"Moneyball" Statistics for High School Track and Field

therunner711 wrote:

I am a high school track and field coach in Kentucky and I have an interesting concept for you all. What if cross country, and track and field has a one stop shop for, "Moneyball" grade statistics. Like not just plain results that we look over for hours, but stats that tell a different story. Like, "The average time of first place in the 1600 at the state meet is this". I think it would be really helpful, and no one is really doing it. I know Dark Horse Athletes is doing it, but they are only working in KY right now and only upon request. What do you guys think? Am I overthinking coaching, or are in-depth, "Moneyball" stats the future of track and field, and cross country?

I think there's something there. As a coach of a team you'd be most interested in developing a model to make incremental gains towards winning. Of course, the sample size of a handful of track meets vs a 162 game baseball season, and as someone above suggested the inherent randomness of high school athletes at various points in their development makes it much more difficult.

Caveats aside, I think the model I would want would look at past results and suggest to a coach which athletes are better off doubling or tripling at a conference meet planning for suboptimal performances in a few events and which athletes should go for the gold in one knowing that they'll only have a shot at a point or two in other events. Someone won't have three optimal performances in a day, but there are some scenarios where a few sub-optimal performances yield superior results (think a few seconds and a third vs. a first and a pair of eighths).

Troll-y Roller wrote:

[quote]therunner711 wrote:

Caveats aside, I think the model I would want would look at past results and suggest to a coach which athletes are better off doubling or tripling at a conference meet planning for suboptimal performances in a few events and which athletes should go for the gold in one knowing that they'll only have a shot at a point or two in other events. Someone won't have three optimal performances in a day, but there are some scenarios where a few sub-optimal performances yield superior results (think a few seconds and a third vs. a first and a pair of eighths).

This is what I imagined you were talking about. When I coached at a tiny 1A school, I had a girl who could potentially medal in 6 different events at the conference meet, but was limited to 4. Using a very basic analysis, it became clear that pulling her from the 4 x 800 (where we were ranked No. 1, and she was our anchor) and letting her do the long jump fresh (where she was ranked No. 3) would be worth about 6-8 points since we had a solid replacement for her. Relays just aren't that efficient. It seemed counterintuitive at the time and took some persuading, but when we won the first conference title in school history no one was complaining about being shuffled any longer (especially when the girl in question was named MVP).

Ryry wrote:

Predictions from test numbers and helping decide what event the athlete should be doing is where this could be useful or determine which is really the best event.

Bad idea. Here is a way to do that: let kids try many events!

People figure out pretty quickly if you are a sprinter or endurance: line everyone up to run 100m. Fastest kids are your sprinters. Line everyone up and run a mile, first kids to finish are your endurance athletes. The field events are all power events, but let the small kid try the shot. Who knows he might be good at it.

Plenty of charlatans out there with tests or even genetic testing (an even worse idea) that supposedly shows what sport your kid should play. Other than perhaps exposing a kid to a sport he/she was not aware of such tests are likely to have a lot of false negatives and get a kid in a sport he/she does not like.

It is the kid's journey, not your's.

inccstats.com

I did this a couple of years ago for the California distance events.

https://docs.google.com/document/d/1GbGLZylkpi7hgfen2X5EB79-l9cVWjjVdSRbLOLhKGY/edit?usp=sharing

rojo wrote:

To me if I'm a 4:15 miler and you are a 4:10 miler, that's irrelevant if I'm improving 5 seconds every year and will improve for one more year than you.

Irrelevant? You would lose to them every year until that final year when you are evenly matched. You would never be the better runner.

Ivy league education...

I've often looked at times from our state track and XC meet to give athletes something to shoot for, (ie - average time to be all state, average time to win, etc). This is how I decide if I want to 'stack' a relay team. If we have a shot at winning, we go for it. If we don't, I usually let the athletes run their individual events.

I think you are underestimating the amount of strategery track coaches are putting into their event placement of athletes to try and place or win state as a team.

runkeller wrote:

I've often looked at times from our state track and XC meet to give athletes something to shoot for, (ie - average time to be all state, average time to win, etc). This is how I decide if I want to 'stack' a relay team. If we have a shot at winning, we go for it. If we don't, I usually let the athletes run their individual events.

I think you are underestimating the amount of strategery track coaches are putting into their event placement of athletes to try and place or win state as a team.

I didn't realize that anyone who looked at numbers is now considered Billy Bean. Yes, that is strategy, but it's hardly "The Moneyball of T&F".

Now if you tell me that a All-American 1600 runners normally averages 5 points at NCAAs over the course of his career and an All-American high jumper normally averages 8 and the 1600 runner is on average offered more scholarship money, then you have found a market inefficiency a la "Moneyball".

Visited a about a dozen colleges with my daughter 4 years ago and met with coaches. Mitchell Baker at Brown was on to something like this, but he was speaking about training and injuries, that is, injury likelihood based upon training errors. Generally impressed with the whole discussion including that topic.

Daughter thought he was a good coach, didn't like anything about the school though, and went elsewhere.

I'd like to see a list of top times that have some sort of altitude adjustment. At least for Cross, 1600 & 3200. Maybe two lists, one ranked by actual time and one altitude adjusted. This may help the kids at altitude trying to get onto teams at sea-level.

I'm stealing this but for a different state.

I am a high school kid who basically has three study halls (study hall, photography, french) and I plan on starting to make some spreadsheets of stats like this? Any recommendations? Currently I am thinking average time from the first/5th/10th/20th/50th/ and 100th (maybe 200th) at each event? starting with 800. Also somehow do all 4 grades on one spreadsheet and look at avg progression of each placement?

I'm also a rep of Dark Horse Athletes. What state, and class are you from? We can find the average time/jump/throw on any place from any event over a ten year period. Just name the meet and class, and we'll find those stats for you.