Fantasy Racers Should Want Qualifying Back

Fantasy-racing payoffs hinge on race outcomes. But if you are a fantasy racer — or a serious bettor — you could make money off qualifying. Not by wagering on it, but by using it as a data source.

Why Qualifying is Necessary to the Sport

COVID forced NASCAR to make a lot of changes, many of which will continue going forward. For example, limiting the number of people and days at the track saves money. One reason NASCAR could do one- and two-day shows was because they eliminated practice at all but new tracks and qualifying almost everywhere.

It seems as though NASCAR will restore qualifying (and a little practice time) in 2022. There was some enthusiasm for the ‘race what you brung’ model: you can’t wreck a car before a race if the car isn’t on the track before the race.

But there are some very strong arguments for qualifying.

  • Teams need some time to shake out the car. Maybe not three practices and qualifying, but a team needs at least a few laps to find errors in set up (or loose ballast).
  • New drivers and new driver/crew chief pairings need time to develop a common language about what the car is doing and what it needs.
  • Even metric qualifying (which is way, way better than NASCAR’s first quasi-random draw scheme) is unfair, especially to smaller teams. Without qualifying, a low-ranking team with a fast car still starts in the back.
  • I’ve shown it costs the driver a lot in lap time and race time to make it to the front. It’s also nearly impossible to win stage 1 points when you start in the back, even if you have the best car
  • There are advantages to some pit stalls. Without qualifying, the same cars get to enjoy those advantages

But drivers and race teams aren’t the only ones hampered by lack of qualifying.

Fantasy Racing Benefits from Qualifying

NASCAR’s bought big into fantasy racing and sports betting is become legal in more and more states. While I’m sure there are people who would bet on qualifying, qualifying is important to fantasy racers because it’s an important piece of data.

Racing is one of the hardest sports to bet on. The chances of a star quarterback being knocked out of the game on the first play are pretty low. The chances of a polesitter getting knocked out of a race are much higher. Weather-shortened races can also turn the finishing list upside down.

Case Study: Texas Motor Speedway

Let’s take the upcoming race at Texas Motor Speedway as a case study for planning our fantasy racing team. The driver with the best average finish in the last five races (6.2) and the best career average finish (a little over 10) is Kevin Harvick.

In the graph below, the red bar shows the last-five-race average and the blue bar is the driver’s career average at Texas.

A comparison of the career average and last-five-race average for top drivers at Texas Motor Speedway

But that’s not the only data we have. We can look at each of Harvick’s Texas races. He’s only got four finishes outside the top 20 in 36 races. And 12 top-5 finishes, including three wins.

Kevin Harvick's record at Texas Motor Speedway.

But…

All of Stewart Hass Racing has struggled this year. Look at Kevin Harvick’s rank and finishes as a function of race for 2021.

Kevin Harvick's 2021 looks really different from his 2020, when he won 9 races.

It’s not horrible. I mean, he is ranked 9th. He’s led laps at 8 of the 32 races. But he hasn’t won a race this year (yet). You’re probably not going to pick him to win at Texas, but what are his chances of getting a top 10? A top 5?

We haven’t had a points race at Texas this year, so we don’t even have any data from that to help us.

That’s where the fantasy racer needs qualifying.

Why Would Qualifying Help?

Because there is a definite correlation between qualifying position and finishing position. It’s not a strong enough correlation that you can predict where someone will finish based on where they start. That’d be boring.

I plotted qualifying position vs. finishing position for races from 1990-2019 in a bubble graph for all the tracks we run where there were at least 20 races since 1990.

Since we’re talking Texas, let’s look at Texas. Here’s a guide to the graph.

  • The location on the horizontal axis is where each driver qualified. To the left is better.
  • The vertical-axis location is where each driver finished. In this case, down is better.
  • The size of the bubble is proportional to how many times it happened.
Fantasy racers can benefit from looking at the vast amount of historical data available from places like racing-reference.info. This graph shows the qualifying positions and finish positions for all drivers who finished races at Texas from 1990 to 2019.

For example: the driver who qualified first finished first on four occasions. The driver who qualified tenth finished first only once. That’s why the bubbles are different sizes.

