On-Track Passing in NASCAR by Track Type

by Sean Wrona

In last week's column I determined which drivers took the lead on track in the largest number of races as well as which drivers had the highest percentage of lead changes in an attempt to measure which drivers were the best duelists in NASCAR over the past 30+ years. In this column I extend that analysis by measuring each driver's lead change record on every track and track type. Just as the previous column provided an interesting and different perspective of who were the best drivers since 1985 over the entire schedule that sometimes differed from the raw finishing results, this becomes even clearer when judging drivers on individual tracks. On almost all tracks, there are certain drivers who are much more dominant duelists than you would have guessed they would be based on their finishing results, and others who were less.

I have divided all tracks that have held NASCAR races from 1985 to present into five track category types. Road courses are defined as all tracks that have both left and right turns. Short tracks are defined as all tracks shorter than 1 mile in length. Rough tracks are defined as all superspeedways between 1 and 1.49 miles in length. Although some would criticize me for grouping Darlington, Dover, Loudon, Phoenix, Rockingham in the same category since they have wildly different shapes and/or surfaces, they have more in common than you might think as many would argue all these tracks drive more like short tracks than horsepower speedways and they all particularly have abrasive surfaces that eat up tires and thereby drivers likely play much more of a role on them than they do on horsepower speedways. Finally, I group all unrestricted superspeedways of 1.5 miles or greater in length together; these tend to be more car-dependent than most of the other tracks. Many people would argue that unrestricted superspeedways with worn pavement (such as Atlanta and Fontana) do test driver talent more than the other tracks in that group, but even then, I think they are more car-dependent than the road courses, short tracks, and rough tracks are. Finally, I group all races at Daytona and Talladega in the restrictor plate category because the racing at these two tracks has little in common with the racing at any other track. For the years prior to 1988 before restrictor plates were introduced, I still do place the Daytona and Talladega races in the restrictor plate category, because the racing at those two tracks even prior to the plates was considerably different from the racing on the other superspeedways with the slingshot passing and the like, which wasn't seen as much anywhere else (I acknowledge there was some slingshot passing at Michigan and Pocono and more than there is today, but it still makes more sense to put Michigan and Pocono in the regular superspeedway category and the pre-1988 Daytona and Talladega races in the restrictor plate category to me, and I think it would to most others as well.)

For each type of track, I list the number of races each driver took the lead naturally along with their cumulative record on all tracks of that type, and then include each driver's record on every individual track. After I have shown all five lists, I will then provide a summary list that includes each driver's percentage on all track types, the number of tracks where each driver led, and the number of tracks where each driver had a .500 record or greater, which I believe are all instrumental in reflecting drivers' versatility in their dueling ability. I have not completely checked whether these data add up to the numbers in the previous column, and I know they do not. There appears to be a discrepancy of two lead changes between the sum total on these lists and the sum total on my previous column, and I can see quickly here that for instance Scott Wimmer had the worst record of 0-10 on the previous column, but it only adds up to 0-9 here, so there are clearly some mistakes and if I go further with this I will have to correct them. However, with only a difference of two lead changes out of over 9000 total lead changes, very few drivers should be significantly affected, particularly those with numerous lead changes to begin with.

Road courses

Since Jeff Gordon is the all-time road course winner in NASCAR history, it should be no shock that he took the lead on road courses more than any other driver, and he was indeed an exceptional duelist as well. Although several other drivers have greater lead change records, Gordon is still highly impressive particularly because he had to go up against his contemporary Tony Stewart, the best duelist of the period, for most of his career. However, Stewart looks considerably more impressive, as nobody comes close to even approaching his 14-4 record among drivers who led many road course races. While Gordon had to deal with Stewart for most of his career, Stewart had to deal with Gordon for his entire career and although he came one win short on the all time road course win list, losing to Gordon 9-8, it's noteworthy that Gordon won three of his road course races before Stewart's career started, thereby giving Stewart a clear advantage in the years he competed in every regard. Based on this, I would probably argue that despite Gordon's six consecutive road course wins (three of which came in races Stewart competed) that Stewart was the greater road racer.

While nobody should be surprised by most of the drivers who are highly ranked on this list, as Gordon and Stewart are generally considered the best Cup road racers of the last 20 years, Rusty Wallace, Geoff Bodine, and Ricky Rudd were the only drivers to manage to win on all three road courses in the period, Marcos Ambrose was considered the best road racer in recent years, and the Busch brothers are probably considered among the best in the post-Gordon/Stewart era, I think a lot of people will be surprised by Dale Earnhardt's high ranking in this list. Earnhardt is often slagged on by some for his weakness in road racing (as if this even really matters from a NASCAR perspective where there have rarely been more than two road course races in a year, although I do think it matters when comparing to drivers across other series somewhat) but this is a complete fallacy. Earnhardt was one of three drivers to take the lead on track on all three road courses (even Wallace although he won two Riverside races did not, as Wallace took the lead in the pits both times.) Possibly the most impressive pass of Earnhardt's career came at Riverside in 1986 when he managed to pass Bodine for the lead through the grass when Bodine was boxed behind a lapped car. Although Bodine managed to retake the lead shortly thereafter, a statement was made. Earnhardt was extremely unlucky to only win one road course race in his career and he was a much bigger threat on them than most people would realize by looking at the raw stats (although even there he did have the best average finish of any driver at Sonoma at the time of his death.) In 1993 particularly, Earnhardt was unlucky as he dominated both road course races until cars wrecked directly in front of him. To be fair, Earnhardt does have a losing record on road courses and has a greater record on every other track type, but he was still a far better road racer than he has been remembered.

