MastersHandicapping

Rather than piecemeal my progress, I was able to write a program that would process the data once I got it.  When I received the data from Concept2, it was just a matter of running the program, trying various scenarios, and settling in on something that I could present.  In any case the initial draft of my findings are now available.  There are tables and charts, which doesn't seem to be appropriate for a blog, so it is in PDF format.

Draft Masters Handicapping Report

Comments, suggestions etc are welcome.  I am planning on submitting it to US Rowing to see if they might be interested in printing it in their mag.

Concept2 provided me with a set of data for 1K Erg times from May 2004 through 2014.  There are about 16,000 records. First step is to take a look at the distribution of data.  The following is a summary of the data, looking only at how many erg times are available by gender, data source and weight classification.

Looking at it there is clearly plenty of data for heavyweight males to around 65, . . . after that, we will need to see.  For female rowers, there is a lot less, about a fifth of what is available for male rowers.  As I further dissect the data, it is clear that light weight data is even more scarce.  The initial analysis will be on heavy weight male data records probably from the IND, IND_V and C2Log datasources.

For now that won't be an issue.  Concept2 to has offered to provide data from their database to use.  I have given them info on what I want and will await the reply.  For now, I have only requested data for 1K distances.  The can provide data for age, sex, either Heavy or Light weight (not specific weight) and source.  Source is interesting, it means I will need to do a comparative analysis to see if there is a significant difference between times per age from different sources.

I'm anticipating eventually looking at all the following:

• Average time and std deviation for each age for all data.
• Average time and std deviation for each age for each source.
• Recompute the above based on a max and min value based on a yet to be determined deviation (outlier).

If there is weight data, it might be interesting to recomputed the above, but instead of simply average time for each age use a power ratio, . . .that could be interesting.

Comments are welcome.

In order to have something worthwhile, I will need at least 50 entries per year for all years starting at age 27.
Concept2 has responded and will help out with data. They need to know more about what kind of data I'm looking for. Will keep you informed as things progress.

This section is for me to provide info and get feedback on the masters handicapping project. To start off, I just sent Concept2 a request to access their database. Don’t know how they will respond, nor do I know if I will be able to use their data. When I started this, I was thinking that I wanted personal best times for 1K and 5K because they are the distances an individual would race and also it would be representative of their effort in a race.

If Concept2 provides me access to their database and I can somehow filter the above type of data from it, then I can pursue my strategy. If not, and all I am left with is lots of data tied to different ages, but without any indication of what it represents, . . . I will have to think of a numerical way to that that into account, . . . hence one of the reasons for this discussion group. Hopefully there are math jocks willing to provide some suggestions on how to refine the data.

So far there has only been a handful of entries. I figure for what I want to do, I will need at a minimum 50 data entry points per year for about 40 years (27 to 67). Of course, the more data, the better. If you are an organization that has collected data for your masters, let me know. If it is in Excel or another database, I could write something to get the data into my DB, . . .

Until I have enough data, there isn’t much I can do, other than dream about the possibilities. If you have data, suggestions, want to help, etc. You input would be greatly appreciated by the community at large!