What's up with Milesplit downgrading races from 5k to 3miles?
09/28/2018 2:41:38 PM
User
SUBSCRIBER
Joined: Feb 2011
Posts: 13
I've seen quite a few races this year drop from 5k to 3 mile via Milesplit's leader boards, IE: Westlake Chaparral Invite and Cedar Park Invitational. I'm fairly certain I heard the announcer at Cedar Park call out the course was certified. What gives? I hope Milesplit admins are not arbitrarily deciding which races are 5k's vs 3 mile. Anyone else seeing this?
I've seen quite a few races this year drop from 5k to 3 mile via Milesplit's leader boards, IE: Westlake Chaparral Invite and Cedar Park Invitational. I'm fairly certain I heard the announcer at Cedar Park call out the course was certified. What gives? I hope Milesplit admins are not arbitrarily deciding which races are 5k's vs 3 mile.

Anyone else seeing this?
09/28/2018 5:16:47 PM
Coach
SUBSCRIBER
Joined: Sep 2004
Posts: 2553
@NIKEPRE We need a classification system showing which courses are long/short. I hate going to courses and getting times that are blown out of proportion and set unrealistic expectations for the rest of the season. I'm glad they seem to be filtering it out a little bit but I would be interested to see how they figured out how long the races were. A lot of people SWORE Southlake was short last year but it was just perfect conditions. Would be nice to have a database/accountability factor.
@NIKEPRE

We need a classification system showing which courses are long/short. I hate going to courses and getting times that are blown out of proportion and set unrealistic expectations for the rest of the season. I'm glad they seem to be filtering it out a little bit but I would be interested to see how they figured out how long the races were.

A lot of people SWORE Southlake was short last year but it was just perfect conditions. Would be nice to have a database/accountability factor.
09/28/2018 9:15:12 PM
Power User
SUBSCRIBER
Joined: Apr 2004
Posts: 1784
@stickstickstick speed ratings are one way to do it. Check out tullyrunners.com I use delta from course to course with common runners. We have only had 1 sub 20 all year, but head to heads are encouraging so I don'tget caught up in times or course lengths. Houston courses are trending to be a smidge long and muddy this year, though. No biggie. It's xc. McNeil/state runs like a 3.05. Always a little quick.
@stickstickstick speed ratings are one way to do it. Check out tullyrunners.com

I use delta from course to course with common runners. We have only had 1 sub 20 all year, but head to heads are encouraging so I don'tget caught up in times or course lengths. Houston courses are trending to be a smidge long and muddy this year, though. No biggie. It's xc.

McNeil/state runs like a 3.05. Always a little quick.
09/29/2018 7:14:16 AM
Coach
SUBSCRIBER
Joined: Apr 2012
Posts: 5
@NIKEPRE Hi, The Cedar Park coach notified all us coaches and the Milesplit staff that the course had to be modified at the last minute due to the heavy rains and mud which resulted in the course only being 3 miles. As a coach, I was bummed too, but I am Happy that they updated the results to reflect the true state standings and for us, school and personal records. Hopefully our kids are motivated to run even faster today at McNeil!
@NIKEPRE

Hi,

The Cedar Park coach notified all us coaches and the Milesplit staff that the course had to be modified at the last minute due to the heavy rains and mud which resulted in the course only being 3 miles.

As a coach, I was bummed too, but I am Happy that they updated the results to reflect the true state standings and for us, school and personal records.