Two other drivers did not lead quite as many road course races as some others but still had noteworthy and spectacular records. Martin Truex, Jr. has always been a great road racer and usually even in his bad years showed speed on them and ran better on them than he did on oval tracks, perhaps indicating a potential that had yet to be tapped. Although now it's almost a footnote in his career particularly because the track is no longer on the Xfinity Series schedule, Truex's win at the Mexico City Busch race in 2005 was considered a rather big deal at the time considering it was the first national touring series race south of the border, and attracted several Cup stars and top-notch road course ringers (especially Adrián Fernández, whom I expected to dominate the race in the Hendrick #5 car) yet Truex dominated when nobody expected him to. His 5-0 record at Sonoma is the best of anyone on any road course, and his 8-1 record is the best record in general among those with a significant level of success. The only time Truex was even passed was in the most recent Watkins Glen race in 2017 where he intentionally let Brad Keselowski pass to save fuel, but he knew Keselowski would not make it to the finish (there are people who would probably argue something like that is not even a competitive pass since Truex was still effectively controlling the race both before and after Keselowski passed.) Although Ernie Irvan lost his road racing dominance after his return to NASCAR competition in 1996, his dominance from 1991-1994 was impressive and his 7-1 record is truly spectacular. Not many drivers could get black-flagged after jumping the start and come back to win the race with no real strategy gimmicks or pot-sweeteners to get them back up to the front as Irvan did at Sonoma in 1992. However, Irvan unlike Truex does have one blemish as he was passed by Kyle Petty for the win in the rain-shortened Watkins Glen race of that same year, which was the only road course race where Petty took the lead on track in his career.

The most overrated road racer of the period is clearly Mark Martin. Although for a while he was talked about in the same breath as Gordon or Wallace on road courses, particularly when he won three straight Watkins Glen races from the pole in 1993-1995, his 1993 and 1995 wins in particular were extremely lucky. Although Martin did have a strong run in 1993 and he passed many cars after he kept having incredibly slow pit stops under caution periods, he did not take the lead on track until Petty spun out while leading and took Earnhardt with him, and I believe Earnhardt likely would have held Martin off considering how dominant he was on road courses and in general that year. I also read in the Bob Zeller biography Mark Martin: Driven to Race that apparently Martin's then-teammate Wally Dallenbach was ahead of him and ordered to let Martin pass earlier. In 1995, Dallenbach managed to go off pit sequence and take a massive lead on Martin, but a caution for oil on the track that many found controversial at the time allowed Martin to catch up and later pass Dallenbach for the win. Furthermore, Martin's Sonoma record of 0-4 is fairly shocking given that he won there and scored many other strong finishes. However, he only inherited the lead after Wallace went off track in that race and then was never contested. Martin was certainly good at maintaining the lead on road courses, but he really wasn't nearly as good as his four wins would indicate: note that he is essentially tied with Jeff Burton, Jimmie Johnson, and Ken Schrader, none of whom come close to Martin's road racing reputation (and Johnson particularly faced stronger road racing fields.) Juan Pablo Montoya, Terry Labonte, and Jamie McMurray look worse than I would have expected as well. Considering Montoya's road racing pedigree, surely he should have taken the lead in more road course races than drivers like Dale Earnhardt, Jr. and Sterling Marlin did. This is to say nothing about Junior and Marlin, whom these data indicate were underrated and better duelists in general than most would expect, regardless of track type (which may be one reason for their restrictor plate dominance, because dueling ability is probably most important in plate races.)

Short tracks

It wouldn't take a rocket scientist to predict the top three drivers on this list, because Rusty Wallace, Jeff Gordon, and Dale Earnhardt easily had much greater short track reputations than any other drivers in the period and no one else comes close (although maybe Darrell Waltrip would if I had more data from 1975-1984 short track races). However, I'm not sure anybody would be able to predict the order between those three. To determine which driver was more impressive in terms of the number of races led, we have to consider how many short track races for each were considered. Earnhardt has 100 races included to Wallace's 130 and Gordon's 147, giving Earnhardt the highest percentage of short track races led with 55% to Wallace's 54.6% to Gordon's 48.3%, and in terms of lead change percentage, these three drivers are ranked in precisely the same way, so I would argue Earnhardt was the most impressive short track driver in the period. Earnhardt, who had the second highest lead change percentage on short tracks (60.32%) to Tony Stewart (62.88%) among major drivers, was also one of only three drivers along with Bobby Hamilton and Harry Gant who had a .500 or higher lead change record on all four short tracks, although three other drivers (Kevin Harvick, Stewart, and Kurt Busch) did not compete at North Wilkesboro but had .500 or higher records on the three short tracks that are still on the schedule.

