NASCAR simulations heating up as we head to Talladega:

In February, I previewed the NASCAR season and encouraged readers to start digging into NASCAR wagering by considering my simulations for each race as a foundation for their strategies. I project a full-field finish every week, and just as it was last year, the early results have been outstanding, particularly over the last month. Whether it has been on projecting race winners, top 3s, top 5s, or head-to-head driver matchups, the simulations are hitting on all cylinders as we head to Talladega on Sunday for the Jack Links 500. In fact, over the last month, simulations have pegged 14 of the last 20 top-5 finishers for a massive profit. You can find my simulations on our NASCAR Hub. 

 

My simulations typically improve as I collect more and more data on each driver, track, or season. So far, the results of 2025 reflect that, as after five fairly volatile races to start the season, things have settled in nicely. I will share more on that thought specifically shortly, but to summarize, things are going well, and I hope bettors have been reaping the rewards. In this piece, I’ll share some of the results from the first nine races, or quarter season, including what has worked best and what hasn’t. This should be a great time to review and hopefully convince you of the possibilities of wagering on NASCAR, as the summer months typically bring some of the more predictable races on the annual slate. 

That said, we are heading to Talladega this weekend, and to be frank, it is always one of the more unpredictable tracks on the circuit using statistical methods. Like Daytona, which will be highlighted in the result wrap-up below, big wrecks can ruin the days of even the best drivers/cars running. However, underdog wagering can prove to be a very fruitful strategy at the drafting tracks like Daytona, Talladega, and now, even Atlanta.

Here is the average margin between my projection and the actual finish for all drivers in all nine races so far. 

1. OVERALL SIMULATION PROJECTION vs. ACTUAL DRIVER RATING: 5.94 positions (usually around 6.5)

2. OVERALL FINAL SIMULATION PROJECTION vs. ACTUAL FINISH POSITION: 8.54 positions (usually around 9.0)

3. OVERALL INITIAL SIMULATION PROJECTION vs. ACTUAL FINISH POSITION: 9.06 positions (usually around 9.5) 

You can see that my FINAL SIMULATIONS (typically released on Saturday) have been slightly more effective than the initial simulations I put out on Tuesday, meaning I would recommend waiting until the final simulations are released if you’re betting races using them as a foundation. 

Keep in mind, too, that these numbers include wild Daytona and Atlanta races that both produced variances over 11.5 on finishes, as usual. Which leads me to an interesting thought, that being finding out which races have been most predictable so far by simulation, as bettors can use this information to decide which races to partake in most, or perhaps even not at all, the rest of the way. With that in mind, here is the order of race finish variations against the final simulation thus far: 

1. Bristol (9th race) – 5.8 position variance
2. Homestead (6th race) – 6.2
3. Martinsville (7th race) – 7.4
4. Darlington (8th race) – 7.9
5. COTA (3rd race) – 8.6
6. Phoenix (4th race) – 9.3
7. Las Vegas (5th race) – 9.5
8. Atlanta (2nd race) – 11.5
9. Daytona (1st race) – 12.1 

This is precisely why I supply the handicap-ability grades for each track, and the reason why the drafting tracks like Talladega and Daytona are the lowest-graded venues. 

The other key point from the list above is that the four most recent races have been the most predictable. I credit this to 1) more historically predictable venues 2) more data available on the strength of the 2025 drivers/teams.

Here are some of the other key stats from my final simulations so far in 2025. As you analyze these numbers, consider that NASCAR races have at least 36 different drivers in each race. 

· FINAL simulations have accurately pegged one outright winner, two others that were projected second, one more that was projected third, and two others that were projected fifth. So, if you were to have placed “to win” wagers on all of the top two projected drivers, you would currently be at +9.35 units. Had you picked all the top three projected drivers, you would be at +8.85 units. And had you backed all top five projected drivers, you would now be at +5.85 units. This is calculated using -1 unit for all non-winners plus Kyle Larson (+285) at Bristol and (+350) at Homestead, William Byron (+1800) at Dayotna, Denny Hanlin (+550) at Martinsville and (+750) at Darlington and Christopher Bell (+750) at Phoenix.

· INITIAL simulations have accurately pegged the same eventual winning drivers in the Nos. 2-5 spots plus an additional one with Christopher Bell (+1600) at COTA. Add that 16-unit win to the +5.85 units described just above.

· There have been nine races so far this season. My top projected driver on the final simulations had the No. 1 or 2 actual driver rating in that race seven times! In other words, my projection of them being the top driver for the day was spot on, but on some occasions, these guys were afflicted by some bad breaks during the race that affected their finish position (i.e., William Byron at Darlington, who finished runner-up after thoroughly dominating the event).

