The 2024 average finishes data show that a number of drivers have much better records than their final positions in the standings suggest. Chase Elliott and Christopher Bell, in particular, had the best average finishes of the year in total races and non-DNF-races respectively.
But let’s look at the entire field to get a complete picture.
2024 Average Finishes by Driver
The chart below summarizes average finishes for drivers who ran 30 or more races.
Races Run | Avg Finish | SD | Best Finish | Median | Worst Finish | |
---|---|---|---|---|---|---|
Elliott | 36 | 11.72 | 8.50 | 1 | 9.50 | 36 |
Bell | 36 | 12.81 | 13.00 | 1 | 6.00 | 38 |
Larson | 35 | 12.91 | 12.23 | 1 | 10.00 | 39 |
Byron | 36 | 12.94 | 11.73 | 1 | 7.50 | 38 |
Reddick | 36 | 13.19 | 11.35 | 1 | 9.00 | 35 |
Hamlin | 36 | 13.89 | 11.63 | 1 | 10.50 | 38 |
Buescher | 36 | 14.31 | 9.23 | 1 | 14.00 | 37 |
Bowman | 36 | 14.75 | 10.60 | 1 | 13.50 | 38 |
Chastain | 36 | 14.94 | 10.16 | 1 | 12.00 | 40 |
Blaney | 36 | 15.22 | 13.19 | 1 | 10.50 | 39 |
Wallace | 36 | 15.28 | 9.64 | 3 | 14.00 | 36 |
Keselowski | 36 | 15.61 | 10.72 | 1 | 14.50 | 36 |
Logano | 36 | 17.11 | 11.51 | 1 | 15.50 | 34 |
Truex Jr | 36 | 17.39 | 10.43 | 2 | 16.00 | 37 |
Gibbs | 36 | 17.42 | 11.15 | 2 | 16.50 | 40 |
Suarez | 36 | 17.86 | 9.49 | 1 | 18.00 | 40 |
Hocevar | 36 | 18.31 | 7.81 | 3 | 17.00 | 40 |
Busch | 36 | 18.31 | 10.31 | 2 | 19.00 | 35 |
Briscoe | 36 | 18.81 | 10.06 | 1 | 18.50 | 38 |
Gragson | 36 | 19.56 | 10.48 | 3 | 18.00 | 38 |
Cindric | 36 | 19.92 | 9.72 | 1 | 20.00 | 37 |
Gilliland | 36 | 20.86 | 8.22 | 6 | 20.00 | 36 |
Preece | 36 | 20.89 | 9.36 | 4 | 22.50 | 39 |
McDowell | 36 | 21.31 | 10.42 | 2 | 21.50 | 38 |
Stenhouse Jr | 36 | 22.06 | 10.37 | 1 | 23.00 | 37 |
A Dillon | 36 | 22.56 | 9.37 | 1 | 24.00 | 37 |
Berry | 36 | 22.72 | 9.92 | 3 | 24.50 | 38 |
Jones | 34 | 22.91 | 8.16 | 5 | 23.00 | 35 |
Hemric | 36 | 22.97 | 8.16 | 9 | 23.50 | 38 |
Smith | 36 | 23.19 | 9.52 | 2 | 23.00 | 39 |
Haley | 36 | 23.42 | 8.02 | 7 | 24.00 | 39 |
LaJoie | 36 | 23.72 | 9.52 | 4 | 22.50 | 37 |
Nemechek | 36 | 25.42 | 8.35 | 6 | 28.50 | 36 |
Burton | 36 | 25.72 | 8.78 | 1 | 27.50 | 39 |
The ‘SD’ column is standard deviation, which tells you how much a driver’s finishes during the season varied relative to the average. Sixty-eight percent of a driver’s finishes are within one standard deviation of their average finish.
You can think of the standard deviation as a measure of consistency. For example, Elliott has not only the best overall finish, he also has one of the narrowest standard deviations.
A low standard deviation is good โ if the average finish is also low. Drivers don’t brad about being consistently 30th.
I also indicated each driver’s best and worst finishes on the table. While every driver has bad finishes (usually due to DNFs) Joey Logano never finished worse than 34th. That’s the best worst finish of any driver โ and especially impressive given that Logano had 6 DNFs this year.
Finally, the median finish is the value where half of the driver’s finishes are above that number and half are below. Elliott’s median of 9.5 tells you that half his finishes were ninth or better and half were ninth or worse.
Best 2024 Average Finishes
The graph below shows the average finishes. For clarity, I’ve included only the best 15 drivers over all 36 races.
Average Finishes Not Including DNFs
DNFs are sometimes a driver’s fault, sometimes due to mechanical failure and sometimes due to circumstances entirely outside of the driver and control. So let’s look at the same numbers, but without the DNFs.