Earnhardt's mastery of Bristol was basically unmatched as he had a lead change percentage of 2/3 there, although Kyle Busch seems to be his closest heir in that regard, as Busch's record of 27-16 is quite close to Earnhardt's 28-14. Although Wallace and Gordon were both positive at Bristol as you'd expect, they weren't the dominant forces anywhere near the same degree as Earnhardt and Busch were, who were the two overwhelming standouts at the track in this regard. At Martinsville, everyone would easily and correctly guess that Gordon led more races there than any other driver, but he didn't hold onto the lead as well as others, and here Rusty Wallace had a massive edge (45-21) over both Gordon and Jimmie Johnson, probably because Wallace didn't really have a rival in his heyday (it seemed like he pretty much took over from Earnhardt on the short tracks in 1993 and years following before Gordon took over from him a few years later), while Gordon and Johnson both had to deal with each other for essentially their entire relevant careers at Martinsville. Denny Hamlin may be particularly unbalanced, but his 31-19 record says a lot for him and says that he might have been just as good as Gordon and Johnson there but just didn't have cars quite as strong as the Hendrick cars usually were there. However, it was Tony Stewart who had the best lead change record there at 23-13, slightly above Earnhardt and Hamlin. Clint Bowyer also did an excellent job at 11-7. At North Wilkesboro, Earnhardt was the obvious standout in terms of battling for the win, leading 14 different races, but Geoff Bodine was the most dominant force in terms of lead change percentage with an incredible 10-4 record, not to mention becoming the last driver to lap the field in a race, which is almost certainly unlikely to ever happen again now with the advent of stage racing and timed cautions. At Richmond bizarrely Gordon and Wallace led more races than anywhere else but they both had losing records. Just like at Martinsville, Stewart had the highest Richmond lead change record at 29-13 and this time nobody else was even really close. One big surprise you won't likely notice is Jamie McMurray's 4-0 record at Bristol. As much as he is thought of as only being a superspeedway driver, that's really selling him short, and this is proof. A lot of the times McMurray led on other types of tracks (like at Bristol) he took the lead right before entering the pits or it happened early in the race so you wouldn't be likely to notice, but he has a few other sneaky things in his lead change record like this that make him look better than you likely realize.

Several drivers never really figured out short track racing and the stereotypical one was Bill Elliott. Elliott was tremendously disappointing on this track type, particularly at Bristol where he went 4-11 and at Martinsville, where he never took the lead on track. Obviously, as is well known, his lack of balance on these sorts of tracks cost him the 1985 championship, but it wasn't so much that he didn't lead the races as he was a complete non-factor in them and often got poor finishes as well. Bobby Labonte looked very similar with a 13-23 record to Elliott's 11-23 but had one fewer race led. Labonte however was more balanced as he didn't look that bad in the races he was competitive except at Martinsville where he had a dismal 1-10 lead change record. Other disappointments include Ricky Rudd's 9-15 at Bristol (but that stems almost entirely from the 1991 spring Bristol race with the odd/even pit rules where Rudd kept gaining the lead under caution before Wallace instantly passed him on each ensuing restart), and Ryan Newman (whose 3-10 at Martinsville looks very bad.) However, in general people who were good on short tracks were good on short tracks and were relatively consistent on each, although some drivers obviously specialized on one (Hamlin and Johnson at Martinsville; Bodine on North Wilkesboro) more than at the others. A lot of people say Richmond is the short track that drives most like a superspeedway, and on this evidence, I'd be inclined to agree because a lot of the drivers who appear dominant on the other track types appear much weaker here while a lot of the drivers who appear weaker on the other short tracks (like Bill Elliott, Bobby Labonte, and Ryan Newman) appear much stronger there.

Rough tracks

As I said above, one may quibble with me grouping all these tracks together since they include an egg-shaped oval, a concrete oval, a paper clip, a triangle, and a D-shaped oval, which seem to have nothing in common, but I think these tracks have a lot in common due to how abrasive they are on the tires, how they tend to drive more like short tracks than superspeedways, and how I think the driver plays a bigger role on them than on some of the other track types. Although Jeff Gordon led in terms of the number of races led again, his lead change records on these tracks weren't particularly stellar, and indeed this was the only track type where he failed to post a .500 or greater record.

Kevin Harvick was the overwhelming standout on this type of track. Although he did not manage to lead at Rockingham naturally before the track stopped hosting Cup races, his 49-23 record (68.1%) is something pretty amazing considering he doesn't have a lead change record greater than 57.1% on any other track type. Although we all know about his absurd dominance at Phoenix compared to his more general respectability at most other tracks, his lead change percentages honestly weren't that different at Darlington, Dover, or Loudon either. At his worst of these tracks (Dover), he still has a 63.6% lead change record, which is higher than any other driver's overall mark except for two. Harvick's boss Tony Stewart as usual is not far behind and in second place in terms of lead change percentage at 64.5%. Although he was surprisingly weak at Darlington, he was very strong at Dover and Loudon, where he took the lead on track twice as often as he was passed. Martin Truex, Jr. is another master of the rough tracks with a lead change percentage of 64.3% although he was only particularly strong at two tracks: Darlington, where he surprisingly had the best lead change record of anyone (7-0, although I did not count Denny Hamlin's pass of him at the end of the 2017 Southern 500 because I counted that as Truex losing a tire), and much less surprisingly, Dover, where his 14-7 lead change record was an even higher percentage than Jimmie Johnson's. Ward Burton was the fourth best duelist on this kind of track with a lead change percentage of 61.5%. However, Ward was kind of a one-trick pony on this sort of track as he looked rather weak on all the other lists (note that he got four of his five wins on these tracks other than his utter fluke Daytona 500 win.) He definitely didn't have the versatility of a lot of the other drivers who are rated in his class (he's not even close to Bobby Hamilton, for instance), but if you're going to be a one-track-type wonder, it's still more respectable to be strong on the rough tracks only than on the superspeedways or plate tracks only, I would say. Much like his brother Ward, Jeff Burton was better on these tracks than any other track type and not merely Loudon, although Jeff was a lot more balanced than his brother.