Hopefully our kids are motivated to run even faster today at McNeil!
09/29/2018 3:12:03 PM
User
SUBSCRIBER
Joined: Jul 2013
Posts: 26
@Aggietom Been creating my own speed ratings based on information at [url=http://www.tullyrunners.com/articles/MakeOwnSR.htm]http://www.tullyrunners.com/articles/MakeOwnSR.htm[/url] along with the dozens of other articles he has posted. Started two or three years ago on small scale but the time to do ratings for just a couple meets was onerous and not accurate, so this year I went all in and built more of a repeatable system with lots of data and now can pretty easily create them. I have rated about 40 Texas courses this year. I have no clue how Bill @ TullyRunners rates as many races as he does, he must have a beautiful database set up plus his years of experience, and I am sure he must spend all weekend at it. As for the Westlake Chapperal course I calculated a course correction factor of 127 so I would certainly believe it was shorter than 5k, probably significantly so. For reference the State course last year had a course correction of 93 and the Texas State course was one of the "easiest" in the nation when back calculating Tully's state meet data and comparing. A track 5k roughly comes out to 90-100 (according to my math) depending on weather and other conditions. Last year Southlake was 111, so yes I bet Southlake last year was probably a little short, but if it was perfect conditions, then maybe not by much. McNiel in 2016 was 108 according to Tulley, this year my preliminary calculations show it to be significantly higher than that, so we will see where it lands, Tulley normally rates this meet so I can cross check with my numbers, it's weird that it is so different from the state meet but they must set the flags a little tighter at McNeil, and the McNeil vs State times seem to support that. Post here and I can check other course corrections to help with 3 mi vs 5k debates. Its not a perfect gauge for distance but it’s something. Course difficulty and weather obviously still play a role as well (i.e. a rough hilly course could be very short but would still have a low course correction.)(And of course I am still perfecting my system, and don’t have near the data built up like Tully so my numbers are rough especially the early season ones when the data was especially sparse)
@Aggietom Been creating my own speed ratings based on information at http://www.tullyrunners.com/articles/MakeOwnSR.htm along with the dozens of other articles he has posted. Started two or three years ago on small scale but the time to do ratings for just a couple meets was onerous and not accurate, so this year I went all in and built more of a repeatable system with lots of data and now can pretty easily create them. I have rated about 40 Texas courses this year. I have no clue how Bill @ TullyRunners rates as many races as he does, he must have a beautiful database set up plus his years of experience, and I am sure he must spend all weekend at it.

As for the Westlake Chapperal course I calculated a course correction factor of 127 so I would certainly believe it was shorter than 5k, probably significantly so. For reference the State course last year had a course correction of 93 and the Texas State course was one of the "easiest" in the nation when back calculating Tully's state meet data and comparing. A track 5k roughly comes out to 90-100 (according to my math) depending on weather and other conditions.

Last year Southlake was 111, so yes I bet Southlake last year was probably a little short, but if it was perfect conditions, then maybe not by much. McNiel in 2016 was 108 according to Tulley, this year my preliminary calculations show it to be significantly higher than that, so we will see where it lands, Tulley normally rates this meet so I can cross check with my numbers, it's weird that it is so different from the state meet but they must set the flags a little tighter at McNeil, and the McNeil vs State times seem to support that.

Post here and I can check other course corrections to help with 3 mi vs 5k debates. Its not a perfect gauge for distance but it's something. Course difficulty and weather obviously still play a role as well (i.e. a rough hilly course could be very short but would still have a low course correction.)(And of course I am still perfecting my system, and don't have near the data built up like Tully so my numbers are rough especially the early season ones when the data was especially sparse)
09/29/2018 4:38:24 PM
Power User
SUBSCRIBER
Joined: Apr 2004
Posts: 1784
If you have a system, post the results here weekly because people are starved for data. Good to hear others are in on this.
If you have a system, post the results here weekly because people are starved for data.

Good to hear others are in on this.
09/29/2018 6:42:11 PM
User
SUBSCRIBER
Joined: Feb 2011
Posts: 13
@stickstickstick Agreed. I am thinking Milesplit could work with meet organizers to certify a course prior to the race so we know it's been vetted. A tall task, I know, but it could give more credibility to courses that are truly the distance advertise.
@stickstickstick Agreed. I am thinking Milesplit could work with meet organizers to certify a course prior to the race so we know it's been vetted. A tall task, I know, but it could give more credibility to courses that are truly the distance advertise.
09/29/2018 6:44:25 PM
User
SUBSCRIBER
Joined: Feb 2011
Posts: 13
@LunaRun I love the idea of Speed Ratings. I prefer algorithms since I know it's not really accurate to compare a 5k course at one place to another. You can also apply these outside of Texas to see how we'd fair against other states. I'm sure this has probably been around a while, but either way, thanks for sharing. If you have other material regarding how you derive the ratings, please post.
@LunaRun I love the idea of Speed Ratings. I prefer algorithms since I know it's not really accurate to compare a 5k course at one place to another. You can also apply these outside of Texas to see how we'd fair against other states. I'm sure this has probably been around a while, but either way, thanks for sharing. If you have other material regarding how you derive the ratings, please post.
09/30/2018 10:37:13 AM
Power User
SUBSCRIBER
Joined: Apr 2004
Posts: 1784
Nike South and McNeil speed ratings from Meylan. [url=http://www.tullyrunners.com/XC2018/Ratings_Sept29.htm]tullyrunners.com[/url]
Nike South and McNeil speed ratings from Meylan.

