Trail Running in Texas & beyond

South Texas Trail Championship Series

For full details about the series go here:

This past spring Joe start talking about wanting to kick off system to reward runners who ran multiple Tejas Races.  There was already the Tejas 300 (Cactus 100m, Bandera 100k, Rocky 100m) and the Tejas 250 (all five 50 mile races) but how could we create a series that spanned the whole year and multiple distances?  We noodled over various ideas and tried to balance a few concepts.

  1. Normalize for course difficulty.
  2. Be interesting for all level and age of runners.
  3. Reward the longer distances.
  4. Incorporate the whole running season as, obviously, Joe has a vested interested in more people running his race.  🙂

After a few conversations, I got started on coming up with a proposed system.  There are 3 main scoring systems for any race series:  place based, time based, and point based.

Place based systems work great for something like track & field teams competitions.  There is a limited depth to the entry field so it is easy to score the top 8 and move on.  It doesn’t work as well with fields of 700+.  It is also complex if you have to start scoring each age and gender group.  Then there is the issue of many different races and distance to choose from.   You are going either going to have to limit the number of scoring races to encourage runners competing head-to-head or introduce other formulas.  Ultimately, it works best when everyone runs the same, limited number of races.

Time-based system only work if everyone is running the same races.  Otherwise, you have the issues of different distances, difficulty, etc.

Point based systems looked like the way to go, so then it was just a matter of designing a system that is balances certain goals.  Percentage of time was a pretty obvious direction.  Percentage works well for a number of reason, primarily it normalized the course difficulty and weather conditions.  The winning time on that given day sets the standard that everyone else then is judged on.  Heat, rain, mud get accounted for in the times.   It would then allow people to run different courses and somewhat comparable resutls.  It is also easy to score across all the different divisions.  You don’t have score each group separately since everyone is compared against the same benchmark, if the age group leader scores 80 points, he still is ahead of who ever is behind him.

There are a number of points based systems out there including Xterra, Skyrunning, Bellingham Trail Series, and Northeast Trail Series.  I was surprised to see how consistent the distribution was across races.  The 20th place person in the 2012 / 2011 races I modeled were generally consistent in terms of percentage.  There is a bit more fluctuation near the top and some races like Hells Hills 25k are super top weighted where the top 20 are very close in times but overall it distributes across races and distances fairly consistently.

The next question was distance factor.  The first model I looked at was to multiple by distance.  100m = 100 points.  50k = 31 points.  I liked how this rewarded those who ran long.   However, as I scored out a mock season, the issue that came up was in some age categories could be dominated by running a handful of races if they were run at the longest distances.   For the under-30 and 40-50 groups, the depth is fine but at 50-60 or up, did we want to reward someone who ran 4 races at the longest, or some who who ran 7 races at a different mix?  Joe and I talked this through and we came up with distance factor where distance matters but not over-weighted.

This is the first year, so we’ll see how this model works out when the data is in at the end of the year.  We may find that we want to reward the 100 mile / 100k distance a bit more in the future, for example.

Here is two scoring examples.  1 is the median place runner, running the max distances, and the other is the 10th place runner running a mix of race races, but only 1 50 miler and no 100 milers.

  Winning Time   Median Time % factor   total points
Bandera 100k 8h 28m 15h 12m 55.70% 90 50.1
Rocky 50m 6h 15m 11h 34m 54.03% 80 43.2
Nueces 50m 6h 30m 11h 33m 56.28% 80 45.0
Hells Hills 50m 7h 35m 11h 33m 65.66% 80 52.5
Pedernalas 60k 5h 5m 7h 55m 64.21% 70 44.9
Cactus Rose 100m 17h 34m 28h 51m 60.89% 100 60.9
Wild Hare 50m 7h 2m 11h 13m 62.70% 80 50.2
Series Total         346.9
  Winning Time   10th place time   % factor total points  
Bandera 50k 4h 4m 4h 38m 87.77% 60 52.7
Rocky 50m 6h 15m 8h 4m 77.48% 80 62.0
Hells Hills 50k 4h 5m 4h 54m 83.33% 60 50.0
Pedernalas 60k 5h 5m 6h 37m 76.83% 70 53.8
Mule Shoe 60k 5h 59m 7h 15m 82.53% 70 57.8
Reveille Ranch 60k 5h 45m 7h 32m 76.33% 70 53.4
Wild Hare 50k 4h 26m 4h 59m 88.96% 60 53.4
Series Total         383.0

It is interesting how these two theoretical runners are at the end of the series.  Obviously, the best approach is to be Steven Moore and go out and win 6-7 races at the 100 and 50 distance but absent that, the system rewards a variety of running mixes.

