher score is consider a winner. For example, suppose that on a particular board the likely contract is 4H for N/S (non-vulnarable). Suppose that teams A and B make the contract (420 pts), team C goes down (50 pts for E/W), and team D scores an overtrick (450 pts). Then teams A and B are considered to have lost to team D (420 < 450), won over team C (420 > -50) and tied with each other. Team D has beaten the other three and team C lost to the other three. Each team gets 1 point per victory, 0.5 points for a tie, and no points for a loss. In this example team D gets 3 points, teams A and B 1.5 points each and team C 0 points. The trouble with this is system is that these teams played different E/W teams on this board. It is not possible to have them play the same E/W team because that team would have seen the board multiple times. It is an imperfect system and N/S team Y may score a win over N/S team Z because Y played a very weak E/W team Here is an example from one of our sessions,. On board 20 A play N/S (and so do B). A(North) opens at 1 NT (15 pts). A(South) has only 4 pts and she passes. The same happens when B play board 20. A's opponents on the board is pair C with (11,10) points. So they pass and A make the contract and earn 90 points. When B play board 20 their opponents are pair D and they bid 3D (in spite of their few points). They go down by 2 and pair B gets 100 points. This gives them a "victory" over A (who got only 90 pts). But the result has nothing to do with the relative skills of A and B. It is due only to poor bidding of pair D! In the long run things average out and the scores reflect reality but not in short plays. The Effect of Not Playing all BoardsClick on this link to see a table illustrating the effect of not playing all boards. When all three boards are played A is the clear winner in N/S and C is second. But in two of the instances of omitting a board A and C are tied. In E/W the rank is F, D, E when all three boards are played. Depending on which board is skipped the rankings are: E and D tied, F third; D and F tied; and E third; F, E, D. Note that the results for N/S are affected less than those for E/W. Why? The board scores were computed on the assumption that B was a much weaker team than A and C while the E/W team were roughly of similar skills. (See the scores on the yellow line.) We may poin out that the scores in the red, green, and blue line add up to the scores on the yellow line. Thus omitting a board does not eliminate the effect of the skill of the teams on the scores. Thus A and C always outperform B. But for teams of similar skills (E/W in this case) it plays havock with the rankings. |