Racermetrics race-database.com


2023 Stock Car Model Update

by Sean Wrona


With the Daytona 500 starting in a few minutes, I wanted to finish up the update of my stock car teammate model just as I did for my open wheel model a week ago. This update includes all races through the end of 2022 but does not include the Daytona season-openers for either the reborn Craftsman Truck Series or the Xfinity Series. Just as with my open wheel model, I am only listing current drivers here since the ratings for any of the inactive drivers will obviously not shift very much in future updates and this keeps the article simpler and shorter. Additionally, so many of the 20th century era drivers on my model (especially those who peaked in the '70s, '80s, and '90s) had wildly inaccurate ratings while the drivers of the 21st century who had lots of teammate comparisons generally did not (with some obvious exceptions like Joey Logano, Ryan Blaney, and Tyler Reddick who remain wildly underrated here, but I do think most other drivers are in relatively reasonable positions overall.)

As before, I compare each driver's ratings in the original 2021 and 2022 (pre-chase) updates of my stock car model. I did make a subsequent update for 2022 at the end of the year (which is obviously very close to the current one), but I did decide to add the three drivers who are expected to make Cup starts in 2023 who were not already integrated into the model. Two of them (Travis Pastrana and Riley Herbst) will be making their debuts today, while the third one (Chandler Smith) failed to qualify but it seems inevitable that he will make his first start before the next time I updated my model, so I threw him in as well.

I also made projections based on my model for the percentage of the time each driver will beat their teammates. For the drivers who are expected to compete in multiple NASCAR divisions, I only included the teammate comparisons for drivers in the series in which they declared for points, so I will only be comparing Riley Herbst to Cole Custer here, not to Cody Ware (but since Ware has declared for Cup points, I will list Ware's comparison to Herbst there so you can still see these projections listed for the teammates that have declared for Cup points.) I only included drivers who are expected to make NASCAR starts in 2023. I threw in Kurt Busch even though it's unclear whether he'll ever make a start again (I hope so) and Kyle Weatherman since I personally think he is going to make some Xfinity starts this year at some point even though he has nothing announced. I suppose I could have used the same logic to include Trevor Bayne as well, but I decided not to. I do think Weatherman is more likely to make Xfinity starts this year than Bayne.

With Jimmie Johnson and Kevin Harvick both in decline, Johnson actually overtook Harvick in this iteration of the model for the first time because he was inactive while racing in IndyCar as Harvick was declining in Cup. Admittedly, Harvick has remained very strong in terms of delivering results and outfinishing his teammates while most of his decline has been in his ability to fight for the lead in races. Chase Briscoe utterly dominated him in terms of leading races and fighting for the win but I still think Harvick had a better season than Briscoe because he was so much more consistent. However, I do kind of expect Briscoe to be regularly faster than Harvick this year and I can see him beating Harvick in terms of head-to-head finishing records in his retirement season, which will lead to further declines for Harvick. It'll be interesting to see whether Johnson or Harvick ends up on top. Johnson I believe has declined significantly more than Harvick has and I think Erik Jones is better than any of the other Stewart-Haas drivers at this moment, so I fully expect Jones to dominate Johnson this year and be probably the biggest gainer in my model in 2024 as he is benefiting from competing against a rookie and a washed-up legend: I could even conceivably see Jones leading my model in much the same vein that Chris Buescher did in 2021. However, Johnson has fewer teammates than Harvick and he is only running a part-time schedule while Harvick is running every race so my guess is that Johnson will decline to a lesser degree than Harvick will even though he is less washed up, but it will be interesting to see who finishes their career on top.