Jimmie Johnson probably best combines relevance in terms of races led and lead change record as he has the fifth greatest lead change record on these tracks (60.45%) and he is one of only three drivers to earn a .500 record on all five of these tracks. His 39 races led is particularly impressive because Rockingham no longer held races pretty early on in his career, although Johnson certainly did benefit from two races a year at Phoenix more than the drivers above him did. The other two drivers were perhaps surprisingly Ernie Irvan and Bobby Hamilton. Irvan although generally regarded as a great driver is not highly regarded on this track type. He only won once on any of these tracks - his comeback win at Loudon in 1996 - but he is a lot better on them than you would think. His 11-4 record at Darlington is particularly eye-popping, especially considering he never won there. This is the type of track where Irvan was most unlucky throughout his career. He was the terminal natural leader in the 1990 Southern 500, 1996 fall Rockingham race, 1997 spring Darlington race, and both Loudon races in 1997. These five races made up most of the difference between Irvan's 15 wins and 22 TNLs. One would expect that had his career not been interrupted by injury, he would have likely fleshed out substantially greater diversity by filling in many of his gaps on this sort of track, because he was actually a better duelist on rough tracks than on any of the other track types except road courses. It shouldn't be as much of a surprise to see Hamilton showing so well here as he won at both Phoenix and Rockingham, but it may be a bit of a surprise considering his short career that he did manage to lead on all of them and score .500 records on all of them (even if he only had two more lead changes for than against.) What is impressive is that Hamilton was the only driver to lead on all five of these tracks *and* on all four short tracks, although to be fair he had exactly a .500 record on five of those nine tracks. Still, no one else did that. Earnhardt didn't do it. Gordon didn't do it. Wallace didn't do it. And Hamilton didn't have anything resembling their level of equipment. When you further consider that Earnhardt intentionally wrecked Hamilton out of the lead in the 1996 spring Rockingham race, a race Hamilton stood a very good shot of winning, it's possible he could have looked even better on this track type than he did.

Disappointments include Kurt Busch and Matt Kenseth, who seemed to particularly be masters on this sort of track from memory but both had sub-.500 records, Ryan Newman (who although he was disappointing almost everywhere in terms of lead change percentages, is generally considered much stronger on these sort of tracks - particularly Dover and Loudon - than he is anywhere else but still had a sub-.400 record), Denny Hamlin (whose 1-7 Dover record is actually fairly shocking), along with Geoff Bodine and Greg Biffle who most people would seem to think would have enough grittiness for this track type. Ricky Rudd having a mere 1-6 record at Darlington is particularly disappointing, as is Carl Edwards's 1-4 at Loudon, and Geoff Bodine's 7-14 at Darlington and 2-6 at Phoenix. Darrell Waltrip's 0-3 is particularly shocking since he won there and one would expect that a track that rewards smooth consistency and veteran smarts would reward him - even late-career drivers like Cale Yarborough, Richard Petty, and Benny Parsons led there but Waltrip did not in the recorded races, and he never really factored there in his later years (and of course he only won the Southern 500 he did win because of fuel mileage.) One of the most amazing records on the other end was Neil Bonnett's 10-3 record at Rockingham, and that did not even include all his wins there as some of them from the '80s did not nearly have enough footage for me to include them with confidence.

Overall I think just eyeballing the lists you can see that these tracks although classified as superspeedways are much more like short tracks, because the rankings on this list come far closer to matching the rankings on the short track list than the rankings on the superspeedway list, and the drivers with strong lead change percentages on the one tend to have strong lead change percentages on the other. They definitely are different though, as Gordon, Earnhardt, and Wallace all have records either just above or below .500, and while they were all great on these tracks, they weren't exactly the best discipline for any of those drivers.

Unrestricted superspeedways

I find this list to be the perhaps the most interesting of all the lists because I think many people would argue that in general the driver plays the least role on this sort of track and one could almost consider this list a measure of who has had on average the strongest cars in their career more than anything having to do with driving strength. It's also a lot harder to make comparisons here in terms of races led considering all twelve of these tracks are on today's schedule but prior to 1994 there were only four such tracks (Atlanta, Charlotte, Michigan, and Pocono), so there is going to be a massive skew in favor of modern drivers and against drivers who peaked in earlier eras. That doesn't mean that information can't be gleaned on certain drivers based on performances on these tracks, but it does mean that it is likely harder to separate the driver and car effects in most of these cases. Having said that, I will grant that driving talent comes more to the fore in superspeedway races with worn pavement and also that I have more respect for drivers who excel at Pocono and Indianapolis than those who excel at most of the other intermediates. As with all other lists, Jeff Gordon leads in terms of races led, but once again (as with all the other lists) he is not the best duelist.