Races Run | Avg Finish | SD | Best Finish | Median | Worst Finish | |
---|---|---|---|---|---|---|
Bell | 30 | 8.23 | 8.59 | 1 | 5.50 | 35 |
Larson | 30 | 9.20 | 8.63 | 1 | 6.00 | 34 |
Blaney | 29 | 10.34 | 9.33 | 1 | 7.00 | 33 |
Elliott | 34 | 10.38 | 6.58 | 1 | 9.00 | 29 |
Byron | 32 | 10.56 | 10.03 | 1 | 6.50 | 35 |
Reddick | 32 | 10.91 | 9.77 | 1 | 6.50 | 32 |
Hamlin | 32 | 11.09 | 8.91 | 1 | 8.50 | 30 |
Chastain | 32 | 12.41 | 7.45 | 1 | 11.50 | 33 |
Bowman | 33 | 12.94 | 9.04 | 1 | 9.00 | 38 |
Buescher | 34 | 13.00 | 7.66 | 1 | 12.50 | 30 |
Wallace | 33 | 13.58 | 8.09 | 3 | 13.00 | 35 |
Keselowski | 33 | 13.94 | 9.54 | 1 | 13.00 | 33 |
Logano | 30 | 13.97 | 9.92 | 1 | 12.50 | 34 |
Gibbs | 32 | 15.25 | 9.68 | 2 | 15.50 | 36 |
Truex Jr | 33 | 15.70 | 9.13 | 2 | 15.00 | 34 |
Busch | 31 | 15.97 | 9.04 | 2 | 15.00 | 31 |
Gragson | 30 | 16.33 | 8.20 | 3 | 14.00 | 34 |
Briscoe | 31 | 16.35 | 8.50 | 1 | 15.00 | 32 |
Suarez | 34 | 16.74 | 8.46 | 1 | 17.00 | 31 |
Hocevar | 34 | 17.24 | 6.53 | 3 | 17.00 | 32 |
Cindric | 32 | 18.06 | 8.61 | 1 | 18.50 | 34 |
McDowell | 29 | 18.10 | 8.89 | 2 | 16.00 | 36 |
Preece | 31 | 18.87 | 8.25 | 4 | 18.00 | 37 |
Berry | 26 | 18.96 | 8.83 | 3 | 20.00 | 36 |
Stenhouse Jr | 29 | 19.28 | 9.58 | 1 | 21.00 | 33 |
Gilliland | 33 | 19.58 | 7.30 | 6 | 18.00 | 32 |
A Dillon | 32 | 21.06 | 8.83 | 1 | 22.50 | 37 |
Smith | 31 | 21.32 | 8.83 | 2 | 21.00 | 36 |
Jones | 31 | 21.84 | 7.72 | 5 | 22.00 | 35 |
LaJoie | 31 | 21.97 | 9.07 | 4 | 21.00 | 36 |
Hemric | 33 | 22.24 | 7.98 | 9 | 23.00 | 37 |
Haley | 34 | 22.88 | 7.93 | 7 | 23.50 | 39 |
Burton | 29 | 23.83 | 8.61 | 1 | 25.00 | 36 |
Nemechek | 30 | 24.10 | 8.47 | 6 | 26.00 | 35 |
Bell leads the 2024 average finishes list for races completed with an average 8.23 finish position and a median finish of 5.5. Kyle Larson and Ryan Blaney also move ahead of Elliott with average finishes of 9.20 and 10.34 respectively when DNF races are excluded.
Best Average Finishes for Completed Races Only
Comparing the graph below to the one above shows how DNFs impact driver productivity. Only two drivers had average finishes better than 10.0 in races completed: Bell and Larson.
Bonus: Who Got Better in the Playoffs
Comparing driver averages in the regular season to those in the playoffs tells you who got hot at the right time.
- Bell had a 15.0 average finish in the regular season and 7.1 after. (7.9 position improvement.)
- William Byron’s 14.6 average finish improved to 8.7 in the playoffs. (5.9 position improvement.)
- Logano had a 18.3 average finish in the regular season and a 13.9 after (4.4 position improvement.)
- Denny Hamlin’s average finish of 15.0 before the playoffs rose to 11.0 after. (4.0 position improvement.)
And then, on the other side of the list:
- Ty Gibbs started the season strong with 14.7 average finish; however, he dropped to a 24.5 average finish in the playoffs.
- Tyler Reddick won the regular season championship with his 11.2 average finish. Unfortunately, that average dropped to a 18.5 in the playoffs.
- Brad Keselowski fell from a 14.1 to a 19.5 average finish.
And that’s the 2024 average finishes post. Remember, if you have questions you’d like answered, drop me a line at ask (atsign) buildingspeed (dot) org.
For the rest of the posts in my by-the-numbers series, see the 2024 redditsummary page.
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.
Discover more from Building Speed
Subscribe to get the latest posts sent to your email.
I’ve felt for a long time that the various methods used to determine a Cup Champion (or other series) today and over the past could stand some changes. The current system up to the point where the “playoff” format goes into effect is OK, but I really believe something quite different should go into effect for the last 10 races.
Further – and I’ll get back to the current scenario(s) and formats – there are 2 things long-term that NASCAR is going to have to confront.
First – the NASCAR pro-series (Cup, Xfinity, Craftsman Truck) season is too long, with too many races and competition from other US sports, especially as the season draws to conclusion. 30 points races should be a target for the entire season going forward.
Second – the “playoff” races, as a percentage of the full season currently, are too heavily skewed toward focus on the final 10 races. This has many downsides, but especially that races too often come down to just those cars/teams that that are in the Playoffs….coverage of the remainder of the field drops significantly, following the 16>12>8 elimination format.
With a reduced number of races (in the future), my belief is that a “playoff” system should play less of a role in the full season. My proposal is to reduce the playoff content of the season to seven (7) races out of a 30 race full season – just over 23% of the races. Currently with 36 races, 10 playoff events, the portion in playoffs is nearly 28%
All teams that advance to playoff status should come under a ranking system that adresses the Average Finish issue – not points. That ended at Race 26. No more win & you’re in, once advancing to Playoffs. Lower golf score = higher race finish, so the final 10 (or 7 under my future season proposal) brings the Championship competitors to the final race based on 10 (or 7) races, 6 of which would be used to determine the eligible Race 7 competitors.
Race 7 – beat all the other 3 eligible contenders, just as is the case today.
It would take some number crunching to see if this really changes anything….probably not much, but it would be nice to find a surprise, at least to the point where the 9 or 6 races deliver the last 4 standing to be eligible for the Championship.
One thing I’d like to see is a possible Wild Card system – it would be activated when those that made the cut are outperformed based on Stage resuots. Spice things up….why not!