Cautions are Down in 2018: Are NASCAR Drivers Getting Better at Avoiding Accidents?

Cautions are down this year — again. And that’s despite adding around 70 cautions per season when stage racing was instituted in 2017.

That same year, NASCAR introduced the damaged vehicle policy: If you can’t get your car back on track and at minimum speed in five minutes on pit road, you’re done for the day. This rule made the penalty for getting involved in accidents a whole lot higher.

So it’s natural to wonder if the reason accidents are down is because drivers are being more careful.

The State of Cautions in 2018

Recall the panic after Martinsville at the start of the year. Cautions were running low at a rate that hadn’t been seen since 1999. 

I’ve been through the reasons why comparing absolute numbers of cautions is wrong before. — especially since there were only 34 races in 1999 and the first six races weren’t always run at the same tracks. It makes a difference if Bristol is one of the first six tracks or not.

What you really want to compare are the number of cautions per 100 miles, which I’m going to call the C-Index. If the C-index is two, it means there are (on average) two cautions per 100 miles. A 500-mile race would have (on average) ten cautions.

Note: I suppressed the zero (meaning that it starts at ‘1’) on the vertical axis to show the variation better.

It doesn’t look any better plotted this way, does it?

But it’s the year as a whole that matters. I’ve shown that trying to predict the ultimate number of cautions (or anything else) based on the first five or ten races is a mistake. You can see how big a mistake by looking at how the C-Index changes over time.

If you’re wondering, Race #8 was Bristol. The caution rate didn’t really level off until the very end of the season. Here’s the season-end C-Index for the last 18 years:

The lesson is not to make predictions after six races. 2018 didn’t end up being the season with the lowest number of cautions. 2012 still holds that distinction. This year was actually pretty similar to 2010. Yes, 2018 was low, but not abnormally so. We’ll get to why that is in a moment.

Five Years of Cautions: In Detail

I did a detailed analysis of cautions for the last five years, which gives us three years before the stage racing/damaged vehicle policy and two years after. Let’s first look at a graph of total cautions for that time period.

What’s interesting here is that NASCAR added 2 stage-end cautions per race when they introduced stage racing. That’s 72 additional cautions, although there have been cases in which accidents ended a stage, so the number varies a little.

The overall number of cautions went up the year stage racing and the damaged vehicle policy were instituted —  but only by 27. (I discussed then whether we could infer that drivers were driving harder due to stage racing.) In 2018, the number of cautions was down 22 from 2016. 

So we added 70-something stage-end cautions, but the total number of cautions went down. Did these policies change the way drivers comport themselves on the track? 

Breaking Down Cautions

In order to answer that question, we have to look at the origin of the cautions. Let’s look at 2018 first.

The largest causes of cautions in 2018 were

• Accidents (45%) – 111
• Stage end cautions (29%) — 71
• Spins (10%)  – 24
• Debris cautions (5%) – 13

Let’s compare that 2014:

There were, indeed, more accidents, but look at how many debris cautions there were!

• Accidents (49%) – 150
• Debris (27%) – 81
• Spins (10%) – 30
• Oil/Fluid (5%) – 17

Let’s look at the change over the five years in a couple categories.

Accidents

Here are the numbers for just those cautions NASCAR attributes to accidents:

There was a small increase from 2016 – 2017 and a pretty significant decrease (57!) this year.

Spins are basically one-car accidents, right? So I thought perhaps I should include spins in the graph.

Accidents for which drivers can be held responsible were definitely down in 2018 after being pretty doggone consistent for the last four years. But if we’re interested in the impacts of stage racing and the damaged vehicle policy, we’re interested in changes from 2016-2017. And there wasn’t a big change. This tells us that the institution of stage racing and the damaged vehicle policy didn’t make drivers any less likely to have accidents.

Why Fewer Accidents in 2018?

Can we explain the smaller number of accidents this year? I think we can. I analyzed which cars were involved in accidents (My analysis didn’t include spins, but I’ll do that in another blog.) Here are the cars that were involved in the most accidents in 2017.

NOTE: Let’s not confuse being involved in an accident and causing an accident. The stats don’t place blame. Sometimes it’s just being in the wrong place at the wrong time.

• Two of the prime offenders (Patrick and Earnhardt, Jr.) aren’t racing in 2018
  • Almirola, in the 10-car, was only involved in 9 accidents in 2018
  • Bowman, in the 88, had 12 accidents  in 2018
• Trevor Bayne only ran 21 races in 2018. The 6-car had 10 accidents in 2018, 9 of which were with Bayne behind the wheel
• Kahne moved to the 95 in 2018 and was only involved in 6 accidents.

Although accidents are down significantly in 2018, I don’t think it has anything to do with stage racing or the damaged vehicle policy. 

What about Debris?

Here are the debris cautions for 2014 – 2018:

This is pretty interesting. The number of debris cautions decreased over 2014-2016 pretty steadily. This is like due to changes in rules for how the car is constructed. 

But there’s a huge drop off in 2017 — right when the damaged vehicle policy was instituted. We see a decrease in cautions due to oil and fluid on the track over time as well, although it’s not as pronounced.

If we combine the two…

Keeping damaged cars from going back on track after an accident and causing more cautions was the primary reason for the damaged vehicle policy.

This graph suggests it’s working.

Counting stage-end caution laps

Many people (including drivers) have complained about counting the stage-end caution laps as part of the overall race lap count. Even doing that, the average percent of races run under caution has gone down. In 2014, 15.7% of the races were run under caution. This year, we’re down to 13.6% cautions.

• In 2014, 15.7% of the laps were run under caution.
• In 2018, only 13.6% of the laps were run cautions.

So even counting the stage-end laps, NASCAR fans are getting more green-flag laps in 2018 than they have in the last five years.