The best duelist on the high-horsepower superspeedways was Davey Allison, whose record of 41-20 is fairly absurd given how equipment-centric this statistic seems to be and how Allison's lead change percentages on all other track types are very disappointing in comparison (51.2% on plate tracks, 33.3% on road courses, 49.0% on rough tracks, and 47.6% on short tracks.) Although I half-recanted on calling Allison overrated in my previous column when considering his overall lead change percentage over 54%, I think these results convince me that I was right the first time. The only tracks where he passed other drivers more often than he was passed were tracks where equipment strength was paramount, while on those tracks where the driver would seem to play more of a role (the road courses, short tracks, and rough tracks) he was not all that special despite having equipment dominant enough to take the lead twice as often on the unrestricted superspeedways as he was passed. Allison was especially impressive at Charlotte (13-2) and Atlanta (7-2) but did have a .500 or higher lead change percentage on all four tracks where he competed. This alone to me is mostly a measure of Robert Yates's dominant engine program, although to be fair I'll note that the other three major Yates drivers (Dale Jarrett, Ernie Irvan, and Ricky Rudd) were nowhere near as dominant on this list, so likely Allison played some role in his dominance on these tracks, although I suspect his famous 007 chassis played a larger role (note that in the first Charlotte race in the fall after Allison's death in 1993, Irvan had one of the most dominant races of anyone in the entire 30-year period leading 328 out of 334 laps, and even Allison never dominated a race like that.) I feel it does say a lot that Mark Martin borrowed a #28 engine for the season finale in 1990 in an attempt to beat Dale Earnhardt for the championship yet Allison finished only 13th in the championship that season while Martin was 2nd. Allison was spectacular on this track type for sure, and if he did anything close to this on any of the other track types, he would live up to the hype he usually receives. However, based on what he did on the tracks where I think the driver plays a bigger role, I think he was more in the same class with drivers like Geoff Bodine, Sterling Marlin, and Greg Biffle than where he is usually rated. He was a very good driver, not a great one, and I'm actually starting to think people have Davey Allison and Dale Earnhardt, Jr.'s legacies reversed. Junior was actually closer to legitimate greatness I believe, although they were both rather second-tier to be sure.

I have no such criticism of Jimmie Johnson, the driver with the second highest lead change percentage (61.22%). Everybody knows Johnson is the best intermediate driver in the NASCAR cookie-cutter era, but he was nowhere near as much of a slouch in the other categories as Allison was, as he had a lead change percentage of 50% or higher on all track types except for restrictor plate tracks. Johnson was one of only eight drivers to lead on all twelve of these tracks. This list skews very modern considering that some of these tracks like Chicagoland, Kansas, and Kentucky haven't been around all that long and had one race a year quite often. In addition to Johnson, Kyle Busch, Matt Kenseth, Tony Stewart, Kevin Harvick, Carl Edwards, Brad Keselowski, and Joey Logano led naturally on all these tracks. However, Johnson was significantly more impressive on this track type than on all the others were, which should surprise no one. Johnson's most staggering dominance came at Indianapolis, where he had an absolutely unreal 11-1 lead change record. Although he was more consistent on many of the other tracks and had quite a few bad races at Indy, when he was in the fight for the win at Indianapolis, he was basically unbeatable. Johnson had a .500 or higher record on eleven of these twelve tracks (all except Homestead, which makes some sense because he tended to take it easy in that race to secure the title so frequently) and only two other drivers had .500+ records on eleven of the twelve tracks: those being Stewart (who barely missed at Atlanta) and Edwards (who barely missed at Charlotte, which is fairly shocking considering his dominance on 1.5 mile ovals, particularly Atlanta and Texas, in general.)

As with Davey Allison, there were other drivers who performed much better on this sort of track than on the other track types. Greg Biffle had a very strong lead change percentage of 57.7% on unrestricted superspeedways, but failed to manage a positive lead change record on any other track type, indicating he is likely also overrated. Carl Edwards and Bill Elliott were both similar but a bit more balanced, as both of them had positive lead change records on all three types of superspeedways (plate tracks, rough tracks, and unrestricted superspeedway) but losing records on both road courses and short tracks. Elliott however was best on the rough tracks, unlike Allison, Biffle, and Edwards, who were much better on the unrestricted superspeedways, so I have considerably more respect for Elliott than the others in this group. Terry Labonte and Kasey Kahne are two more drivers who had above .500 records on unrestricted superspeedways but had sub-.500 averages everywhere else, probably arguing once more that their cars were better than they were, although Labonte does have a few other things in his factor that I will discuss in a future column. Joey Logano had positive records (barely) on superspeedways and plate tracks, but negative records on the three tack types where the driver played a bigger role. Bobby Labonte had a below-.500 record on all five track types, probably marking him as the worst champion of the period, but as one would expect, he had a slightly higher percentage on the unrestricted superspeedways than on the other tracks indicating his cars were doing more of the work.