· Only three of the nine race winners thus far were projected outside the top 5 in my simulations — Christopher Bell (13th at Atlanta, 12th at COTA) and Josh Berry (27th at Las Vegas). However, if you recall reading my Atlanta predictions in Bill Adee’s morning newsletter, Berry was one of my top long-shot picks to win at 75-1. I picked after Berry’s huge Saturday of qualifying and practice vaulted him above normal simulation levels for that race.

· Of the actual 27 top-3 finishers this season, my final simulations pegged eight of them accurately, good for 29.6%, a slight drop of 1.0% compared to last year. This has produced a profit of +3.4 units and an ROI of 12.6%.

· Of the actual 45 top-5 finishers this season, my final simulations pegged 22 of them correctly, good for 48.9%, a noteworthy 7.2%+ jump over a year ago. This has also produced a profit of close to 4 units on the season. Perhaps the greatest of any highlights for this season for the simulation is that 14 of these 22 have come in the last four races, 70% of the actuals!

· On top-10 finishes, the projections on the final simulations versus the actual finishes are over 50%, at 46 out of 90. This is almost identical to a year ago in terms of success. However, the profit for this year is up to about 5 units, thanks in large part to scoring four out of 10 at Daytona, with enhanced underdog odds.

Simulation Projection Results by Driver

Naturally, different drivers will have different predictability levels based on their overall consistency at certain tracks. In looking at the full-time drivers that my simulations have projected most closely in this season’s first nine races, when considering variance between projection and actual finish, they are:

1. Ryan Blaney: 2.67 (average margin)
2. Kyle Busch: 4.86
3. Chase Briscoe: 5.13
4. Chase Elliott: 5.44
5. Joey Logano: 5.63
6. Denny Hamlin: 5.67
7. Chris Beuscher: 5.88
7. Ty Gibbs: 5.88
9. Noah Gragson: 6.14
10. Erik Jones: 6.75 

Incidentally, the least predictable driver this season has been Josh Berry (13.0), followed by John Hunter Nemechek (11.78). Both of those drivers are off to better-than-expected starts and have been climbing up the charts recently. 

Driver Variations from the Norm

One of the key handicapping principles that I have always embraced in betting NASCAR is in finding drivers who are supposed to do better or worse than usual in any given race. For instance, when my final simulations typically project a driver around 15th in any given race, but they are suddenly pegged 10th in a certain event, I believe that driver has betting value. This is what happened with Berry at Las Vegas, when he was projected 27th after having been in the mid-30s for the races prior. 

So far in 2025, there have been 98 instances that didn’t involve trouble in a race in which my final simulations projected a driver at least 2.0 positions better than his average. Of those 98 instances, 62 (63.3%) of the projected drivers finished better than their season average finish position, by an average of 7.1 positions. The vice-versa situation is similar but not as dramatic. 

Using this powerful information for handicapping purposes, these are the average final simulation projections and average finishes for the 2025 season thus far, using my handicapped races (95% or better laps). Consider these going forward to spot drivers expected to do better or worse than usual.

Driver – AvgFinSimProj / AvgFinPos
AJ Allmendinger – 22.38 / 16.38
Christopher Bell – 7.89 / 9.78
Josh Berry – 22.43 / 16.86
Ryan Blaney – 7.17 / 8.5
Alex Bowman – 10 / 12.14
Chase Briscoe – 12.5 / 12.63
Chris Buescher – 13.25 / 12.13
Kyle Busch – 15.14 / 11.71
William Byron – 3.78 / 9.11
Ross Chastain – 15.38 / 10.88
Austin Cindric – 11 / 16.63
Cole Custer – 27.71 / 26.14
Austin Dillon – 24.25 / 16.75
Ty Dillon – 30 / 21.43
Chase Elliott – 7.67 / 11.56
Ty Gibbs – 14.75 / 18.38
Todd Gilliland – 26.22 / 20.78
Noah Gragson – 21 / 21.43
Justin Haley – 26.63 / 18.63
Denny Hamlin – 10 / 9.67
Riley Herbst – 34.63 / 24.63
Carson Hocevar – 19.71 / 19.57
Erik Jones – 25.25 / 20.75
Brad Keselowski – 19.43 / 22
Kyle Larson – 3.88 / 9.25
Joey Logano – 11 / 15.38
Michael McDowell – 23.38 / 17.75
John Hunter Nemechek – 29.89 / 19
Ryan Preece – 22.56 / 18.11
Tyler Reddick – 7.67 / 12.44
Zane Smith – 25.5 / 17.25
Ricky Stenhouse – 27.11 / 19
Daniel Suarez – 26 / 18.43
Shane Van Gisbergen – 22.33 / 24.83
Darrell Wallace – 9.25 / 16.5
Cody Ware – 36.86 / 29.71 

Sunday will be NASCAR’s first stop this season at the famed 2.67-mile Talladega Superspeedway. Known for its size, speed and pack racing, Darlington is considered one of the toughest tracks on the circuit for drivers. Many drivers would love to skip this race and the one in the fall for the potential dangers it brings. Those running out front at Talladega have the best chances of avoiding a “big one.” Typically, only the best drafting racers rise to the top. Thus, not surprisingly, the initial simulation finds three Penske Racing Drivers among the favorites. Austin Cindric is projected first, followed by William Byron of Hendrick Motorsports, then Joey Logano and Ryan Blaney. With that said, let’s dig right into the key stats for Sunday’s event. 

