good point, 190 seems to be slightly more attainable. one of these days Ill get around to using previous year XC data with Track times to give some indication.
4:10 and 9:00 seem like good cutoffs to start with.
good point, 190 seems to be slightly more attainable. one of these days Ill get around to using previous year XC data with Track times to give some indication.
4:10 and 9:00 seem like good cutoffs to start with.
jabouko wrote:
good point, 190 seems to be slightly more attainable. one of these days Ill get around to using previous year XC data with Track times to give some indication.
4:10 and 9:00 seem like good cutoffs to start with.
Meylan himself did an analysis like this about 15 years ago. You would need to do a new linear regression, with modern numbers, but you can look at his articles about it as an example. I myself did this with 2017-2018 ratings, and found these numbers:
For 1600: time (in seconds) = 399.7 - (0.7725 * SR)
For 3200: time (in seconds) = 883.2 - (1.7795 * SR)
I also added 5000 times with his normal 3 second slope and a 15:00 as the equivalent of a 185 (this was just a guess but seems pretty accurate based on the track 5k he rated this year). If I had to guess, you will find similar slopes but with a slightly lower intercept, as speed ratings have gotten a bit lower on the whole since 2017. I should also add two more things; these are overall seasonal ratings, NOT individual highest (mainly because overall ratings are much more reliable). These are also just New York ratings, you would get very different numbers if you used a different state or did it on a national basis. You would also likely get slightly different results if you did the same things with girls times. For your interest, these regressions gave times like this:
200 - 4:05.2 - 8:47.3 - 14:15.0
190 - 4:12.9 - 9:05.1 - 14:45.0
180 - 4:20.7 - 9:22.9 - 15:15.0
170 - 4:28.4 - 9:40.7 - 15:45.0
the funny thing is speed ratings will actually decrease a little bit when compared to all the speed ratings of the past!
At least my theory is that with more Runners having better training, more access to information and knowledge about other Runners across the country, better nutrition, just more access to technology and equipment
My thought is that you will have more parody and it'll be hard to really separate yourself. So I think it'll be really tough to break some of the speed rating performances of the past
Anyway correct me if I'm wrong but it would seem different than track times which will continually be improving
Not sure if it makes sense but it might be reasonable to just increase speed ratings by a point or so every decade to keep up with the fact that Runners are just getting faster on the whole each decade
my question is say you had the ability to take the top performances of all time and measure them on some standardized course
And there wouldn't be any competition or adjustment for someone who could have ran a little faster. Just whatever effort they gave on that day in history
What would have been the top performances of all time?
Anyway as has already been said, I think if people see speed ratings as where you rank compared to your fellow competitors vs some absolute historical guide of the all-time best runners , it's way more useful and accurate!
Hutchins ran the best performance of all-time but the speed ratings are lagging. I don't think any other runner from the past could have broken 16 minutes at RunningLane. The only runner who could run faster is also running this year, Thorvaldson. Too many here try to compare ratings from previous years which just does not work.
Just some general observations that may be of interest.
I do spend some time looking at, what I call, “the average-of-the-average” … meaning the group that comprises the typical average high school XC runner … Since collecting data from the late 1990s to the present, I do not see data that indicates any significant improvement or deterioration of this group of runners for the purpose of speed rating.
But I have noticed an increase in the size distribution of “above-average” group of runners and a decrease in the overall “average” group … The “Elite” group varies from year-to-year and varies between the boys and girls.
High Foot Locker Ratings … Maybe looking at high speed ratings NOT achieved at Foot Locker can yield some things of interest … Four brief examples:
Kate Tuohy shattered the Holmdel Park course record at the NJ Shore Coaches Invite (10/6/2018) by running 16:21.0 for a speed rating of 177.00 … and many observers thought she was under-rated … However, several NJ coaches did note that Holmdel was running unexpectedly fast that day possibly due to a “Tuohy-Effect” (meaning everybody knew she there to set a record and more than a normal number of kids ran fast because of it, and that lowered her speed rating) … That 177 was Tuohy’s best speed rating ever, and nobody questions its accuracy in terms of being too high.
Edward Cheserk set the Holmdel Park course record at the NJ Shore Coaches Invite (10/1/2011) by running 14:53.0 for a speed rating of 208.3 … the Holmdel course was running a bit slow that day … Several NJ people told me Cheserk would likely have run 10-15 seconds faster on a better day plus he went out a bit too fast during the first mile which did not help … No body complained about the speed rating being too high … I wish Katelyn Tuohy had run her 16:21 Holmdel time that day because her speed rating would have been 181.0, and complaints about Amber Trotter’s 180.3 rating at the top of the girl’s list might go away … Edward Cheserk went on to win Foot Locker that year by 0.6 seconds over Futsum Zeinasellassie for respective speed ratings of 209.8 and 209.6 … Why do people think those Foot Locker ratings are too high?