Chase Elliott and Kyle Larson continue to keep switching positions back and forth while competing to see which Hendrick teammate is currently the best driver in the sport, a debate that at this point is far from resolved. Since the only difference between the 2022 and 2023 models is performance in the playoff races, few drivers made large movements but there were some exceptions. The drivers who made significant movements included Ty Gibbs (+.072), Andrew Ranger (+.039), Chase Briscoe (+.037), Bayley Currey (+.033), Todd Gilliland (+.027), Cody Ware (+.027), Corey LaJoie (+.026), Kyle Weatherman (+.022), Anthony Alfredo (+.017), William Byron (+.017), Noah Gragson (+.015), Michael McDowell (+.014), Brett Moffitt (+.014), Kaz Grala (+.013), Ben Rhodes (+.013), Ryan Truex (+.013), A.J. Allmendinger (+.011), and Christopher Bell (+.011). Most of these drivers did have significant highlights relative to their usual performance in 2022, particularly in the playoffs. Obviously the more inexperienced drivers such as those who recently or currently competed in Xfinity and had relatively few teammate comparisons make the most staggering jumps, as we see for instance from Gibbs. Some of these jumps such as McDowell's and Moffitt's were primarily effects of the new drivers who were added to the model.

Of the three new drivers, Chandler Smith may have been the only driver to miss the Daytona 500, but he's still overwhelmingly the best of them. For a recent truck graduate and Xfinity rookie to be rated as high as -.062 is extremely impressive when you consider that a lot of drivers with a lot more experience than he has are lower and sometimes significantly so. Smith is ahead of Gibbs by a large margin still (although I acknowledge that gap will likely significantly narrow this season) and close to interchangeable with Noah Gragson, so that leads me to believe that Smith might be a leading championship contender in Xfinity this season, but he is admittedly driving for Kaulig, which probably will continue to be not as fast as JRM or Gibbs as they were last year. Since Allmendinger failed to make the final four in Xfinity last year despite in my opinion delivering the best performance in the series and Smith is rated significantly lower than him, it might be difficult for him to put big numbers up, but I do think he will deliver impressive performance. However, the two drivers making their first Cup starts in the Daytona 500, Pastrana and Herbst, start as the 4th and 3rd lowest rated active drivers in the model respectively. They are not up to usual Cup standards. Even though a lot of people gave Herbst considerable hype in Xfinity last year, I'm not really sure why when you consider that Cole Custer had a seven-win season and Chase Briscoe had a nine-win season for the team in previous years, and both of those drivers are below average in my model as well. However, for all the criticism Custer gets, he has still slightly improved through every iteration of my model and he indeed overtook his much older replacement Ryan Preece in this iteration of my model. A lot of people are now expecting big things from Preece especially after he took the lead on track in the Clash at the Coliseum, but temper your enthusiasm about him. He'll probably run about the same and although I don't think either of them will ever be great, Custer does seem to have marginally more potential for growth. I decided to count all the Rick Ware and B.J. McLeod cars with Stewart-Haas backing in the Xfinity Series as Stewart-Haas entries when I was calculating Herbst's teammate records, and this may be the reason Custer overtook Preece because he had a 1-0 record against Herbst while Herbst had a 2-1 record against Preece. I realize Preece did beat Custer yesterday, so it's possible that might have come close to reversing. However, given Custer's history in the series relative to Herbst's, I fully expect him to utterly dominate Herbst to rebuild his cred and honestly I do probably expect him to overperform his 81.0% expectation even though he lost to him yesterday.

However, the driver I expect to gain the most is obviously Jones given the fact that he has a rookie teammate and a washed-up legendary teammate. I also expect Tyler Reddick to make up a lot of ground. One of the rough things about my model is that when both teammates are underrated or on the rise (as with Joey Logano and Ryan Blaney or Ross Chastain and Daniel Suárez or Reddick and Bubba Wallace), that one driver will have to improve at the other driver's expense. Wallace is clearly a driver on the rise but Reddick is clearly a driver on a meteoric rise so I think he will far outperform his expectation of beating Bubba only 52.8% of the time. That's no knock on Bubba and more a reflection of the fact that I think Reddick is currently a top five driver in the series even though my model does not reflect that yet. If Reddick does improve, I think one of the biggest benefactors might be Ryan Blaney as Blaney beat Reddick 15-5 when they were in the NASCAR minor leagues together and they are probably both the most underrated drivers in the model at the moment. Even though he's probably somewhat overrated in the model, I also think Corey LaJoie will continue to make moves as I think he will dominate Ty Dillon by even more than the 75.8% of the time my model predicts. Once again, drivers who beat the projections listed here will generally move up while drivers who fail to do so will move down, but that can be complicated by second-order effects. Younger drivers will have more room to improve simply because they have fewer teammate comparisons to begin with and also because a lot of the drivers they were teammates with in the minor league series will later be integrated into the model. I imagine if Brandon Jones ever makes a Cup start, a lot of the other JGR drivers like Ty Gibbs and Harrison Burton will probably move up a lot, and that will indirectly help a lot of other drivers. Since the drivers who do make it to Cup are generally better, the drivers who make slower ascents to Cup (like Herbst for instance) will continue to help elevate the ratings of the younger drivers who made it to Cup faster when they are added to the model.