tullyrunners.com
09/30/2018 11:24:01 AM
Power User
SUBSCRIBER
Joined: Aug 2016
Posts: 13
Can someone explain this correction factor to me? Does anyone know of someone who does the 2 mile/3200 races? TAMUCC had some ridiculously fast times this year for the teams I've been watching and I felt suspicious of every runner on the team taking over 30 seconds off their time from the week prior. There's been a couple of other races... Covington and Graford especially. Does anyone have any idea how the number of runners in a race affects times? I've noticed from year to year a race might have a huge difference in participants and I wonder if it makes a difference. Since I have no idea what a course looks like it's hard to be sure. Mae Simmons is a prime example for us though because in a huge race you have to know that the race funnels you into a narrow path up a hill and across it about 300m after the start. And if you're not out front then you're stuck behind a bunch of people for quite a while. Do these correction factors take these things into account?
Can someone explain this correction factor to me? Does anyone know of someone who does the 2 mile/3200 races? TAMUCC had some ridiculously fast times this year for the teams I've been watching and I felt suspicious of every runner on the team taking over 30 seconds off their time from the week prior. There's been a couple of other races... Covington and Graford especially.

Does anyone have any idea how the number of runners in a race affects times? I've noticed from year to year a race might have a huge difference in participants and I wonder if it makes a difference. Since I have no idea what a course looks like it's hard to be sure. Mae Simmons is a prime example for us though because in a huge race you have to know that the race funnels you into a narrow path up a hill and across it about 300m after the start. And if you're not out front then you're stuck behind a bunch of people for quite a while. Do these correction factors take these things into account?
09/30/2018 12:40:23 PM
Power User
SUBSCRIBER
Joined: Apr 2004
Posts: 1784
@Riggers30 It depends on who is doing them. I do it on delta vs. previous or "known" course. So a SHSU vs. CFISD comparison might have been +90s for girls near the front +1:45 for midpack and 2:00+ for girls in the back. Don't use those numbers as official ones, but I've noticed and take into account the following: 1) Does the course penalize runners for a slow start (this is usually a tight path not far in or heavy mud.) 2) Does the field size effect runners? Large fields tend to affect slower runners more. Some of them are better when they're in the lead and mentally don't deal as well with being a fish in a stream. They also run longer races (not significantly, but it does add some distance) because they run midline (due to said "stream of fish" effect) vs a tangential path that a frontrunner takes. Backpackers are also in their own world...often looking down, etc. 3) Is there any emotion or extra motivation for a team at this particular race? There were a couple of early showdowns where a few teams were trying to make statements. They ran better. Same thing happens later at post-season meets. Some teams get to state and go for it and others are happy to be there and run like it. Same for NXN. Same goes for that 7th guy or girl who barely made the varsity roster. A lot of times the JV kid runs faster. Why? Because they're a) frontrunning (energizing, etc) and b) don't have the pressure and comparative mental state that a 7 would have in a varsity race. 4) Environmental factors - weather, humidity, heat, etc. The delta for course to course usually controls for most of those, though.
@Riggers30 It depends on who is doing them.

I do it on delta vs. previous or "known" course.

So a SHSU vs. CFISD comparison might have been +90s for girls near the front +1:45 for midpack and 2:00+ for girls in the back.

Don't use those numbers as official ones, but I've noticed and take into account the following:

1) Does the course penalize runners for a slow start (this is usually a tight path not far in or heavy mud.)

2) Does the field size effect runners? Large fields tend to affect slower runners more. Some of them are better when they're in the lead and mentally don't deal as well with being a fish in a stream. They also run longer races (not significantly, but it does add some distance) because they run midline (due to said "stream of fish" effect) vs a tangential path that a frontrunner takes. Backpackers are also in their own world...often looking down, etc.

3) Is there any emotion or extra motivation for a team at this particular race? There were a couple of early showdowns where a few teams were trying to make statements. They ran better. Same thing happens later at post-season meets. Some teams get to state and go for it and others are happy to be there and run like it. Same for NXN. Same goes for that 7th guy or girl who barely made the varsity roster. A lot of times the JV kid runs faster. Why? Because they're a) frontrunning (energizing, etc) and b) don't have the pressure and comparative mental state that a 7 would have in a varsity race.

4) Environmental factors - weather, humidity, heat, etc.

The delta for course to course usually controls for most of those, though.
09/30/2018 12:42:42 PM
Power User
SUBSCRIBER
Joined: Apr 2004
Posts: 1784
btw this is why I don't care about course lengths. I was asked by a few people yesterday if Nike South was short. Who cares? I know Natalie Cook is a good runner and she ran x:xx and if our girls are only xx behind her then it's all good! Have y'all ever noticed that Houston area meets are all slower than meets elsewhere? Not sure if it is course conditions or what. But it isn't necessarily team quality - Tompkins boys just went up and won McNeil.
btw this is why I don't care about course lengths. I was asked by a few people yesterday if Nike South was short. Who cares? I know Natalie Cook is a good runner and she ran x:xx and if our girls are only xx behind her then it's all good!

