With Bristol coming up this weekend, thoughts turn naturally to carnage. Short track racing means accidents. Given the greater-than-usual level of aggression among drivers this year, we can expect at least a few collisions.
But does short-track mayhem hold a candle to Daytona and Talladega?
Accident vs. Spin
Racing-reference.info (one of the best places on the web to find NASCAR statistical data) posts a chart like the one below for each race listing cautions, reasons, who was involved, how long the caution was and who got the free pass.
Nascar divides cautions into about ten different types: stage end, debris, oil/fluid on track, etc. They distinguish between spins and accidents.
The first yellow flag in the chart above was for the #20 spinning. The second yellow flag was also a single-car incident: Something in the 2-car broke and sent it into the wall. The distinction is:
- A spin produces little, if any damage
- An accident means there was contact.
We’ll only deal with the latter here.
How Many Accidents Per Season?
Here’s a chart showing the total number of accidents for the last five years, plus what we’ve accumulated so far this year.
I wrote elsewhere about why accidents went down in 2018 (spoiler: damaged vehicle policy plays a role). We’re up just a little (75 vs. 65 at this time last year), but we were were already over 100 accidents other years shown by this time in the season.
Are There More Accidents in the Playoffs than the Regular Season?
The stakes rise as the season continues and you might wonder if the numbers increase as the year goes on. I’ve separated accidents for the first 26 races vs. the last 10.
The regular season is 72.2% (26/26) of the entire season. The red line (at about 72%) shows you where the split should be if accidents happened at the same rate during the two periods.
While it varies from season to season, there is no indication that there are more accidents in the playoffs vs. the regular season.
How Many Accidents in a Typical NASCAR Race?
NASCAR doesn’t really have a ‘typical’ raced because the series visits a wide range of track types. But if we combine the data for the last five years (plus 2019 through the first 23 races), we get a very reasonable histogram that is almost Poisson-ish in nature.
- Out of the 203 races, about 5.9% (12 races) had no accidents. This year, we’ve already had three accident-free races.
- The most likely number of accidents in a race is 2 or 3
- The biggest number of accidents over these years is 15, which happened at Darlington in 2015.
- Only 6.4% races (13) had 9 or more accidents. You’re about as likely to have a race with 9 or more accidents as you are to have a race with no accidents
Are All Years Similar?
Here’s the same data, but broken down by year.
- There is a very similar pattern each year.
- We haven’t had more than 10 accidents in a race since 2016.
- We’ve had at least three accident-free races in the last two years
- That big red spike shows that 28% of the races (10/36) in 2018 had two cautions — and none had more than eight.
Which Tracks Have the Most Accidents per Race?
The question I posed at the start was whether the mayhem at Bristol was comparable to the chaos at plate tracks. The table below shows the five races with the most accidents in each of the last five years, plus 2019 to date. I highlighted the plate tracks in blue and the mile-and-a-halfs in yellow.
- Most accidents in a single race:
- Darlington 2015: 15
- Fall Richmond 2016: 14
- Spring Bristol 2016: 13
- Fall Martinsville 2015: 12
- The plate track with the largest number of accidents is This year’s Daytona 500, with 8.
Surprisingly, plate tracks don’t make much of a showing in this chart. There are actually more mile-and-a-halfs than plate tracks.
A pie chart makes the variations in track type even clearer. This chart shows how many times each track appeared on the list of top-five tracks for accidents each year.
- Plate tracks (green) made up 13.3% of the 30 tracks
- Large tracks (orange): 6.7%
- 1.5-milers (blue): 20.0%
- Smaller tracks (shades of red): 60%
To be fair, we haven’t run Talladega this year, but we also haven’t run fall races at Bristol or Martinsville, both of which typically have larger numbers.
Does Race Length Matter?
Short tracks have more accidents than longer tracks, even though they often run shorter distances. Bristol is 266.5 miles while races at Daytona and Talladega are 400 or 500 miles. Charlotte has a 600-miler.
So let’s look at how many accidents there are per 100 miles, which evens out the length disparity. Here’s the analogous chart.
Now there’s only one plate track represented: This year’s Daytona 500 had 1.55 accidents/100 miles. Spring Bristol had 1.7 times more.
The highest rate is Spring Bristol in 2016 with almost five per hundred miles. If Spring Bristol that year had been 500 miles, there would have been 24 or 25 accidents.
This pie chart shows how many times each track is represented in the table.
We’ve got a different balance of responsibility using this metric.
- Plate tracks (green): 3.3%
- Large tracks (orange): 6.7%
- 1.5-milers (blue): 16.7%
- Rovals (red): 3.3%
- Smaller tracks (shades of red): 70%
So Are Short Tracks More Dangerous?
There’s one more variable we should consider. The 2019 Daytona 500 only had 8 accidents. But here’s a chart showing how many each car was involved in.
While there were only eight accidents, almost everyone was involved in at least one (and often more than one) of them. It’s not too surprising that the first and second-place cars were two of the four vehicles that avoided all the crashes.
Even with fewer accidents at plate tracks, you have a higher likelihood of being caught up in each one.
The chart below tallies the number of accidents and the total number of cars involved. I only have three years worth of data available here.
You can see the tradeoff:
- In 2019, the Daytona 500 only had eight accidents, but they involved 51 cars.
- The Charlotte race (spring) had three more accidents than the Daytona 500, but only involved 19 cars.
- For each of the last three years, a plate track took the prize for the most numbers of cars involved with 51, 51 and 46
- So far this year, 190 cars have been involved in wrecks.
- 95 of those cars (50.0%) were at the three plate races.
- 79 (41.5%) were at Daytona.
- 45 (23.7%) were at Bristol
But, of course, we still have to run Talladega this fall!
Which Tracks Eat Cars?
Here’s a pie chart showing how the 313 cars involved in accidents breaks down by track in 2018.
- Plate tracks (green) : 35.5% of all cars in accidents
- Large tracks (orange): 10.0%
- 1.5-milers (blue): 21.7%
- Rovals (red): 15.0%
- Smaller tracks (shades of red): 17.8%
But note that Bristol hasn’t disappeared, even in this category.
- Bristol had more cars involved in 2018 than the rest of the short tracks combined
- Bristol even outdid Talladega.
NASCAR’s Most Accident-Prone Tracks Are…
To really determine which tracks are the most accident prone, you have to look at how many accidents they have per 100 miles and how many cars are typically involved in an accident.
As a first attempt at a figure of merit, let’s just multiply those two numbers together. The tracks that consistently end up with the largest numbers (2017 through 2019-to-date) are Daytona, Bristol and Talladega.
|Track||DLP Danger Index|
I’ve starred the Charlotte Roval because we only have one data point. You’ll notice that road courses in general rank very low in terms of propensity toward accidents. We’ll have to see what happens this year.
The list is imperfect because of the low numbers of data points. And remember that there’s variation from year to year.
- Daytona’s DLP Danger Index ranges from a low of 6.18 to a high of 12.1.
- Bristol’s DLP Danger Index ranges from 4.12 to 9.38
- Sonoma’s Index is low because it went two years without accidents, but the one year there were accidents, its Danger Index was 2.74
Although plate tracks seem to have the market on accidents, Bristol holds its own.