The Better than Average Surfski Race Scoring System
A few weeks back I was cornered by a fellow Durban Paddler, David Gwynn, after a rough Pirates to Umhlanga and back race. He wanted to discuss our current system of scoring and rating surfski paddler’s performances at events. This was something I had never thought much about, but after speaking with David, it quickly became clear that there a few shortcomings with our current system. He has devised a very simple and ingenious system that needs serious consideration.
Below is his explanation of the short comings of the current system and the benefits of doing things a little differently.
Thoughts on Surf-ski Race Results and their Analysis and Presentation.
Our results are unique in that they not only reflect an athlete’s skills and fitness level, but are also influenced strongly by several environmental factors. Presenting simple times and distances as in athletics, cycling etc is plainly inadequate.
We thus have that interesting PERCENTAGE column in our current results.
The current system uses the winners time as the reference point for the entire field. If the winner has a relatively poor performance he makes everyone look better when in fact they have not improved. Similarly if he outperforms, the rest of the pack will appear to have deteriorated when in fact they have not. The same applies when the upper echelons are absent from the race (e.g at river races/racing Mauritius), the entire result is skewed.
Another problem may arise when a temporary change in race condition benefits / hinders this SINGLE paddler and misses the majority of the participants. For example a SW buster hits Durban minutes after the leaders finish but the rest of the field is still out on the water.
The current system , though often working reasonably well, is not as robust as it could be. One often finds unrealistically large swings in a paddlers week to week percentages.
This makes it difficult to use the results in a meaningful way.
Is my training starting to benefit me?
Am I improving race on race or season on season?
Is this boat faster?
Is my race strategy bearing fruit?
Are the Durban Boys better than the Cape Town Boys?( or Aussies)
Am I better in downwind races?
Is this week’s winner as good as last seasons winner? Under the current system all winners are equal – 100% !
This systematic fragility may also interfere with grading and legacy issues. I believe this weakness has been exploited by paddlers wanting to boost their ratings before important events.
Build a system that gives the paddler real feedback on his/her performance in each race and to develop a rating system for paddlers that reflects current form best.
What I am NOT doing is trying to build a system that decides who would win a series/competition. That is up to individual race organizers.
How I do this:
Just use the AVERAGE time of the ENTIRE field as the marker rather than the winner!
Elegant and Simple.
David has been testing his system by taking the results from the races thus far and running them through his calculations. I have been so impressed with the outcome that I have published his results here . You can also find them under the results tab above. Just LCS results at the moment. This is a work in progress so there maybe a few errors. I suggest opening the link in a new tab and follow along with me as I point out some examples of the strengths of the new system.
Using my own results – Robin Tindall, as point of comparison, you will see that I am pretty consistent with an avg rating of 103.4% This means I am 3.4% above avg, out of the avg race field in Dbn. (Bit humbling as I thought I was better than that…) All my races had different winners though. Therefore, my percentages would have varied wildly race on race under the current system. Consider this weekends Rhotec race. Ollie Burn took the win. Under the old system he would be scored 100%. I would have scored a higher than usual percentage as I much closer to Ollie than to Hank or Matt. So I would have had an incorrectly inflated result. But when comparing my result to the avg, its virtually the same as previous races, as the avg hardly varies race on race. This lets me compare races with a degree of accuracy to see if I am improving or not.
One can also rate Ollie’s performance (113.1%) against the likes of Hank. Hank usually wins with a 120% score, Matt with 119% and Grant with 118%. Interestingly Jasper won SA Champs with 121.8% which is better than Hank’s best result this season. Who is the better paddler….?
The potential for this system is endless. For me, a big advantage is one can generate a seasonal “ladder” rating for every paddler. For example, I am currently rated 103.4 and lying 47th for the season. I can try and improve my rating and my ladder position as the season goes along. David and I are still working out the kinks. But I will be posting updated results using the new system after each event. So keep checking back.
If anyone is statistically or mathematically minded please comment. We would really appreciate feed back so we can fine tune and improve.