Have y'all ever noticed that Houston area meets are all slower than meets elsewhere? Not sure if it is course conditions or what. But it isn't necessarily team quality - Tompkins boys just went up and won McNeil.
09/30/2018 2:22:34 PM
Coach
SUBSCRIBER
Joined: Sep 2012
Posts: 78
@Riggers30 TAMUCC changed the course from previous years. Then, with the heavy rains making one section not runnable, they had to cut out 300m. The race distance this year was approximately 4700m.
@Riggers30

TAMUCC changed the course from previous years. Then, with the heavy rains making one section not runnable, they had to cut out 300m.

The race distance this year was approximately 4700m.
09/30/2018 8:14:07 PM
User
SUBSCRIBER
Joined: Jul 2013
Posts: 26
@Riggers30 CC Islander Splash I calculated a course correction at 127 so the course was pretty fast. Makes since if it was actually 4700. For Tully’s speed ratings, Course correction is basically the difference between what the “average” runner would have ran at the "standard" race which for Tully’s method is the SUNY Utica Tech 1999 sectionals (Nov. 6, 1999). So basically, if there is a course correction of 100 for the State Meet, a runner that ran 20:00 minutes at SUNY would have gotten a time of 18:20 at if doing the same performance at the State Meet. What the “standard course” is really isn’t important though and is arbitrary. The important thing is that every course gets related back to one common standard for comparison. So if there are two courses one that is 126 vs one that is 100 the first is on average 26 seconds faster. The only additional thing to remember is that weather and other conditions also apply, so you could have a boys race in cool weather in clouds, then a big thunderstorm, and then a hot muddy race, which could result in dramatically different course corrections for the girls race.
@Riggers30

CC Islander Splash I calculated a course correction at 127 so the course was pretty fast. Makes since if it was actually 4700.

For Tully's speed ratings, Course correction is basically the difference between what the "average" runner would have ran at the "standard" race which for Tully's method is the SUNY Utica Tech 1999 sectionals (Nov. 6, 1999). So basically, if there is a course correction of 100 for the State Meet, a runner that ran 20:00 minutes at SUNY would have gotten a time of 18:20 at if doing the same performance at the State Meet.

What the "standard course" is really isn't important though and is arbitrary. The important thing is that every course gets related back to one common standard for comparison. So if there are two courses one that is 126 vs one that is 100 the first is on average 26 seconds faster.

The only additional thing to remember is that weather and other conditions also apply, so you could have a boys race in cool weather in clouds, then a big thunderstorm, and then a hot muddy race, which could result in dramatically different course corrections for the girls race.
09/30/2018 8:45:30 PM
Coach
Joined: Nov 2015
Posts: 46
@Aggietom Couldn't agree more. I base our performances on our spread and an athletes individual placement. In my opinion if your athlete is used to placing at our near the front each week, then he or she should be able to do that during the post season because they have conditioned themselves for that outcome. Time is so variable because of the significant number of factors that take place each weekend. For example, this past weekend at Aldine my number 1 male finished 18th but ran 30 seconds slower than last year on the same course. However, last year he placed 38th. The field was slightly smaller, but I was much more satisfied with his individual placement because it created a domino affect for our other varsity athletes to place higher individually.
@Aggietom Couldn't agree more. I base our performances on our spread and an athletes individual placement. In my opinion if your athlete is used to placing at our near the front each week, then he or she should be able to do that during the post season because they have conditioned themselves for that outcome.