Who are the statistical leaders at Darlington?

These are the leaders from my weekly simulation factors in some categories I deem to be key in analyzing race data. Note that Talladega is one of only two superspeedway tracks on the circuit in which drafting skill is paramount. The Track and Track Designation (with Daytona) Ratings are shown below.

Steve’s Track Handicap Ratings:
1. Ross Chastain – 98.4
2. William Byron – 91.1
3. Cole Custer – 89.1
4. Noah Gragson – 89.1
5. Austin Cindric – 87.5

Steve’s Track Designation Ratings:
1. Austin Cindric – 98.2
2. William Byron – 91.8
3. Kyle Busch – 88.2
4. Chris Beuscher – 87.1
5. Bubba Wallace – 85.6

In looking at the leaders above, two things stand out right away. First, our top two projected drivers on the overall initial simulation are the only two drivers to appear on both lists. Second, some unusual names rarely appear on top-5 lists for any other venues, including Custer, Gragson, Beuscher and Wallace.

Steve’s Last 10 Overall Ratings (Momentum)
1. William Byron – 108.1
2. Kyle Larson – 106.1
3. Ryan Blaney – 101.9
4. Christopher Bell – 98.9
5. Tyler Reddick – 97.0

The lists above are a good place to start your handicapping. Noticeably, Cindric is missing from the momentum list as he currently ranks 11th in that stat heading into Sunday as he typically only has a shot to win at tracks like Talladega, Daytona or Atlanta. Ryan Blaney has a good amount of momentum now and is still in search of his first victory for the season.

Keep in mind, the following variances are off of my INITIAL SIMULATIONS from Tuesday and can/will change based on qualifying/practice on Saturday morning. Recall that the first list would theoretically be the drivers to follow, the next list the drivers to fade on Sunday.

Drivers projected higher than average simulation

1. Gilliland Todd: 16 / 26.22 (-10.22)

2. Cindric Austin: 1 / 11 (-10)

3. Busch Kyle: 8 / 15.14 (-7.14)

4. Logano Joey: 4 / 11 (-7)

5. Dillon Austin: 18 / 24.25 (-6.25)

6. Van Gisbergen Shane: 17 / 22.33 (-5.33)

7. Keselowski Brad: 15 / 19.43 (-4.43)

8. Blaney Ryan: 3 / 7.17 (-4.17)

9. Dillon Ty: 26 / 30 (-4)

10. Custer Cole: 24 / 27.71 (-3.71)

 Drivers projected lower than average simulation

1. Hocevar Carson: 34 / 19.71 (14.29)

2. Haley Justin: 35 / 26.63 (8.37)

3. Allmendinger AJ: 30 / 22.38 (7.62)

4. Briscoe Chase: 20 / 12.5 (7.5)

5. Gibbs Ty: 22 / 14.75 (7.25)

6. Suarez Daniel: 32 / 26 (6)

7. McDowell Michael: 28 / 23.38 (4.62)

8. Jones Erik: 29 / 25.25 (3.75)

9. Bell Christopher: 10 / 7.89 (2.11)

10. Bowman Alex: 12 / 10 (2)

I prefer to wait until after qualifying on Saturday to finalize my top picks, but here are some early predictions, with odds courtesy of DraftKings:

Top 5: Blaney (+1000), Cindric (+1600), Busch (+1100), Byron (+1200), Wallace (+2000)

Top long shots: Cindric (+1600), Wallace (+2000), Gilliliand (+2500), Buescher (+3000), Berry (+4000)

Favorites to struggle: Larson (+1400), Hamlin (+1400), Bell (+1600)

I will have my final race simulation for the Jack Links 500 available on VSiN.com sometime between the wrap-up of the qualifying sessions on Saturday morning and Sunday morning. Note that there is no practice session this week, and for the most part, starting position hasn’t proved very critical at Talladega.

Races at Talladega are always entertaining and intensely competitive, perhaps unparalleled on the circuit. Do yourself a favor and give it a watch, but remember to expect the unexpected!

RELATED ARTICLESMORE FROM AUTHOR