[back to Home Page] [back to Articles Page]
Standardized Race Time Ratings
by Bill Meylan (TullyRunners Webmaster)
(update draft: January 2001)
This article is an extension to the article "Computer Ranking in Cross Country?". In that article, I describe the difference between the actual cross country race time ("raw" race time) and the "standardized" race time. I use "standardized" race times to rank both individual runners and teams. However, when actual race times and "standardized" times are listed side-by-side in a runner's past performance listing, it can be confusing to a casual observer ... this is the response I received from several reviewers who thought it might confuse too many readers. Here's an example for Skaneateles's Lia Cross (the first time is the actual time and the second is the standardized time):
11/13/99 NY States-Class C 5 20:12.2 19:42.2 11/6/99 Section 3 - Class C 2 19:59.3 19:59.3 10/30/99 OHSL 1999 4 19.36.0 20:06.0
The confusion is understandable for most readers ... (1) which time is which?, and (2) even when the reader does realize the difference, the times have little meaning for most readers.
Therefore, I decided to convert the standardized time to a "standardized race time rating" that most readers can quickly comprehend with some understanding. Cross country is a sport that has virtually no performance statistics that common sports fans can relate to ... baseball has batting percentage, ERA, etc ... basketball has scoring average, shooting percentages, etc ... football has rushing yardage, completion percent, etc ... soccer has goals scored, assists, etc. ... and track & field has race times for each distance and heights / distances for field events. Cross country is unique, and similar statistics are not really possible.
Most fans understand plain numbers. For example, basketball can have "power ratings" ... as a casual fan, you don't need to understand how the rating was produced, you just need to understand that "higher is better". For cross country, I decided to convert the "standardized" race time to a plain number rating where "higher is better". The conversion method is explained below. The resulting "standardized time rating" is similar to the Beyer Speed Figures that appear in several versions of the Daily Racing Form for horses ... speed figures for horses have been available for many years (and I am thoroughly familiar with them ). The obvious reaction is: "Are you comparing people to horses?" ... NO ... I compare horses-to-horses and people-to-people. However, similarity in performance lines between people and horses is actually quite astonishing!
Converting a "standardized" time to a number is simple as shown in the conversion table:
|Standardized Race Time||Standardized Speed Rating|
I arbitrarily selected the rating numbers ... I assigned a value of 100 to 21:00 because 21:00 on the SUNY Utica course corresponds roughly to the median "above average" level in section 3. I also decided that one rating point equals three seconds. Note: ratings are rounded up/down to whole numbers.
Now here's the Lia Cross performance list from above with actual race times and standardized speed ratings:
11/13/99 NY States-Class C 5 20:12.2 126 11/6/99 Section 3 - Class C 2 19:59.3 120 10/30/99 OHSL 1999 4 19.36.0 118
My statistics are now based upon the SUNY Utica Tech race course at the time of the 1999 sectionals (11/6/99) as being the "standard" race course ... All "standardized" times are relative to this course. Therefore, the actual race times and the standardized race times are exactly the same by definition (for the 1999 sectional race) for every section 3 runner who ran last year at sectionals.
Here is the actual computer equation I use to generate speed ratings:
Speed Rating = (1560 - (actual race time in seconds) - (course correction factor)) / 3
where 1560 is the number of seconds in 26 minutes ... 26 minutes is used because it corresponds to zero in the chart above ... the course correction factor is how fast or slow (in seconds) a race course is in relation to the SUNY Utica standard course ... the entire expression is divided by 3 because one point equals three seconds. As an example (using Lia Cross), the course correction factor for the NY State Meet is +30 seconds (I statistically determined that the Westchester CC State course was 30 slower on 11/13/99 than the SUNY Utica course on 11/6/99) ... Lia's State time of 20:12.2 equals 1212.2 seconds; therefore, her State speed rating is calculated as: Speed Rating = (1560 - 1212.2 - 30)/3 = 125.93 = 126 (rounded). The course correction for the Section 3 Meet is zero by definition ... for Lia, Speed Rating = (1560 - 1199.3)/3 = 120.23 = 120 (rounded). The course correction for the OHSL Meet was -30 (30 seconds faster than SUNY Utica) ... for Lia, Speed Rating = (1560 - 1176 - (-30))/3 = 118.0 = 118 (rounded).
Evaluation of Speed Ratings
Speed ratings are useful for evaluating running performances. They are a decent indicator of running ability ... but they are NOT absolute! They need to be interpreted with some care ... time permitting, I hope to write a more comprehensive article about evaluating speed ratings in the future. In general, better runners have higher speed ratings (this is obvious) ... but other factors must be considered, for example, (1) Is the runner improving?? Are the existing speed ratings for a runner in the process of "getting in shape"??, (2) Do some ratings correspond to a runner who was sick or injured?? (3) How competitive is the runner?? ... a really competitive runner can often overcome lower speed ratings and beat "supposedly" better runners.