Metric analysis

A new metric for NASCAR betting analysis

Handicapping NASCAR racing is all about projecting speed.

The two avenues used by bettors to project speed are:

  1. Review of track activity (practices, qualifying) during the race weekend
  2. Using past races on the current track and similar tracks

There is a lot of value in midweek betting. However, this occurs before any activity on the track. This means that we focus on the second method.

Metrics such as average green flag speed, average race position, driver rating, fastest laps, and laps led are interesting, but they often don’t tell the whole story.

Projection speed: an example

As a good example, let’s look at Tyler Reddick’s run in the Coca-Cola 600 earlier this year.

Reddick ran solidly for 322 of the scheduled 400 laps. Without a doubt, he had a car in the top 10.

However, on lap 323 the rubber buildup started building up in his tire and impacting his performance. Eventually, this caused the tire to shred, costing Reddick a lap and damaging his car. He couldn’t run at the same speed the rest of the race.

Graphically, we can see that Reddick was trailing between 30.5 and 31.25 seconds laps for most of the race.

After the rubber buildup and subsequent incident (indicated by the red arrow), Reddick was only two laps faster than 30.5 seconds over the final 91 laps, a rate of 2.2% .

Prior to the incident, he had completed 62 such laps over a period of 322 laps, a rate of 19.3%.

Additionally, Reddick had only recorded six race speed laps above 31.25 seconds prior to the incident. After the incident, he was 10.

A new speed metric: FLAGS

Obviously, in the example above, Reddick’s actual speed was affected by an incident.

Although incidents do occur, projecting them is largely beyond our control as punters. They are basically random events, so the best option is to project the actual speed.

To do this for Reddick, I simply look at his healthy lap times (before lap 323) and compare them to the rest of the field.

Doing this for each driver lap by lap gives us a fixed number of healthy cars on the track each lap. This number can obviously change, so it behooves us to adjust the time of each lap based on the size of the field of healthy cars.

When all is said and done, we get my new metric called Ffield-Llevel Aadjusted gqueen Spee, AKA, FLAGS.

Going back to Reddick’s example, NASCAR’s average green flag metric speed shows that Reddick was ninth fastest.

However, this average includes all of his times after lap 322.

Because he finished in the first lap in sixth place, largely due to attrition from others, bettors and analysts may think the whole race was indicative of his pace.

But, by performing this deeper analysis, we know that was not the case.

Using my FLAGS metric, Reddick actually placed fourth in Charlotte. This gives us a truer picture of his running pace before any speed impacting incidents.

How to use FLAGS

Each week, I will have a FLAGS ranking for each race. This weekend, as we head to Michigan, I decided to use regular season data from the 1.5 mile ovals in Las Vegas, Kansas and Charlotte.

All three are good proxies for speed in Michigan and correlate well. By averaging my FLAGS ranking on these three tracks, we get the following ranking:

At these races, Kyle Busch can be seen leading all the drivers in speed.

If we just look at other speed rankings that don’t account for post-incident speed, those rankings will be much lower for a driver like Christopher Bell, who encountered speed-impacting events in all three races. .

Nonetheless, Bell still managed two top-five and one top-10 finishes in those three races, despite not having a perfect car at the end of each race. Imagine what he could have done with a healthy car.

It is also possible to add other somewhat similar tracks in Michigan. Both Auto Club Speedway and Pocono have some correlation with Michigan in the horsepower department, but for different reasons, I’m not including either of them.

Going forward, I’ll try to look at different subsets of similar tracks to give you a complete picture of which riders should be fast each week before race weekend!

Good luck with your betting and can you take checkers…ahem, FLAGS…this week!