German Fernandez won the California State Meet in 2007 with a time of 14:24.0 and a speed rating of 209.0 … No complaints then … But it did lead to complaints that I under-rated Nico Young’s CA State Meet win in 2019 with a time of 14:28.5 for a speed rating of 201.5 … This is case where profiling all the CA State Meets from 2007 through 2019 shows the difference in course speeds and why those ratings exist … Also, Fernandez was 49 seconds faster than anybody else and only 7 other runners were within a minute of him … Nico Young was 15 seconds faster than runner-up Matt Strangio, and 44 runners were within a minute of Young … In 2007, Jordan Hasay had a CA State Meet speed rating of 158.
Dathan Ritzenhein has 2 of the top 4 Foot Locker speed ratings (213.7 in 2000 and 211.7 in 1999) … In 2000, Ritz beat Alan Webb by 20 seconds (207.0 rating for Webb) and Ryan Hall by 24 seconds (205.7 rating for Hall) … I was asked to rate Ritz’s Michigan State record time of 14:10.4 at the MI State Meet course which came out at 210.53 (211 rounded up … I did post an article about how this rating was derived) … It is consistent with his Foot Locker ratings, and was accomplished with no helping competition in-state … Sometimes high speed ratings at Foot Locker are the result of talent … Nobody has ever specifically questioned Ritz’s speed ratings, possibly due to respect for his high school abilities.
Personally, I consider the speed ratings from past years comparable to current speed ratings ... I do not consider the current speed ratings to be lower than older speed ratings especially when all the groups I evaluate are examined ... Just looking at the very best runners from each year is insufficient.
What is the girls HS 5k track record? How many u19 college freshmen have broken 16, on the track or in XC? That may give a good indication of whether others may have been able to break 16, but just never ran on a fast enough course on a fast enough day.
I think any attempt to guess who is faster between Hutchins and Thorvaldsen is pointless. Any ranking should rank them equal.
They did not have a real track season so we don't know what they would have run. I predict that 2 girls will break 9:30 this year which should solidify who the fastest XC runner all-time is at 15:58. RunninLane is fast but that doesn't mean a 9:40 girl can break 16 there.
This is very interesting to me, because I actually thought it was because the average runner is getting faster. This makes sense, though, instead of being a few lone runners out in front and a big pack of average runners, most good races now have a lot of good, but not great runners in the 160s, 170s, and 180s (for high level meets) that are above average, but not running 200+. This extends to track times; people have been running sub-4:00 in the mile since the 60s, but there are so many more people nowadays that can run sub-4:30 compared to then. Just interesting.
This I'm not quite sure about. I did an analysis of several courses in New York that have relatively consistent ratings within a single year, and are contested multiple times every year. I used SPAC, VCP 4k, and Bowdoin Park for this analysis. I did a cross-sectional analysis within a single year (to find which courses were most consistent), as well as a time series analysis from 2004-2020. There is a consistent decrease in the average speed rating for a particular time from year to year. To me, this kind of contradicts what you're saying. The courses are not changing. The conditions are relatively consistent. It's not any easier to run a 16:00 on Bowdoin Park now than it was in 2004. But the data shows that the average speed rating has declined by about 0.8 every single year, with a fairly consistent drop from year to year.
my data wrote:
...
This I'm not quite sure about. I did an analysis of several courses in New York that have relatively consistent ratings within a single year, and are contested multiple times every year. I used SPAC, VCP 4k, and Bowdoin Park for this analysis. I did a cross-sectional analysis within a single year (to find which courses were most consistent), as well as a time series analysis from 2004-2020. There is a consistent decrease in the average speed rating for a particular time from year to year. To me, this kind of contradicts what you're saying. The courses are not changing. The conditions are relatively consistent. It's not any easier to run a 16:00 on Bowdoin Park now than it was in 2004. But the data shows that the average speed rating has declined by about 0.8 every single year, with a fairly consistent drop from year to year.
It may seem that SPAC, VCP 4k, and Bowdoin Park have been consistent in speed over the time period 2004-2020, but that's really not accurate ... All of the courses have undergone some changes.
