Here is an outline of how I do the overall scoring for the NJ Scout Event, along with the reasons for choosing them.
First, remember that the Troop "competition" is really for fun. There is no prefect way to score it, and I have chosen one method. There are many others that could be justified, and each will likely give a different overall order.
Second, the methods presented here were chosen to score the Second NJ Scout Championships where there were only about 100 total scouts involved. I'm not sure if that makes a difference.
Third, I've chosen to include all scouts in the scoring, not to separate them into Boy Scouts, Girl Scouts, Cub Scouts, etc. This could make a difference in the scoring of the individual courses, and this fact should definitley be taken into account when looking at the overall standings. Typically age range (and hence, number of different courses) is wider for a boy scout troop than for a girl scout or cub scout troop. Hence, they will likely not score as high overall, so you should look to compare the relative order of girl scouts vs. girl scouts, etc. However, it is very impressive to see a girl scout troop, say, score high overall. It shows the depth of their orienteering skill.
For the individual courses I do the following. First, I base each individual score on time behind the winner. The winner gets 20 points, and a time of twice the minumum time gets 5 points. Then everyone else is fit linearly with no score less than zero and DNFs and such being given 0. Twenty was picked so a "perfect" troop score would be 100, and values of 5 points and twice the time seemed to be about right. Then, given these points on a course, I do a weighted sum of the teams for each troop, where whighest gets full credit, second gets "epsilon" times as much, third "epsilon" squared, etc. I picked "epsilon" to be sqrt(10) so that the third best time would be weighted as 1/10 of the best.
Why not just add up every team's score for a troop, or just the highest score? Basically I feel that an aim of this competition is to encourage whole troops to compete, and there should be a possible advantage to the troops which have more participation. However, some troops are just not as big, and they should not be penalized, either. My compromise was to weight the lower scoring teams by less so that their scores count, but that a troop with only one team which is very good, will score better than a troop with several mediocre teams.
Where does the value for "epsilon" come from? I picked it out of a hat because it "seemed about right."
The only other problem was whether to set the maximum total for a course to be 20 or to let it go over. I went with the latter. The maximum that a troop could get is around 28-29 with the epsilon that I chose.
For the score-O I just base the score on the total points, with first place getting 20 points and 25% of the highest score getting 5 points. Then, as above, everything else is fit linearly with no score being allowed to be negative. In fact, I believe that this scale implies that any score-O total greater than zero will have a positive score.
Why not use the average score? Well, there are problems with that. (I present the average score for informational purposes only; it is not used in the scoring.) The min problem is that the parks ware too small and the scouts are too good. Typically, a decent, single team can get somewhere between a third and half the total points. Adding another team (and not allowing for duplicating controls found), doesn't get another third, or half, because part of the ground that they have to cover has already been covered by the other team. Therefore, the average would go down as the number of teams goes up because each new team can get fewer points. One possibility would be to allow for controls to be "found" more than once by a troop, but that could allow all of the teams of a troop to find all of the same controls. It boils down to the total point scheme favoring larger troops, and the average points favoring smaller troops. I decided to favor the larger troops.
Of course, there is a serious problem with cub scout packs and brownies which have 6 to 10 teams who aren't going to be able to go for some of the far out controls, and may be forced to go after the nearby controls more than once just to give the teams something to do. That is a problem that I don't know how to deal with.
Finally, I just total the score for each course and Score-O and sort them.
Please direct questions or comments to: Rick Slater <rdslater@splash.Princeton.EDU>
Last updated: 25 January 2000