A new category scoring method

Traditional category scoring

In most competitions, besides the overall scoring, we also produce various category scorings: Female, Sports Class, or all pilots of one nations for their national championship.

Traditionally, those category rankings are produced by "filtering": The pilots who fall into the cateogry in question are filtered out from the overall scoring. Their score is always identical with their score in the overall ranking.

The problems with filtering

This filtering method is very easy, but it can lead to very tight rankings, where pilots are separated by only a very small number of points. Which potentially leads to a chance competition winner, especially in slower categories in competitions with a good number of fast wings and pilots: The slower pilots - competing in Sports Class, for example - will be bunched up towards the bottom of the ranking, with very few points to separate them. The pilots who do well within their categoriy and beat their opponents into goal are rewarded much less than the overall task winners. This because time points distribution becomes "flatter" towards the end of the race, when the slower pilots arrive in goal. For them, their differences in time translates into very small differences in time points. After a certain time, which depends on the overall winner's time, they may not even get any time points any more, which leaves them all with the same amount of distance points, and the few leading points they may get.

Another issue is that if a task is set too ambitiously for Sports Class, but the higher-end wings do just fine, the day still counts fully for everybody, even if all the Sports Class pilots landed shortly after the start.

A new approach: Separate scoring

To counter these effects, and to have competitors in slower categories fully benefit from the GAP formula's task validity and points distribution schemes, some associations and competition organizers have started scoring categories separately: While everybody is still ranked in the overall ranking, just like today, the categories are scored from scratch, with only the pilots of that category. This means that the category task winner can and often will score 1000 points, or close to that. If the task is set badly for the category, the category's task validity will be reduced accordingly, regardless of the overall day validity. Category results thus reflect better the relative performance of the pilots within that category. They become more decisive, giving more weight to achieving top rankings within the category in each task.

One thing we will need to get used to with separate scoring: The order of pilots in the overall scoring may end up being different from their order in the category scoring. As the category is scored with only the category members, the points distribution within the category may be quite different from that in the overall ranking, which can change the pilots' order in the two rankings. But that is only fair: One is the order in which those pilots are ranked considering all the other, probably faster pilots. And the other is the ranking that takes into account the performance of only that category. These can and should be different.

Using separate scoring in FS

Until now, scoring categories separately meant a lot of additional work for the score keeper. But with FS 2020, using separate category scoring has become effortless:

When creating a new competition, note the new setting just below the scoring formula choice: "Category scoring". Here you can pick between "Filter (traditional" - the way most competitions have scored categories to date, see above - and "Separate scoring". To keep changes to a minimum, we use the filter method by default, but encourage organizers of FAI Category 2 competitions to consider using separate scoring going forward.

If the competition file was created by an external competition management system like Airtribune, you can change the setting by opening the file in FS, then going to File - Edit competition details...

Once you have set this up, everything else works just as before: Enter participants, define tasks, import and evaluate tracks, score competitions, produce reports. You will notice that when producing category results, your computer will work a few seconds longer than before - it has to re-do the scoring for that category in the background. But that's the computer working, as it should be...

If you have any questions about FS, and about category scoring, write to: support@…

Last modified 4 months ago Last modified on May 24, 2020, 8:12:13 PM

Attachments (3)

Download all attachments as: .zip