Mijn RS Aero uitslag webstek heeft een rankinglist. Op de profiel pagina van elke zeiler staat een grafiek met de ranking trend. Ik kan die grafiek uiteraard real-time genereren, maar ze zijn behoorlijk ‘static’, dus ik genereer ze in batch. Daarvoor is een C# programma gemaakt. Het grafiek deel maakt gebruik van ScotPlot.
The calculation is inspired by the ELO rating that is used in chess. ELO however is not exclusively for chess! The Elo rating system is a method that estimates the relative strength of a player in a game. In this system, beating a higher-rated opponent gives you more points than beating one with a lower rating. At the same time, your opponent loses as many points as you win.
For all details go visit the Omnicalculator. Remarks:
- Each new sailor starts with a rating of 1058. This is the current (2024) median value of ELO ratings (i started with 1000). The median is corrected on January 1st;
- I am using one ‘development coefficient’ (k-factor): 5. This results in relative small changes per event;
- After each event a comparison is made between each sailor (for a regatta with 10 entries: 45);
- For regattas that use a handicap system (like PY) that comparison is done per rig (for a regatta with tree 5’s, two 6’s, four 7’s and one 9: 10 comparisons.);
- Rig ‘agnostic’ regattas are not use for the calculation (for obvious reason);
- Mixed fleet (different boattypes in 1 class) regattas are not used for the calculation;
- The results are processed sequentionally in date order. If late results arrive this could impact the ratings points of already processed events; meaning: the ratings are changing almost weekly;
- Initially the start rating is hidden. After 100 comparisions the rating is visible;
- The minimum gain per comparison is .6 (In effect this means that very good sailors do not earn points when winning from sailors with a much lower ranking.
- Sailors without a result in the past 2 years wil become ‘inactive’ (= invisible in the ranking list), they will not loose their ranking points;