TullyRunners - Article
Quick Look at Newbury Park Boys 2021 at CA States Compared to German Fernandez 2007
by Bill Meylan (November 29, 2021)
I have received some e-mails wondering about the Speed Ratings of the top Newbury Park boys at the 2021 California State XC Meet versus the Speed Rating of German Fernandez at the 2007 California State XC Meet.
The California State Meet is held annually at Woodward Park 5K course in Fresno, CA. German Fernandez set the current Woodward Park course record of 14:24 at the State Meet in 2007 which I speed-rated at 209.0.
The Newbury Park boys achieved the highest-ever single race Team Speed Rating Average at the CA State Meet 2021 ... The remarkable Newbury Park results and Speed Ratings were as follows:
The valid question being asked is ... Based on the race times of the Newbury Park boys on the same course run by German Fernandez in 2007 (only a few seconds slower than German Fernandez), the Speed Ratings of the Newbury Park boys seem TOO LOW.
My Brief Answer to the Question ... Based on multiple race results, Woodward Park is running faster now (and recent years) than it did in 2007 (and the years around 2007) ... That response has not been "well received" by a several California folks.
My ONLY concern in speed-rating California races is to derive reasonably accurate numbers so I can compare California individuals and teams to other teams and individuals around the nation ... The CA State Meet (in non-covid infected years) yields the teams that qualify for NXN Nationals, as well as individuals to NXN Nationals ... So I want decent speed ratings from the CA State Meet for that purpose ... and I'm satisfied with my California speed rating results over-the-years (even though many observers from other States believe I frequently over-rate CA teams and individuals).
So I decided to post some brief statistics and a graph that provide some data of consideration to those who might be interested.
The CA State Meet has been relatively constant in its format over many years ... Five divisional races for both the boys and girls ... Total number of runners about 950-980 boys and the same for girls ... This is great for somebody who wants to speed-rate races because it helps in direct comparisons.
Below are some table statistics that I rarely use in speed rating, but find useful in showing "trends" and find helpful in explaining "anomalies" in occasional races ... For speed rating, I focus on roughly the top 60% of runners at any race ... For the Merged Results of the CA State Meet, I focus mostly on the top 600 finishers and their seasonal speed ratings in my database.
I start out by graphically comparing the top 600 finishers (as a group) to previous CA State Meets of known speed as determined by my speed ratings ... This gives me an initial starting point iteration "guess" for "How fast was Woodward Park running today compared to previous years of known speed" ... I use that "guess" and then start adjusting the guess by fitting the current seasonal speed ratings of the individual runners to their final race times and time differences between other runners of known speed.
The tables below show the median and average times of the top 600 boys and top 600 girls at various CA State Meets ... It also shows the number of runners that achieved selected times at the State meet:
The numbers in the tables clearly indicate that the races in 2005-2007 are slower than the races between 2016-2021.
Assuming the quality of the runners (especially the average group) is roughly the same (does not need to be exactly the same ... there is usually some variance), then the course was running slower for the the runners during 2005-2007.
Various factors must be considered when making this type of direct comparison:
Some observers suggest that the California high school runners making the State Meet are better (higher quality) currently than back at German Fernandez's time ... I have not seen or generated any statistics that support that suggestion ... You can definitely make a case for specific better teams & individuals ... But as a group (especially mid-level runners as a whole) I assume little to no difference over the time span being considered ... There is always some variances (up & down) with quality of specific leagues, sections, individuals, but that should not effect my discussion here.
Merging multiple races ... Merging 5 races at the CA State Meet involves compromises ... (1) different times of the days, (2) maybe different conditions (temperature, weather, etc), (3) different pacing of races ... So merging races can affect speed ratings and the method I apply to derive the ratings.
For the California State Meet, I have always derived a single race adjustment-correction and applied that adjustment to all races (separate derivation for boys and girls, but often the same result) ... Since I am usually interested in the teams going to NXN, races with those teams get my focus ... During the season, I focus on specific races at large invitational meets for my race adjustments.
I love graphs (some people dislike them) ... Graphs can give a quick depiction of the data ... The graph below shows the Boys CA State Meet for four years (top 600 finishers) ... I plotted only 4 meets to make the data easier to see ... It is obvious the races were slower overall in 2007 & 2006 compared to 2021 & 2019 ..... Side-Note - a graph of the girls for the same years looks very similar to the boys.
It appears that 2019 was faster than 2021 (same as the tables above) ... But normalizing the data and projecting it back to the origin (in combination with individual runner speed ratings), results in the same race correction for both years ... Pacing of specific races & times of the day of specific races could also contribute to the apparent variance (a compromise of merging).
Speed ratings are a runner-to-runner measurement ... More separation from runners of known ability yield higher speed ratings ... and the graph nicely illustrates German Fernandez's larger separation from his competition compared to the Newbury Park guys ... That's why German Fernandez had a speed rating of 209 (and not 202).