The calculation is started by a call:
score2 -b [-s <number of samples>] <movement-file>
The program first shows the results for an assignment as given in the movement-file with the strength coupled to pair number as described. Then the scores are shown for 10 arbitrary assignments of pair numbers. For instance, the renumbering "4 2 3 1" means the original pair 1 now gets pair number 4, etcetera However, for the score calculation the original pair numbers are used, so pair 1 remains the strongest pair even though it plays with a different pair number.
Subsequently the average deviation (standard deviation) of the ideal value is determined by taking a large number of random assignments of pair numbers. This average deviation should be taken with a grain of salt as the distribution is by no means a normal distribution. Deviations of more than 3 times the average deviation are not uncommon.
To make the results look a little less extreme, Bussemaker gave an extra constant number of MP's to every pair before calculating the total score in percent. This has not been done here, but a similar effect is obtained by simply dividing the difference from 50% by 3. These reduced values are shown in a separate column.
Finally a probability table is output. This is a N x N - matrix showing the percentage probability that pair i ends up at place j in the final ranking. (N is here the number of pairs, i and j are numbers from 1 to N). In the ideal case this is a diagonal matrix with all diagonal elements equal to 100.
score2 -bw <movement-file> <strength1> <strength2> <strength3> ...Note that
score2 -bw <movement-file> 1 2 3 ...is equivalent to the original Bussemaker model, while
score2 -bw <movement-file> 1 1 2 2 2 2 ...is the case of the "2 strong pairs" model that was the purpose of 'score2' in the first place.
score2 -b -D<n> <movement-file>assures that the subgroups are kept separate when assigning the random pair numbers. If, e.g., the number of pairs in the movement is 14 then option "-D2" results in keeping pair 1 ..7 apart from pair 7 ..14. Notice that the score is always calculated across the whole field, only the participants of the each group are kept separate. This option may be combined with Weighted Bussemaker. Example:
score2 -bw -D2 24p6r.txt 1 2 2 2 2 2 2 2 2 2 2 2 3 4 4 4 4 4 4 4 4 4 4 4BEWARE! The program assumes that the pairs in each subgroup have consecutive numbers.
-v nyou can specify for how many cases (normally 10) detailed output is wanted. E.g. with "-v 1" you only get the output for the initial assignment 1 2 3 4 ..., with "-v0" no details at all.
The detailed output normally consists of the result (Pair, strength, MP, percent). With the option
-Syou get, in addtion, the scores per board.