Drivers who may be underrated are those who had significantly worse lead change percentages on this track type than on the others where the driver may play more of a role. The foremost such driver appears to be by far and away Bobby Hamilton. He had a 50% lead change percentage on all the other four track types (and in fact was the only driver in Cup in the entire 30-year period to have a .500 or greater record on all four short tracks and all five rough tracks, along with a solid 1-1 road course record at Sonoma in 1998 where he passed Ricky Rudd before being passed by Jeff Gordon for the win.) However, on the unrestricted superspeedways, where car probably plays the biggest role, Hamilton had a miserable 2-5 record. No other driver has even close to as large a difference between their record on the horsepower tracks and all other tracks. This further cements my opinion of Hamilton as the most underrated driver in Cup of the last 30 years and it makes sense. Petty Enterprises had not won for 13 years prior to his first win there, and only won once at Martinsville since (where Hamilton had dominated for years and they still likely had great setups when John Andretti drove for them), then he strongly improved on the also-underrated Sterling Marlin at Morgan-McClure finishing 10th in points a year after Marlin finished 25th well after their heyday, and then he gave Andy Petree one of his only two wins as a car owner proving that when the equipment was equalized (on the plate tracks) he could be competitive, passing no less than Tony Stewart, the best duelist of his generation to win.

Hamilton was not however the only driver who was substantially better on all the other track types than on the unrestricted superspeedways, and I think most of the rest are underrated too. Dale Earnhardt's lead change percentage of 50.75% here is surprisingly weak indicating his equipment may not have been as strong as everyone thought it was (and it may explain why he only won the most races in a season twice; however, I also sometimes think that may be because Richard Childress made an intentional tradeoff in favor of reliability instead of speed, which continues to this day, because the NASCAR points systems invariably reward reliability over speed). Harry Gant had a 53.9% lead change percentage or higher on all track types except this one (and road courses, where he never led) indicating that he was greatly pulling above his weight, which everyone already knew since he was the only driver to score wins for the Hal Needham team and the only driver to score an unrestricted win for the Leo Jackson team. Alan Kulwicki was slightly weakest on the unrestricted superspeedways; however, he was fairly balanced, as he was between 46.2% and 47.0% on all unrestricted ovals (Kulwicki never led naturally on a plate track or road course.) I think Kulwicki may deserve more credit as a team manager than he gets, but somewhat less as a driver than he gets. Considering how weak everyone says his equipment was, I expected a larger disparity between his short track and superspeedway results than what we see here. However, I think Kulwicki is also hurt a lot because he was such a great qualifier, as he earned 24 poles compared to his 5 total wins and was passed on the start a lot; he might look a lot better if I did not count start and restart passes. Although he didn't lead very often, A.J. Allmendinger also had a substantial difference between his performance on the unrestricted superspeedways and everywhere else, as he had a .500 record or better on plate tracks, road courses, and rough tracks, and a 37.5% lead change record on short tracks, but only 25% on unrestricted superspeedways, so it appears he is pulling above his weight too (and before I did this analysis I thought he was overrated, not underrated, but now I am forced to change my mind.) Marcos Ambrose has a very similar record all-around to his ex-teammate Allmendinger, which makes sense since they likely had similar equipment. Morgan Shepherd and Ryan Newman were about even on all unrestricted superspeedways, but were worst in this category, but I don't think Newman is underrated, since he is much higher rated than any of the other drivers I am mentioning here. Clint Bowyer, Sterling Marlin, and Jamie McMurray have higher records in three of the other four categories, with McMurray worse on road courses, Marlin worse on short tracks, and Bowyer worse on rough tracks indicating they all may be underrated.

Additionally, if you judge this as simply a measure of equipment strength, there are a few anomalies that don't seem to make sense. For all the criticism of Dale Earnhardt, Jr. supposedly being a great superspeedway driver and mediocre everywhere else, the reality is almost the exact opposite. Junior had greater than a 50% lead change percentage on all track types except for unrestricted superspeedways, a very similar record to Denny Hamlin, who was above 50% on all track types except for rough tracks and superspeedways, and also was the weakest here (if you exclude Hamlin's stunning 20-8 Pocono record, he was a miserable 52-80 despite championship caliber equipment) and Brad Keselowski (who is weaker on this track type than any other except for road courses.) I suspect in Hamlin and Keselowski's case, a lot may have to do with pit stops. Hamlin has on average seemed to have the fastest pit crew in his career, thereby placing him in the lead more often, and Keselowski's team makes more pit strategy gambles than any other. Maybe they don't have as fast cars as everybody seems to think they do and are overachieving in them. As some would argue it's often easier to make a pass on this sort of track than the others (as the fastest cars go straight to the front) it could just be that drivers like Hamlin and Keselowski do a much better job of holding off drivers on tracks where it is harder to pass. However, on the flip side, many people also argue that most of these tracks are especially sensitive to aero push, which would argue just the opposite. As a result, I can't really come to a conclusion in regards to explaining these results. It's particularly puzzling since they seem to have had championship-caliber equipment for most of their career. I can't figure this one out. In Junior's case, maybe the strong backlash he received because he was the most popular driver of his generation but not the best driver of his generation masked the reality somewhat. Junior did lead more races naturally than any other driver in 2002 and 2003, which is at least as solid a record as Davey Allison (I would actually say superior) yet nobody thinks of him the same way at all. However, Junior delivered the first wins to DEI, while the Harry Ranier team (before it became the Yates team) had been winning for many years with many different drivers before Allison got there, and remained just as dominant, if not more, with Irvan and Jarrett. However, DEI collapsed without Junior, even with a future champion in Martin Truex, Jr. Maybe Junior's DEI equipment wasn't as good as we think (except on plate tracks.) Maybe it was more of a Morgan-McClure style team that specialized on plate tracks and wasn't as good everywhere else. It's something to think about. Jeff Burton was also very similar to Junior in that he had higher lead change percentages on all track types except for this one, indicating maybe Roush equipment was a lot weaker than I thought it was in the late '90s, although that still doesn't explain why Kurt Busch, Matt Kenseth, and Carl Edwards surpassed what Burton did so quickly.

