I did not go racing last Sunday, but took a look at the weather data to see what sort of conditions my friends had for racing that day. Surprisingly, since a long distance windward/leeward race was called for, the wind direction was extremely consistent.
This illustrates the steady wind conditions observed for last Sunday’s race (September 9, 2007). The stable pattern had developed around midnight and persisted throughout much of the day.
Later on, I learned the race course was set for 5 degrees, not even close to the 333 degrees the steady winds were coming from when I had looked at the weather.
Wondering if there had been some weather changes to the extremely stable weather pattern, or if the RC had blundered by setting a skewed course when conditions allowed something much more ideal, I took another closer look at the wind data and the racing times of the boats on the course.
The easy-to-read wind data available online in real-time was now stored in a less friendly archived format. After taking a rough look at some raw data, which reaffirmed the steady wind direction pattern but left some questions about race course conditions, I decided to use the opportunity to dust off some software and start developing some rudimentary data modeling tools for sailboat racing…
Here is an interesting look at wind conditions for racing on September 9, 2007:
Crib data was used as source for wind direction. Most of the boats I race on do not have equipment to record wind data (not to mention I was not racing last Sunday!).
There may be some discrepancies with regard to using this information for race analysis. Specifically, the geographic location of the wind instruments in a fixed location not on the race course and higher than most masts; and the resolution of the data (each data point represents a 5 minute time interval).
Given these limitations, however, the information still provides some useful observations that would not otherwise be available.
Among other things, the graphical view seemed to present some insight into managing fleets in different section breaks to optimize time on the water. In hindsight, the advance decision to hold a single long distance race prevented the opportunity to benefit from the stable wind conditions. Alternatively, an earlier start schedule would have also allowed a long course or multiple races during stable and consistent racing conditions.
I imagine experienced long-time race committee planners have plenty of knowledge for assigning start times and sequencing races, but the issue seems to generate a lot of discussion and opinions, and graphic analysis such as that shown above might tend to show the pro’s and con’s of various scenarios more clearly to assist in better planning and evaluation.
Playing with more sailing data and exploring more possibilities for improving race performance and analysis is intriguing…