Time is so variable because of the significant number of factors that take place each weekend. For example, this past weekend at Aldine my number 1 male finished 18th but ran 30 seconds slower than last year on the same course. However, last year he placed 38th. The field was slightly smaller, but I was much more satisfied with his individual placement because it created a domino affect for our other varsity athletes to place higher individually.
09/30/2018 9:23:22 PM
User
SUBSCRIBER
Joined: Jul 2013
Posts: 26
[quote=Aggietom]Nike South and McNeil speed ratings from Meylan. [url=http://www.tullyrunners.com/XC2018/Ratings_Sept29.htm]tullyrunners.com[/url][/quote] @aggietom Looks like Tullys course corrections came out to 105 for the boys and 108 for the girls, which are the exact numbers he used for 2016. (in his documentation he says sometimes he uses his previous values, which is what I think he must be doing here) [url=http://www.tullyrunners.com/Data/Articles/EarlySeasonSpeed.htm]http://www.tullyrunners.com/Data/Articles/EarlySeasonSpeed.htm[/url] Basically my data came out at 124 so 16 seconds faster than Tully (so take off 5 or 6 pnts or so of Tully's numbers) and my data is pretty robust in this case, so I very confident the runners are being over rated (at least compared to my data set, which is necessarily "calibrated" to Tully, but the other courses he calculated that I also calculated have all been within a pnt or 2 (3-6 seconds). With the exception of Woodbridge which I think he was overrating runners by about 4 pnts. Woodbridge was interesting because there were 16 common girl runners with the Friday Night Lights race in Houston a few weeks before. When you directly compared those 16 results it suggested that the girls Friday Night Lights winner would have been very competitive at that meet, I am bias to Houston runners. As for Nike my data is a lot closer to Tully's but I actually think the girls are underrated in his numbers so add a point or two to the girls results for comparative purposes. (though my data for Nike isn't as Robust as I didn't get good agreement between the boys and girls, anyone know if the lighting distrupted that meet? or any other potential change between the boys and girls races?) With respect to Houston races being slower, if anybody's is worried about the Prowl in Park results this weekend being fast, don't be, my guess is that course was less than 3 miles, rendering the times meaningless.
Aggietom wrote:
Nike South and McNeil speed ratings from Meylan.

tullyrunners.com


@aggietom Looks like Tullys course corrections came out to 105 for the boys and 108 for the girls, which are the exact numbers he used for 2016. (in his documentation he says sometimes he uses his previous values, which is what I think he must be doing here) http://www.tullyrunners.com/Data/Articles/EarlySeasonSpeed.htm

Basically my data came out at 124 so 16 seconds faster than Tully (so take off 5 or 6 pnts or so of Tully's numbers) and my data is pretty robust in this case, so I very confident the runners are being over rated (at least compared to my data set, which is necessarily "calibrated" to Tully, but the other courses he calculated that I also calculated have all been within a pnt or 2 (3-6 seconds). With the exception of Woodbridge which I think he was overrating runners by about 4 pnts.

Woodbridge was interesting because there were 16 common girl runners with the Friday Night Lights race in Houston a few weeks before. When you directly compared those 16 results it suggested that the girls Friday Night Lights winner would have been very competitive at that meet, I am bias to Houston runners.

As for Nike my data is a lot closer to Tully's but I actually think the girls are underrated in his numbers so add a point or two to the girls results for comparative purposes. (though my data for Nike isn't as Robust as I didn't get good agreement between the boys and girls, anyone know if the lighting distrupted that meet? or any other potential change between the boys and girls races?)

With respect to Houston races being slower, if anybody's is worried about the Prowl in Park results this weekend being fast, don't be, my guess is that course was less than 3 miles, rendering the times meaningless.
09/30/2018 10:41:50 PM
Coach
SUBSCRIBER
Joined: Sep 2004
Posts: 2553
@LunaRun My kids GPS's averaged 2.96 - was a fun meet regardless and the park/footing was great. Reports from my group were that there were spots where the course was being cut by runners. I encourage anyone to go to this meet for those not running down south.
@LunaRun

My kids GPS's averaged 2.96 - was a fun meet regardless and the park/footing was great. Reports from my group were that there were spots where the course was being cut by runners.

I encourage anyone to go to this meet for those not running down south.
10/01/2018 2:43:15 PM
User
SUBSCRIBER
Joined: Apr 2013
Posts: 4
@LunaRun where do you find the actual course ratings for MCNeil and Nike South on Tully’s site? This speed rating system has really peaked my interest. Want to learn more.
@LunaRun where do you find the actual course ratings for MCNeil and Nike South on Tully's site? This speed rating system has really peaked my interest. Want to learn more.
10/01/2018 5:04:53 PM
Power User
SUBSCRIBER
Joined: Apr 2004
Posts: 1784
@rschoppe they're linked in my post above.
@rschoppe they're linked in my post above.
10/01/2018 6:29:12 PM
User
SUBSCRIBER
Joined: Apr 2013
Posts: 4
@Aggietom i see the individuals, but not the overall course rating. I may be missing it though, not sure, sorry if I am. I will look again.
@Aggietom i see the individuals, but not the overall course rating. I may be missing it though, not sure, sorry if I am. I will look again.

You must be logged in to comment.

Click Here to Log In.