Restrictor plate tracks

Plate tracks are as everyone knows the biggest crapshoot in NASCAR. However, when you consider that Daytona and Talladega are the most likely tracks to have a high number of lead changes, they may be the tracks where passing is most paramount and where dueling ability is most rewarded. Because of the endless passing and repassing that pack racing generally creates, there is generally a smaller range in lead change percentages in this category than in all the other categories. Indeed, while almost all other tracks have drivers with lead change percentages of 60% or higher and a significant number of races led, plate tracks are the big exception. The high number of lead changes that these tracks produce (not to mention the amount of passing and repassing seen at these tracks) tend to moderate almost every driver's overall lead change percentage, making the weak duelists look stronger and the strong duelists look weaker. However, that's not to say there isn't a correlation between the drivers who are the strongest duelists on plate tracks and those on other tracks because there definitely is. And as you will see below, as messy and disorganized as many plate races look when you are watching them, there is definitely a lot of meaning in the below results, as the drivers who were most dominant in plate racing overall do appear at the top of this list.

This list definitely matches the conventional wisdom that Dale Earnhardt was the greatest plate racer ever. Although Jeff Gordon led more races than Earnhardt within the given time period, Earnhardt certainly led more races naturally than Gordon if you consider the years before 1985 as well. Earnhardt also had a staggeringly higher percentage of races led than did Gordon or his son, with Earnhardt taking the lead on track in a ridiculous 53 of his 63 Daytona/Talladega races counted in the period, while Junior's ratio of 48/71 and Gordon's ratio of 56/92 both look considerably more modest. As with the short track list, all three of the clearly most dominant plate drivers are ranked in the same order in both percentage of races led and lead change percentage, with Earnhardt first in races led percentage and lead change record among the three main leaders, and Gordon third on both lists, with Junior second in both of those categories on those list, as opposed to the short track list, where Rusty Wallace was second in both categories. While a lot of people liked to criticize Earnhardt, Jr.'s success due to the fact that Michael Waltrip was dominating in the same equipment (and indeed he was: the difference between Waltrip's 32 races led on plate tracks alone and only 9 on all other track types is definitely one of the most unbalanced records in NASCAR Cup history, just as everyone already thought it was) it doesn't seem all that well-founded. While Gordon's plate career still impresses me more than Junior's, and while I think Junior generally had an equipment advantage on restrictor plate tracks because of his 2001-2004 seasons, it's clear that objectively they were close, and there is a case for Junior being the second greatest plate driver ever.

A lot of people throw Sterling Marlin and Michael Waltrip in the same category in that they were both substantially more dominant in plate races than on all other track types, and won all their races in a three-year period when they were driving for the dominant restrictor plate team. However, this argument doesn't really hold water for a number of reasons. While both were most dominant on plate tracks, Marlin led only 39 of his 94 races led on them (41.5%) while Waltrip led 32 of 41 (78.0%). Although Waltrip led plate races a lot even in his non-DEI years (particularly because he continued on into the tandem drafting era where over half the field led most of those races while Marlin didn't), he rarely was any sort of significant factor for the win when he was not driving for DEI. That was not the case for Marlin, who was the TNL of the 1992 spring Talladega race before Davey Allison beat him out of the pits, managed to finish 2nd for the Stavola Brothers in the summer Daytona race in 1993 (and went airborne in the Daytona 500 that year mid-race and still somehow recovered to a top ten finish), and was the TNL again while driving for Ganassi at the 2002 Pepsi 400 before Waltrip beat him out of the pits (and none of his Ganassi teammates were even competitive on the plate tracks.) Marlin led Daytona/Talladega races naturally as early as 1986 and as late as 2007, both times in rather questionable equipment, and Marlin based on his record was a very good/borderline great driver everywhere while Waltrip was generally mediocre everywhere except the plate tracks (and strangely, the rough tracks, where he was somewhat decent.)