Most of the data are concentrated along a broad swath that goes from the lower left to the upper right. It makes sense, right? Drivers that qualify well tend to finish well. Drivers that qualify in the back often finish toward the back.

If everyone started where they finished, all the data would lie on a straight line of slope 1, which I’ve shown below in red.

This bubble graph is exactly the same as the one above (qualifying positions and finish positions for all drivers who finished races at Texas from 1990 to 2019.), but I added a slope-one line to show what a perfect correlation would look like.

There is more data below the line than above it. That’s because, almost every race, one or more drivers fail to finish. As far as our polesitters go, one DNF driver came in 32nd, while three finished 37th. (The same number of pole sitters had a DNF as who went on to win the race.

I put the DNFs in red on the graph below.

Another bubble graph for Texas, this time with the DNFs added.

DNFs are one of the elements that makes betting on racing harder than many other sports. When’s the last time one baseball team dropped out in the middle of the game?

This pattern doesn’t just hold for Texas. It holds for most tracks. It’s weaker at superspeedways and stronger at other places. Like Michigan.

A bubble graph showing finishing positions vs. qualifying positions for 60 races at Michigan International Speedway
If you didn’t see the streak running lower left to upper right before, you should definitely see it on this graph. Right?

But Does It Hold for 2020?

Not nearly as well as the pattern holds for the years in which we had qualifying. We’re at a disadvantage having only 1-4 races in the COVID era for each track. But COVID races, where qualifying is determined by algorithm, are much less correlated. We know that sometimes P10-P20 teams get a really good car at a track their driver loves. They’ve got a higher-than-their-average change of finishing well.

But without qualifying, they don’t have a chance — and you don’t have that information.

Qualifying Doesn’t Predict Finishing, But…

I know you’d like to be able to predict who will end up where, but racing is far too random to have a foolproof algorithm. There wouldn’t be any fantasy racing if such a thing existed!

We can, however, talk percentages. For those drivers who finished the race at Texas:

  • The polesitter won the race about 11.8% of the time.
  • 43.3% of the time, drivers starting in the top 5 finished in the top 5.
  • 53.8% of the drivers starting in the top 10 finished in the top 10.

Qualifying gives you valuable information and it’s the closest data you can get before the race.

It’s not just the final order, either. If you can listen in on the driver and crew chief during practice or qualifying, you tell a lot about where each of their heads are that weekend. Is Harvick showing quiet determination? Or is he frustrated and sniping at his team. Television and radio interviews can provide some of the same information — but the drivers can’t say much about how their car is handling if they haven’t had a chance to handle it.

If we had qualifying this weekend, it would give us some really additional data that would help us weigh how much we want to lean on Kevin Harvick (or any other driver) and this track.

Fantasy Racing Other Tracks

I ran the numbers on all of the tracks for which we had at least 35 races. I show below the frequency of how often a driver qualifying in the top five finishes in the top 5.

Looking at percentages -- in this case, the number of drivers who qualify in the top 5 and go on the finish in the top 5 -- can help the fantasy racer decide who to race and who to keep in the garage.

Michigan has the lowest rate, at just under 1/3 while Texas has the highest rate at 43.3%.

We can do the same thing for the top 10. Now we’re looking at around even odds that those drivers who qualify in the top ten finish in the top ten — even at superspeedways.

A vertical bar chart showing how many drivers qualified in the top 10 and finished in the top 10 for a variety of tracks.

By the time we get to top 20…

Yet another vertical bar chart, this one showing how many drivers qualified in the top 20 and finished in the top 20 for a variety of tracks.

If someone fails to qualify in the top 20, you might want to think about leaning on them too hard for this race. We don’t have that data this season. I sure hope we do next season.

Please help me publish my next book!

The Physics of NASCAR is 15 years old. One component in getting a book deal is a healthy subscriber list. I promise not to send more than two emails per month and will never sell your information to anyone.

Be the first to comment

Leave a Reply

Your email address will not be published.


*


This site uses Akismet to reduce spam. Learn how your comment data is processed.