Implementing and testing algortihms for assigning golfers to foursomes in a way that all golfers play with other golfers about the same number of times. I (Ph.D. in decision sciences and long career in IT) have developed the algorithms.
Consider 12 golfers on a trip to play 6 rounds of golf. For each round there will be 3 foursomes (i.e., 4 people playing together). in 6 rounds of golf, a golfer will be with 3 other golfers 6 times, for a total of 18 companions. Since there are 11 other golfers, it would be ideal if every golfer played with 4 other golfers 3 times and 3 other golfers twice (4*3 + 3*2) =18. This is not possible to do. So, how close can we come? I have a formula for scoring "pairings" - the assignment of golfers to 4somes for each round. I have the algorithm for generating pairings. Need it implemented and tested. Based on learning, we will refine algorithm. This effort is simply to implement algorithm i provide.
No real UI work. input is simply number of golfers and rounds. output is best pairing we can find, sing scoring method i provide.
CHANGE: we need the ability to specify that certain golfers will not be in certain rounds. e.g, golfers 1 and 5 will not be in round 1; golfer 4 will bot bne on round 2 or 4. hence, we will have both foursomes and threesomes in rounds where there are missing players.
Understanding of basic combinatorics (remember probability 1?) of the number of ways to select k items out of n. Experince is algortihms and heurisitics would be good.
some understanding of combinations would be helpful but not necessary. this is a very straightforward project. when we get it done. we can possibly work together to add things such as golfer 1 and 3 should never be in same foursome. certain golfers are not playing in certain rounds (there will be some threesomes in addn to foursomes).
the problem will be restricted to 8 or 12 golfers and 3 or 6 rounds to start. this should not be hard coded but considered in design. input is number of golfers and number of rounds.
* * *This broadcast message was sent to all bidders on Saturday Jan 17, 2009 3:18:25 PM:
there were some problems reading zip file. i have redone with older compression method. this file contains more detail on problem and data structures needed. i do not want to reveal exact algorithm until i have selected someone. the algorithm will be supplied in detail and is not difficult to implement
* * *This broadcast message was sent to all bidders on Wednesday Jan 21, 2009 8:01:10 PM:
Please do not do any further work on this bid. i have found someone with good experience to do the job. Thank you for your interest.