I'm sure nobody will read this at the time since the Daytona 500 is starting in minutes, but I wanted to get this out there in the nick of time regardless. I wanted to finish this faster but I was doing a completely separate analysis where I was calcuating all the speed percentiles for every IndyCar race back to 2013 (the first year that IndyCar has lap times for every race available for every driver) and I will probably share those results in a month or two, but I see that as being nowhere near as much of a rush as this was, because I did want to get it out before the start of the Daytona 500 and I just barely made it. Whew.

The Model

Driver202120222023Projection
Jimmie Johnson0.2460.2490.25080.1% vs. Gragson, 78.4% vs. Jones
Kevin Harvick0.2480.2510.24984.9% vs. Preece, 79.8% vs. Briscoe, 76.8% vs. Almirola
Chase Elliott0.2200.2240.22364.1% vs. Bowman, 63.2% vs. Byron, 50.5% vs. Larson
Kyle Larson0.2180.2260.21863.6% vs. Bowman, 62.7% vs. Byron, 49.5% vs. Elliott
Kyle Busch0.2020.2000.20171.5% vs. A. Dillon
Kurt Busch0.1720.1730.173
Denny Hamlin0.1650.1630.16679.5% vs. Gibbs, 59.0% vs. Bell, 51.8% vs. Truex
Martin Truex, Jr.0.1510.1530.14877.7% vs. Gibbs, 57.2% vs. Bell, 48.2% vs. Hamlin
Brad Keselowski0.1180.1160.11753.9% vs. Buescher
William Byron0.0730.0740.09150.8% vs. Bowman, 37.3% vs. Larson, 36.8% vs. Elliott
Ross Chastain0.0780.0880.09068.4% vs. Suarez
Joey Logano0.0890.0840.08783.2% vs. H. Burton, 74.4% vs. Cindric, 57.3% vs. Blaney
Corey LaJoie0.0420.0580.08475.8% vs. T. Dillon
Alex Bowman0.0690.0790.08249.2% vs. Byron, 36.4% vs. Larson, 35.9% vs. Elliott
Chris Buescher0.0640.0770.07846.1% vs. Keselowski
Christopher Bell0.0000.0640.07570.5% vs. Gibbs, 42.8% vs. Truex, 41.0% vs. Hamlin
A.J. Allmendinger0.0140.0380.04965.9% vs. Haley, 61.0% vs. C. Smith
John Hunter Nemechek0.0550.0440.04768.2% vs. R. Truex
Ryan Sieg0.0390.0440.046
Brennan Poole0.0160.0240.01663.4% vs. Currey
Ryan Blaney0.0110.0100.01476.0% vs. H. Burton, 67.1% vs. Cindric, 42.7% vs. Logano
Ricky Stenhouse, Jr.0.0080.0130.010
Parker Kligerman-0.0030.0000.001
Matt Crafton-0.007-0.004-0.00255.3% vs. Rhodes
Austin Dillon-0.012-0.012-0.01328.5% vs. Ky. Busch
Aric Almirola-0.025-0.020-0.01958.1% vs. Preece, 53.1% vs. Briscoe, 23.2% vs. Harvick
Michael McDowell-0.047-0.043-0.02965.9% vs. Gilliland, 58.9% vs. Z. Smith
Justin Allgaier-0.026-0.027-0.03062.7% vs. Berry
Erik Jones-0.041-0.036-0.03451.7% vs. Gragson, 21.6% vs. Johnson
Chase Briscoe-0.150-0.087-0.05055.1% vs. Preece, 46.9% vs. Almirola, 20.2% vs. Harvick
Tyler Reddick-0.057-0.055-0.05084.3% vs. Pastrana, 52.8% vs. Wallace