Among drivers who led a fairly sizable number of plate races, perhaps surprisingly Darrell Waltrip had the highest lead change percentage. Although this period was not exactly Waltrip's heyday and he only won one restrictor plate race due to fuel mileage, his lead change percentage of 57.41% was higher than any other driver that led ten or more races naturally. However, the era in the late '80s when Waltrip was most competitive on plate tracks did not have as many lead changes per race as some other eras did (such as the era at Talladega before the plates were introduced, or the wicker bill era, or especially the tandem drafting era), so it was easier to have a more top-heavy lead change percentage as the leader during a given run would not automatically be passed by the end of a run. For these reasons, I consider Brad Keselowski's dueling record on plate tracks to be the most impressive. His percentage of 57.29% is barely behind Waltrip's but he led many more races and came to prominence in the tandem drafting era, when almost all drivers who led those races naturally had lead change records within one lead change above or below .500 because nobody could hold the lead. For him to rack up such a high percentage despite competing in this era (although to be fair his period of greatest dominance in plate racing came shortly after the tandem period ended) says a lot for him. This record for Keselowski did not include his last lap pass in his debut win at Talladega in 2009 when Carl Edwards blocked him, Keselowski refused to back down, and Edwards went airborne into the catchfence. Had that pass been included, he would have a higher lead change percentage than even Waltrip despite competing in a much more competitive era. Keselowski definitely comes closer to matching Dale Earnhardt's cleverness in managing the draft than maybe any driver since, as many commentators have already noticed. In addition to earning the only win for a single car team between Ricky Craven in 2003 and Martin Truex, Jr. in 2015 in his first plate victory, he also managed to take the lead in the 2012 spring event from Matt Kenseth (who had a much faster car) on a green-white-checkered and break the draft to prevent Kenseth's teammate Kyle Busch from getting a run on him, considered an unheard of move at the time not to mention claiming a last-gasp win to advance in the chase in 2014 and he probably would have repeated the feat in 2016 had he not blown an engine. When you further consider that Keselowski was driving for Penske Racing, for whom neither Rusty Wallace or Kurt Busch managed to win a plate race, his record looks even more impressive. He effectively almost singlehandedly brought Ford back to relevance in restrictor plate racing for the first time since the heyday of Robert Yates Racing in the '90s after a string of long domination by the Chevy teams of Hendrick, DEI, and Childress. Because he hasn't won a Daytona 500 yet, he is at the moment overlooked despite his many Talladega wins. Other drivers who were strong duelists include once again Tony Stewart (who was third at 55.07% behind only Darrell Waltrip and Brad Keselowski among drivers with a large number of races led, in spite of never winning a Daytona 500 and only winning one Talladega race in controversial fashion). Not surprisingly, David Ragan (the millennial Michael Waltrip) is one of the top duelists (in fourth place at 55.00%). Much more surprisingly, Casey Mears is directly behind Ragan. The remainder of the top ten (Earnhardt, Marlin, Earnhardt, Jr., Dale Jarrett, and Jamie McMurray) makes much more sense.

Four drivers led ten or more races at Daytona and did not win at Daytona 500, and one can argue that both Kyle Busch and Tony Stewart were very unlucky not to win Daytona 500s, especially Busch, who has led more races at Daytona than all drivers except Gordon, the Earnhardts, Marlin, and his teammate Denny Hamlin. Ken Schrader was very unlucky not to win at Daytona generally as he led eleven races there and he was considered one of the best restrictor plate drivers for most of his Hendrick heyday. Greg Biffle was also a much better restrictor plate driver than many people realized, but didn't get the wins, probably in part because Roush simply wasn't as competitive as the Chevy teams on plate tracks in most of his peak seasons. At Talladega, Kurt Busch was very unlucky not to have won as he led 18 races there. One can argue Jeff Burton was as well because he led sixteen races there, although I can't think of many races that Burton really should have won (he was badly inflated by the tandem drafting races and RCR's domination of them). Although Ryan Newman is a weak duelist on almost every track, he looks surprisingly good at Talladega, where he has not won but has led fourteen times. However, his rather lucky Daytona 500 win definitely makes up for that. Kasey Kahne has more races led than any other driver with no restrictor plate wins, but that probably overstates his competitiveness because most of those were accumulated , and he wasn't really a race factor. I think the number of races led does have value, but it has very little value among drivers who competed in the tandem drafting era, since plenty of people led those races (sometimes even often) despite not really being threats, so I tend to find most stars of the 2000s a bit overrated here. I think on this list the lead change percentage may have more value than the races led totals, which is not something I necessarily agree with on the other lists.

Although I already mentioned the interesting results in the earlier sections, I will conclude with a summary table that lists each driver, the number of tracks where they led naturally, and the number of tracks where they had a .500 or greater record below. Only three drivers: Tony Stewart, Kyle Busch, and Brad Keselowski managed to lead naturally on every track where they made a start, although Stewart did so only barely taking the lead for one lap at Kentucky on a restart, but that did give him the edge over Jeff Gordon, who led everywhere else but there. Almost all legendary drivers given enough included races had a .500 or greater record at ten or more tracks, although three drivers you might not expect: Jamie McMurray, Jeremy Mayfield, and Bobby Hamilton (until this article anyway) are on this list as well. Tony Stewart and Jimmie Johnson were as usual the biggest standouts with regard to having the most tracks with a .500 record or greater, with Stewart managing to do so on 22 of the 24 tracks he competed on, only missing at Atlanta and Darlington by one lead change each. Jimmie Johnson did so on 20 of the 23 tracks where he led, failing to lead only at Watkins Glen at the moment and also having a losing record at Bristol, Homestead, and Talladega. Ryan Newman as usual appeared especially weak by this metric, only managing a .500 or greater record on four of the 21 tracks where he led.

Sean Wrona is the Managing Editor of racermetrics.com, the Webmaster of race-database.com, the winner of the 2010 Ultimate Typing Championship at the SXSW Interactive Conference in Austin, and the ratings compiler and statistician for the Mensa Scrabble-by-Mail SIG. 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. You may contact him at sean@racermetrics.com.

---