VCP 4K went through a period of park construction at times which affected the speed ... the course has changed several times in distance from 4K to 2.5 miles (48 meter difference) ... I still get yelled at by people who frequent VCP, but in a year when the course was switched back to 4K, 3 guys broke the course ... different "2.5 mile" courses are also used (and have been used) by different people (one starts and finishes at the hare & turtle and one doesn't) ... the "2.5 mile course used by the CHSAA this year is clearly slower than a typical Manhattan Invite course ... Bowdoin is faster now on good days than back in 2004 and its speed can varies significantly with weather and ground softness in areas ... and SPAC varies whether they use the 3.04 course, the 3.05 course or run the course frontwards or backwards in sections (and it has undergone some changes as well).
Hey Bill two questions didn't Katelyn Tuohy run at the end of the day at Holmdel?
I'm sure you take a look at that but wouldn't it be tougher to run at the end of the day after a lot of people have already run the course
Also what is the standardized course that these times play out on? Is it Bowdoin Park in New York? Or is it just some imaginary standardized course that works for the math?
But in theory these times all do translate to performance on a standardized course, right? Sorry, I should say these ratings?
Not to be mean but your predictions are bordering on delusional! All these girls seem like cool kids that work hard so I'd like nothing but the best for all of them. But to just throw some crazy times that fly in the face of all reason evidence and history is a little annoying I can guarantee you know girl is breaking 9:45 in a 3200 M on a track. And the only way they're going to hit that is with a perfect race and almost an unhealthy dedication to running which won't serve them well in the future All of these girls would be better off learning from recent history and focusing more on just being healthy and making small progress and thinking about the future, rather than killing themselves for some meaningless high school glory , and taking a chance on injury Jennifer Hutchins ran a 950 3200 meters at a big meet with 2 Pacers. Which really translates to about a 956 at a regular High School meet To say a girl like that is going to run 20 or 26 seconds faster in a matter of months is just crazy. Girls don't constantly improve their track X, they have their ups and downs. But also there is just normal minor aches and pains and small injury, over-running, and so on that there's no guarantee that they'll necessarily be faster or that much faster And almost every girl hits a time in their life where there times decrease. I guess there could always be an exception to the rule but it's just the facts of life for most young female runners. And they would all be better off not pushing crazy hard for high school records
predictor of this wrote:
They did not have a real track season so we don't know what they would have run. I predict that 2 girls will break 9:30 this year which should solidify who the fastest XC runner all-time is at 15:58. RunninLane is fast but that doesn't mean a 9:40 girl can break 16 there.
ContextisKing wrote:
Hey Bill two questions didn't Katelyn Tuohy run at the end of the day at Holmdel?
I'm sure you take a look at that but wouldn't it be tougher to run at the end of the day after a lot of people have already run the course
Also what is the standardized course that these times play out on? Is it Bowdoin Park in New York? Or is it just some imaginary standardized course that works for the math?
But in theory these times all do translate to performance on a standardized course, right? Sorry, I should say these ratings?
The New Jersey Shore Coaches Invite at Holmdel is a big invitational ... On the day Tuohy ran 16:21, there were 27 races starting at 9:00am (13 varsity races plus JV, Frosh) ... Tuohy was in the last varsity race of day (Girls Varsity A) that went off at 3:50 pm according the race schedule ... Looking at the race videos, the course seemed to be in very good condition when Tuohy ran ... I didn't know until today, but somebody at NJ MileSplit predicted before the race that Tuohy would run 16:23 at Holmdel based on her speed rating at the Ocean State Rhode Island race two weeks prior (176 for a 16:06.87 time).
When originally devising my speed rating method, my first base-line course was the course SUNY Utica College which was the NY Section 3 Champ course at the time and would be the NY State Meet course in 2000 ... It was a relatively slow course, but worked nicely for Section 3 which has a decent diversity of big and small schools for NY (Fayetteville-Manlius, Liverpool and down to Tully and many smaller schools) ... That course set my arbitrary scale where a 26:00 time equals a 0 speed rating and every 3 seconds faster adds 1 point ... e.g. 18:00 time is a 160 speed rating, so the SUNY Utica course is slow, which turned out to be good for my purposes.
When I speed rate any race, I determine how much faster (or slower in rare situations) the race was in relation to the SUNY Utica course in terms of seconds ... Over the years, I have developed a library of base-line courses that have been correlated to SUNY Utica ... Base-line courses for individual courses, individual races (which are the most useful), and approximated base-lines for different quality of runners (e.g. average invitational, good invitational, NXN qualifier, Foot Locker qualifier, etc.) ... These base-lines are matched (profiled) with the race being speed rated to approximate the time variation from SUNY Utica ... In addition and separately, the seasonal speed ratings of the individual runners in the race being evaluated are correlated with their final times to determine the variation in speed from the SUNY Utica scale I set-up.