Noah Gragson-0.066-0.05148.3% vs. Jones, 19.9% vs. Johnson
Ben Rhodes-0.074-0.068-0.05544.7% vs. Crafton
Chandler Smith-0.06254.8% vs. Haley, 53.5% vs. Hemric, 45.1% vs. A. Dillon, 39.0% vs. Allmendinger, 23.7% vs. Ky. Busch
Bubba Wallace-0.100-0.083-0.07881.6% vs. Pastrana, 47.2% vs. Reddick
J.J. Yeley-0.091-0.090-0.09059.5% vs. Ware
Cole Custer-0.112-0.099-0.09281.0% vs. Herbst
Daniel Suarez-0.097-0.095-0.09331.6% vs. Chastain
Josh Williams-0.095-0.09682.5% vs. Bilicki
Daniel Hemric-0.082-0.096-0.09751.3% vs. Haley, 46.5% vs. C. Smith, 41.6% vs. A. Dillon, 35.4% vs. Allmendinger, 20.1% vs. Ky. Busch
Ryan Preece-0.117-0.093-0.10144.9% vs. Briscoe, 41.9% vs. Almirola, 15.1% vs. Harvick
Justin Haley-0.122-0.110-0.11045.2% vs. C. Smith, 34.1% vs. Allmendinger
Zane Smith-0.125-0.117
Bayley Currey-0.145-0.151-0.11836.6% vs. Poole
Gray Gaulding-0.133-0.126-0.120
Brett Moffitt-0.121-0.135-0.121
Ty Gibbs-0.202-0.13029.5% vs. Bell, 22.3% vs. Truex, 20.5% vs. Hamlin
Matt DiBenedetto-0.139-0.136-0.134
Ryan Truex-0.159-0.149-0.13631.8% vs. Nemechek
Andrew Ranger-0.102-0.177-0.138
Austin Cindric-0.330-0.155-0.15758.9% vs. H. Burton, 32.9% vs. Blaney, 25.6% vs. Logano
Josh Berry-0.251-0.139-0.15737.3% vs. Allgaier
Kyle Weatherman-0.183-0.186-0.164
Ty Dillon-0.171-0.172-0.17324.2% vs. LaJoie
Cody Ware-0.255-0.212-0.18571.8% vs. Herbst, 50.3% vs. Gilliland, 40.5% vs. Yeley
Todd Gilliland-0.215-0.18849.7% vs. Ware, 42.9% vs. Z. Smith, 34.1% vs. McDowell
Jason White-0.188-0.187-0.188
Timmy Hill-0.187-0.190-0.207
Kaz Grala-0.261-0.247-0.23431.6% vs. Reddick
Anthony Alfredo-0.310-0.257-0.24054.3% vs. Smithley
Harrison Burton-0.189-0.247-0.24641.1% vs. Cindric, 24.0% vs. Blaney, 16.8% vs. Logano
David Starr-0.245-0.245-0.248
Jeb Burton-0.275-0.272-0.271
Garrett Smithley-0.294-0.287-0.28345.7% vs. Alfredo
B.J. McLeod-0.305-0.303-0.305
Joey Gase-0.344-0.338-0.332
Jeffrey Earnhardt-0.348-0.352-0.35453.4% vs. Ellis
Ryan Ellis-0.496-0.389-0.38846.6% vs. Earnhardt
Travis Pastrana-0.39318.4% vs. Wallace, 15.7% vs. Reddick
Riley Herbst-0.40319.0% vs. Custer
Josh Bilicki-0.446-0.430-0.421
Spencer Boyd-0.809-0.726-0.723
Sean Wrona is the Managing Editor of racermetrics.com, the Webmaster of race-database.com, and the winner of the 2010 Ultimate Typing Championship at the SXSW Interactive Conference in Austin. He earned a master's in applied statistics from Cornell University in 2008 and previously digitized several seasons of NBA box scores on basketball-reference.com. He released his first book, Nerds Per Minute: A History of Competitive Typing, in 2021. You may contact him at sean.wrona@gmail.com.