Fans entered the 2022 season with high expectations for the Next Gen car. Midway through the regular season, it seems that one of the biggest expectations — a broader group of drivers capable of competing for wins — has been fulfilled. That success may be one reason the vitriol aimed at the All-Star Race was so strong. We’ve been spoiled.
Given variables like track preparation, weather and sheer luck, it’s impossible for every race in a season to be great. But 2022 has, so far, brought an awful lot of good racing.
Closer Endings
For example, I’ve arranged the margins of victory for races in the 2022 season from smallest to largest in the graph below. Dover is not shown because it ended under caution.
Of the 12 races this season that ended under green:
- 100% had margins of victory less than 1.5 seconds.
- 83.3% had margins of victory less than one second.
- 58.3% of the races were won by less than a half second.
Only 56.2% of races in 2021 were won by less than 1.5 seconds. Last year’s spring Las Vegas race was won by 3.156 seconds compared to 0.178 seconds this year. Even the much-maligned 2022 Martinsville race had a 0.303-second margin of victory compared to 1.972 seconds in 2021.
Of course, margin of victory doesn’t tell the whole story. A driver could lead the entire race and win by a narrow margin of victory, but few fans would call that an exciting race.
Final passes for the lead
One much-noted statistic about this year’s racing is that the last pass for the lead came in the final 10 laps of the race 10 times in the first 13 races. That’s a record. How meaningful a record is it? Without context, you don’t know if the previous record was nine races in the first 13 or five races.
To test how impressive 2022’s last-pass-for-the-lead record is, I surveyed every race from 1970 to the present. It’s difficult to go back much further because there’s less reliable information about lead changes for earlier races.
I excluded races shortened by bad weather or darkness, then determined how many races had final passes for the lead within the last 10 laps. Because each season included a different number of races, I express results in percentages.
The next graph shows the 10 seasons with the highest percentage of races in which the final pass for the lead happened within the last 10 laps.
Not only does the 2022 season-to-date rank highest, it does so by 21.3% over the next closest season, 2017. The 2017 season beats the third-ranked season by 7.0%, and then we settle in to a series of seasons separated by much smaller differences.
Only eight of the 53 seasons considered had 40% or more races satisfy the criterion. The lowest two seasons in my dataset are 1992 and 1973, both of which had only 11.1% of races with a final pass for the lead within the last 10 laps.
All of the seasons that made the graph, with the exception of 1981, are in the 21st century. 1981 was an exceptional season, with last-lap passes in five of the 31 eligible races. That 16.1% of races with last-lap passes is highest in this dataset.
This season has had two last-lap passes for the win — at Bristol dirt and Talladega. With 23 more races, including three superspeedways, 1981’s record may also fall this year.
A five-decade-plus trend
This year is part of an overall trend toward closer racing. That’s not surprising — drivers don’t win races by multiple laps these days — but it is satisfying when expectations align with the data.
I made a box plot for the percentage of races with final passes for the lead within the last 10 laps, grouping the data by decade. The leftmost box includes races from 1970-1979, the next box from 1980-1989 and so on.
If you’re not familiar with box plots:
- 50% of the data points fall within the shaded area. Since there are 10 years in all but the 2020 decade, that means the five middle years are represented by the boxes.
- The horizontal line shows the median value: half of the data points are above this value and half are below.
- The ‘whiskers’ — the thin vertical lines ending in horizontal lines — represent the expected statistical minima and maxima of the dataset.
- The two dots for the 1990s indicate seasons with percentages so far from the median that they’re considered outliers. The top dot is 1994 and the bottom one is 1992. The two outliers are more indicative of the rest of the races being very similar than they are of anything else.
- The 2020s have only 2.36 seasons of data, but I wanted to include them to show the trend’s direction.
The median values for the percentage of races that have passes for the lead in the last 10 laps has climbed over the last five decades. But even with this trend, 2022’s value (so far) of 76.9% is truly exceptional.
A broader look at end-of-race lead changes
To check whether 2022 had such strong stats only because we chose to look at 10 laps from the end, I ran the same analysis, with different numbers of laps from the end.
2022 leads all other seasons for final passes for the lead within the last one (30.8%), two (38.5%), five (46.2%), 10 and 20 laps. 2022 doesn’t have any races with last lead changes between 10 and 20 laps, so the average is the same for both: 76.9%
Where 2022 doesn’t rank is in passes within the last 50 laps. Why look so far out? Because races where someone leads that many laps at the end tend not to be as exciting as those with lead changes close to the checkered flag.
The graph below shows the 10 seasons with the most final lead changes within the last 50 laps of the race. In other words, these are the seasons with the fewest races in which one car dominated at the end.
The 2010 season wins this competition, with 94.4% of its races having the final lead change within the last 50 laps. 2022 just misses making the cut with its 84.7%. The lower number is because, of the three races without a lead change in the last 10 laps, the last lead changes happened at 23 laps (Phoenix), 52 laps (Dover) and 82 laps (Martinsville).
Going forward, the question is whether the shake-up will continue, or whether one driver (or one owner or manufacturer) will figure something out and dominate the summer.