For nearly all races I speed rate, I add time to the final times of the runners because most races are faster (some much faster) than a base-line race at SUNY Utica.
Example ... for the AAU National Champ races in Florida this past weekend, I added 81 seconds to the boys race and 78 seconds to the girls race to calculate the speed ratings ... those adjustments came directly from a correlation of the seasonal speed ratings of nearly 150 individual runners in the races ... that determination was checked with course adjustments I found at the same course for the Florida State Meet (and Pre-State Meet), and the difference was only 3 seconds.
For the RunningLane National Champ, my adjustment was adding 108 seconds to the boys and 102 seconds to the girls ... Not part of any speed rating determination, a quick search found 28 runners who raced at both AAU and RunningLane and all 28 ran faster at RunningLane ... 3 so much faster that I excluded them as outliers, and another 3 to 6 are potential outliers ... including and not including the potential outliers yields an average difference of 23.2 to 29.8 seconds faster at RunningLane (median of 21.5 to 28.8 seconds faster).
Calling people normally indicates a problem with the name caller. Did you know that Brown ran a 9:39 TT? Did you see her 4xmile workout that indicated sub 9:40 fitness? Any rational person would look at those two data points and place a goal this year at 9:30 for her. Anything else would be insulting. Maybe she won't do it, but that should be her goal. Did you see what Thorvaldson did to her last week over 5k? So somehow a reasonable goal for Brown is 9:30 but 9:50 makes more sense for a girl who demolished her? You could say that it seems a bit fast but to not even be able to see it as a possibility shows that you have some type of bias that is not allowing you to think clearly. Maybe you are Tuohy's coach or Dad or you are just a person who desperately does not want to see records broken, but they will be. You seem to think that women can't improve. That is sexist. What if I told you that a senior in high school who comes from generations of runners and has a father and grandfather as a coach, runs 10:30 through consistently high school and I asked you how fast she would run some day. You would say that she is at her peak. Delusional would be predicting that she would run 9:45. But the fact is that she ran 9 flat this year. Impossible I know.
Fantastic stuff, thank you sir
Again, thank you Bill - your site is great for just poring through data.
Something else I thought of - was the Newbury Park performance at XCTown MOC the best boys team race ever?
Newbury '20: 191 | 191 | 189 | 185 | 181 | 179 | 177
Loudoun '18: 189 | 189 | 187 | 185 | 183 | 167 | 156
Gig Harb '13: 196 | 193 | 191 | 183 | 174 | 166 | 157
GreatOak '15: 192 | 191 | 189 | 179 | 178 | 174 | 171
F-Manlius '14: 192 | 185 | 185 | 184 | 175 | 174 | 154
The other team data was taken from NXN champs only, but I suspect Great Oak and others ran faster in their state meets. Any thoughts on all time boys teams Bill?
What is your prediction for Hutchins this Friday? I say 15:41.
I guessed in another thread that a 182 SR is equal to about a 15 minute perfect conditions track 5k. 202 equal to a 14 minute track 5k. I assume others have done the work on that before, but I haven’t seen it.
If that is accurate and her fitness hasn’t changed, if she has ideal conditions her 171 SR would translate to 15:33.
Wow. I would love to see her set the record so that she holds the XC national record and the 5k national record. I do think that Thorvaldson will go even a bit faster in the spring. But even better to have it broken by 2 different girls in the same year.
shoes shoes shoes wrote:
I guessed in another thread that a 182 SR is equal to about a 15 minute perfect conditions track 5k. 202 equal to a 14 minute track 5k. I assume others have done the work on that before, but I haven’t seen it.
If that is accurate and her fitness hasn’t changed, if she has ideal conditions her 171 SR would translate to 15:33.
I tend to place less emphasis on speed ratings under unusual race conditions as I think they can be impacted by how differently the conditions impact different runners. Some runners handle the tough conditions better than most. The results from last Year NYSPHSAA XC meet show that. If it was me, and I am not making a prediction, I would go with something in the high 160s.
I’m a D2 female runner. Our coach explicitly told us not to visit LetsRun forums.
Great interview with Steve Cram - says Jakob has no chance of WRs this year
Guys between age of 45 and 55 do you think about death or does it seem far away
2024 College Track & Field Open Coaching Positions Discussion
adizero Road to Records with Yomif Kejelcha, Agnes Ngetich, Hobbs Kessler & many more is Saturday
RENATO can you talk about the preparation of Emile Cairess 2:06