Competitiveness in IndyCar

This year’s IndyCar championship came right down to the wire. There were multiple drivers in the hunt heading into Sonoma and Josef Newgarden ended up winning the title by just 13 points. Helio Castroneves, who finished fourth in the championship, was less than 50 points behind Newgarden when the checkered flag flew. This was a highly contested championship, and it got me thinking about competitiveness within an IndyCar season. There are definitely seasons where there is a dominant driver and ones where multiple drivers are battling it out all season and everyone is close.

To answer this question of competitiveness, I first had to decide how to measure it. I wanted to look at how competitive the top half of the field was in a given year. I chose the top half to eliminate the problem that some seasons had many lower-level drivers with a lot of DNFs, and account for the fact that different seasons ran with different numbers of drivers. This made the process easier and more valid in my opinion. In the future I might look at competitiveness across the entire field if I come up with a good way to deal with the differing number of drivers and the DNF problem. Looking at the top ten competitiveness gives us a good idea of how hotly contested the championship is and how the races are likely to turn out — plus, most of the important action throughout races happens in the top ten. Less competitive seasons will see repeat winners and more straightforward races for the most part. Since realistically the entire field does not have a shot at winning the title, looking at the top ten is sufficient for our discussion.

I used a measure similar to the one The Stats Zone used to measure competitiveness across soccer leagues in Europe to measure competitiveness in IndyCar.

I took the total number of points the champion of that season scored and subtracted from it the number of points the 10th place driver in the championship scored. I took this number and divided it by the total possible points a driver could have scored in that season to account for the different points systems and number of races across seasons. This gives us a ratio from 0 to 1, where 0 would mean a perfectly competitive season (all drivers score the same number of points) and 1 would mean a perfectly non-competitive season (there is one driver winning every point in every race and other drivers aren’t competing). Obviously both of these scenarios are basically impossible, but they represent the extremes and what we mean by perfectly competitive and perfectly non-competitive.

Here is a chart of all of the competitiveness ratios from 2000-17:

By competitiveness ratio, 2015 was the most competitive IndyCar season. In this season, Scott Dixon and Juan Pablo Montoya ended up tied for the points title after Sonoma, and the former won it on the tiebreaker. Three drivers were within 100 points of Dixon and it was a hard fought championship until the end. The ’13 and ’12 seasons fall in spots two and three in competitiveness. These years saw four and five drivers respectively finish within 100 points of the champion, and ’12 was won by just three points by Ryan Hunter-Reay. The least competitive IndyCar season was 2001 when Sam Hornish Jr. won the championship by over 100 points, finishing on the podium ten times in the 13 race season. 2002 was also an interesting year: while Helio Castoneves was only 20 points behind Hornish Jr., who won the championship again, the rest of the top ten dropped off quickly after that, giving the season a competitiveness ratio of 0.32.

Looking at competitiveness ratio by season, we can see an interesting trend. Right around 2012 there was a drop off and the series has been more competitive since. This could be because of the introduction of the Dallara IR-12 chassis at the start of the 2012 season. The new chassis might be have leveled out the competition more and be the reason for the drop off we see. Aero kits were introduced in 2015, but there wasn’t much of a change from the two years prior in terms of competitiveness. It’ll be interesting to see how the universal aero kits being introduced for the 2018 season will change the competitiveness of the series.

Competitiveness in IndyCar is an interesting topic and I’d definitely like to explore it more in the future. Finding a way to incorporate full field competitiveness is a goal of mine and I’ll be posting more about the subject in the future.

Read part two of this series on competitiveness in IndyCar here.

by Drew

Updated 11/17: Fixed an error in allocating bonus points for certain seasons. The least competitive season is now 2001 instead of 2008. The chart and article have been updated with the small change.

Power Cleared For Sonoma

Most people figured this would be the case, especially with two weeks before the final race, but now it’s official. Will Power was cleared after having concussion-like symptoms and will race in Sonoma.

The important parts about all of this are (a) Power is okay and (b) he can participate in the test on Thursday at Sonoma. That’ll not only help us get an early indication of how this final race is likely to play out, but it’ll also give Power a chance to find that extra speed he’ll need to take the title fight to Pagenaud.

by Drew

Power Needs Pagenaud To Pull a Montoya

There’s just a 16 percent chance that the points lead (and championship) will shift to Power in the last race of the year. Pagenaud has looked strong all season leading wire to wire and sits with a 43 point lead heading to Sonoma. That’s the fourth biggest lead a driver has held heading into the last race in the past eleven years. Of the drivers who had a larger points lead, only one ended up losing the title: Montoya in 2015.

I went back and looked at the results of the past eleven championships, noting who was leading the points heading into the final race and who ended up winning. The seasons highlighted in yellow are ones in which the driver who led the points before the last race didn’t win the title.

In all seven “yellow” seasons bar 2015, the gap before the last race was under 20 points. Montoya’s loss in 2015 was a fluke, not the norm — at least not yet. Double points in the final race is still a fairly new concept and it’s yet to be seen if it will change this theme. That’s bad news for Power who needs to come back from a deficit closely resembling Dixon’s last year. He’ll be looking for Pagenaud to either (i) get into some sort of crash and drop back or (ii) have a DNF. While both of these are seemingly long-shot scenarios to happen, they’re more likely than Pagenaud simply having a bad race on his own.

A lead of 20 or more points has been safe four times out of five. Even when you look at the 2015 incident, Montoya didn’t lose the championship outright. It took a tiebreaker to declare Dixon the winner. If the championship ended in a tie this year, here’s what would happen:

If Power or Pagenaud won the race, that driver would have the most first place finishes and win the championship.
If neither driver wins the race and Power doesn’t come in second, Pagenaud wins.
If neither driver wins the race and Pagenaud comes in second, he wins.
If neither driver wins the race and Power comes in second, Pagenaud and Power will be tied in second place finishes, and Power will win based on third place finishes, so long as Pagenaud doesn’t come in third.
If one of those scenarios doesn’t happen, well, we’ll figure that out when the time comes.

Luckily (unluckily?) we probably won’t have to deal with a tie in the championship. There’s just a 0.56% chance of the championship being tied after Sonoma.

by Drew

A Quick Model Update — 9,000 More Simulations, Please

Our IndyCar Championship model has been updated and will now run 10,000 simulations of the remaining races instead of the previous 1,000.

This will help improve the accuracy of the model, especially early on in the season when more championship possibilities are at play. As it sits right now, there are only two contenders for the championship, so simulating the remaining race 1,000 times works just fine. It’s really more of a benefit for next season than this one.

The graph on the site has been updated after running these 10,000 simulations.

by Drew

Championship Update: That’s Not What Power Wanted

Watkins Glen was pretty much a disaster for Will Power.

He got taken out of the race after coming together with Charlie Kimball on lap 38, resulting in a 20th place finish for the Aussie. Not only did the crash end his race, it may have ended his championship hopes too.

Here’s how the odds look with one race remaining. Pagenaud pretty much has this locked up barring a DNF out in Sonoma along with a top-four finish for Power.

You can read about how our model works here, and the latest forecast can be found here.

by Drew