A few notes:  the median place at Rocky 100 is about 10 percentage points lower than at Cactus Rose 100.  The fast course plus elite runners really spreads out times.  It won’t get any easier with Mike Morton, Karl Meltzer and others coming to Rocky this year.  You would have running about a 20 hour Rocky to get the same points as running a 29 hour Cactus.  If you are picking your races, some races do have slightly different biases.  For example, running Bandera 100k, Rocky 100m, Nuece 50m, you are going to have to accept a lower percentage for a mid-pack runner due to elites showing up.   But here is the thing, you are not competing against them, they won’t be part of the year end scoring so you really are competing with different set of runners.  But it does set the curve which has its impact.  It will be interesting to see how it plays out by year’s end.  I still think you are better off running the longer distances but a mid-pack runner at Bandera won’t score that many less points in the 50k due to the smoking times at the top of the 100k when the likes of Timmy Olsen, Dave Mackey, and Erik Stanley are toeing the line.  A slightly above median runner will score more points in the Rocky 50 miler than the median or below runner in the Rocky 100m.  I am talking about places and not tracking one runner over the course of a season in the above example.  In my modeling, I picked runners and tracked them for a whole season and the “Peter Bray” runner scored very close to the “Dave Brown” runner.

Joe and I are committed to revisiting the system to ensure it provides a good balance and to address any rules that need to be adjusted as we learn more about the system plays out over the course of the year.  Running different scenarios is a poor substitute to real life data and interaction (unless you have more time and money then I do).

Scoring the sub-ultra category is a bit different.  You definitely are best off running the 25k / 30k distance.  You’d be hard press to win by sticking with the 10k only.

One Response to “South Texas Trail Championship Series”

  1. admin says:

    Here is a reply to a message I received which may provide additional color:

    Using the median runner example may not work well in my writeup because someone who in the middle in one race may be closer or further from the lead in another. The issue with Rocky is that the winning time is a deviation is higher from the race of the pack thanks to people like Hal or Ian. So the person at Rocky who is the median (exact middle) is further away percentage wise than at Cactus. Don’t think of it in terms of hours back but percentage. If Mike Morton runs near 12 hours this year, 20 hours would be 60%. If Steven runs 17 hours next year, 29 hours would be 59%.

    Here are some examples:

    Steven: 75 points (10th of 158)
    Peter: 46 points (99th of 158)
    David Brown: 60 (1st 50k)

    Scott Rabb: 73 points (11th of 218)
    Nick Polito: 57 points (77th of 218)
    Peter: 45 points (211 of 218)
    David Brown: 77 points (2nd in the 50)

    Steven: 100 points
    Peter: 52 points (44 of 47)
    Nick: 56 points (18th of 128)

    Now keep in mind Peter isn’t your median runner, he more of a back of the pack runner. I tossed Nick in there as he as another data point.

    The idea on the series was to reward quality running, regardless of the race but to provide a bonus for doing so at longer distances. If you are like David Brown and win most of the 50m and 50k races you enter, you have a great shot at placing high overall but you will still lose the someone like Steven if he runs enough races. Heck, David would probably be outscored by those who generally finish in the 5 slots behind Steven, if they run enough. The issue is that most of those top finishers don’t run 7 races throughout the year. Someone like Steven may be able to sneak a out title with only 5 races, but it would depend on who else is running, a Jeff Miller type probably could overtake him with 7 races, depending on the mix.

    To me, I think the overall Male and Female will sort out pretty naturally and at the end of the year the top 5 will be pretty much the top 5 you’d expect. A lot depends on who runs what mix of races. Rachel Ballard would score well over the course of the year simply because she is top 3-5 and runs the longest races. However, Melanie Fryer would probably win even if she did nothing but run a mix of 2 50s, 2 60ks and a 50k simply because she’d win them all. Which resume should win? Mostly 3rd to 5th place finishes at the longest distance and 1 win or 5 wins? It depends on the goals of the series. If we were overly weighted towards the longer distances someone who finished say mid-front in all the races at the longest distance would beat someone who won everything but never ran a race over 60k. For this year, we are trying to make the series meaningful for the both sets of runners. It isn’t easy and I am sure there will be tweaks as we learn.

    As for course record bonus, we’ll see on the 25%. It is pretty much average for other series that I researched, some are higher. But yes, there should be a reward for course records and it should be meaningful but the question in my mind is whether the records are stout enough that breaking them for it to be meaningful. That is why there is a rule that that race has to have some bit of history. There were 3 course records set last year by people who may have placed in the series: Steven x2 (Ped Falls, Cactus), Melanie (Ped Falls). All other course records where set by people who were non-locals and only ran 1 race.

    On the loop course races, there are none on the schedule for now so it is moot point.