On-Track Passing in IndyCar

by Sean Wrona

Here I begin to undertake the same analysis I did for NASCAR over the past two months, only this time for IndyCar racing. Just as I did for NASCAR, I have similarly collected every lead change from every CART, Champ Car, IRL, and/or IndyCar race from 1982 to present. The 1982 Indy 500 is not yet counted in the analysis because that was technically sanctioned by USAC and was one of the two Indy 500s along with 1981 in that era that did not count for CART championship points. Because IndyCar racing was more popular than NASCAR in the early '80s and the races also tended to be shorter, there was much better coverage of CART races in the '80s than that of NASCAR. Additionally, races particularly in that period tended to have far fewer on-track lead changes than NASCAR races did, so usually they would be far more likely to all be cited in contemporary newspaper reports even if there was no video footage. As a result, there were only six missing races for which I did not feel I had enough data and the results for the earlier seasons are likelier to be more accurate than those for NASCAR. Theoretically, there is likely enough data for me to go all the way back to the CART debut season in 1979 (not to mention earlier Indy 500s that were also covered in full), but I'm choosing to release it now because I wanted to start this series before the 2018 IndyCar season opener at St. Petersburg next week. Similarly, there are certain NASCAR seasons that still have information available and/or race footage on YouTube that I have not covered. I will try to fill in those gaps (along with those that will result from the current series races) later. I actually collected a lot of these data before I collected a lot of the NASCAR data last year, but decided to release the NASCAR data first because the season started earlier.

I generally used the same definitions for what counted as a lead change and what did not count as I used for the NASCAR analysis. Inheriting the lead did not count when the previous leader at the line crashed, spun out, hit a wall, went off track on a road course, had a mechanical failure, ran out of fuel, had a fuel pickup issue, or had a penalty. However, I did count all other instances where a driver made an on-track pass for the lead and this time (because I did most of this analysis earlier) I counted all instances where there was an on-track pass for the eventual lead: for instance if a driver who was running second among those among the main pit sequence but fourth on the track passed the leader who was leading those among the main pit sequence but third on the track because two other drivers stayed out and had not pitted yet, I would count this as a pass for the lead. For the IndyCar analysis, I did this for all seasons, including races such as the Long Beach race in 1995 where Al Unser, Jr. passed Michael Andretti to take control of the race even though it was not for the lead, and Josef Newgarden's 2015 win at Barber, where his pass of Hélio Castroneves in the midfield did likewise. I since decided eventually that this was making it far more complicated (even if it was also in a certain way making things more accurate) and I eventually will want to rerun this analysis counting only passes for the lead at the line, but that will take months and I'd rather release what I've already done first. As with my NASCAR analysis, I will continue to count passes for the lead entering the pits if the same driver held and maintained the lead after the pit sequence ended, as with Nigel Mansell's pass of teammate Mario Andretti at the 1993 Indy 500.

One of the likeliest flaws of this analysis is that I usually did not count passes on outlaps if the passes were made within one lap after the driver being passed left the pits (I think there were a couple of exceptions here that I did include such as Scott Dixon's pass of Charlie Kimball at Pocono in 2013 and Kimball's pass of Simon Pagenaud at Mid-Ohio a few weeks later... both of those were fairly debatable in that regard.) I realize especially on road courses, many would argue that instances when a driver passes another driver on cold tires are actually among the most impressive sorts of passes, but in those cases, it's hard to tell how much credit to give the pit crew and strategists and how much to give the driver. There's a strong chance I'll change my mind with regard to these sorts of passes for both IndyCar and NASCAR when I rerun the analysis because I really haven't been as consistent about this as I should have been, and it's more likely to make more of a difference here because in many Champ Car and IndyCar road course races, there are extremely few on-track passes for the lead (quite often none) and therefore any sort of pass no matter how rare makes more of an impact, so I acknowledge this as one of the main flaws of the analysis at present. It would perhaps make more sense to count all these as passes for the lead (even on ovals), especially since I always counted passes when one driver had fresher tires than the other or one driver passed another driver when they were both trying to conserve fuel mileage (such as Danica Patrick's pass of Castroneves at Motegi in 2008, or Bertrand Baguette's pass of Patrick at the 2011 Indy 500, which bizarrely made him TNL of that race even though they were both generally non-factors), and I will likely change my mind and do that in future. I've never had good instinct on whether passes on out laps should count, and they are frequent enough that it might make a considerable difference. Additionally, as usual, particularly in the '80s and early '90s, there were a lot of races where I had to decide whether passing a car that was having a slow mechanical breakdown counted as an actual pass or not, and fans can quibble about these. However, in the oval races, I think I did a much more consistent job. I'm not sure this sort of analysis works as well for road racing generally.

Regardless of the flaws of this analysis, as usual there are drivers that turn out to be consistently overrated and/or underrated by all metrics. This time I am going to attempt to combine the analysis of two of my earlier NASCAR articles into one article this time. The below table combines the statistics I used in the original On-Track Passing in NASCAR article with the Lead Shares and Clutch Passing article. The table lists each driver's number of races led naturally from 1982-present (minus the six races not counted), lead change record, lead shares, cumulative races led, terminal natural leads, wins, races leading the most laps, and races with the most lead shares. I also list each driver's TNL lead change record (each driver's lead change record on the final pass of the race only) along with a clutch statistic (which is a difference between each driver's lead change record on TNL passes to lead change record on all passes... the drivers who have scores the furthest above 0 are the most clutch and those who have scores the furthest below 0 are the least clutch.) Listing all these statistics for all drivers at one glance makes it even easier to draw conclusions. Overrated drivers will generally have CRL > LS (indicating that their teams place them into the lead more often than they do), wins > TNL (indicating that they win more races due to luck than on-track passing), and LML > LSL (indicating that they benefit more from their teams putting them into the lead more than they do through their own efforts.) The most overrated drivers will also have more disappointing lead change records and clutch lead change records than you would expect. I'd hesitate more to call drivers overrated if they had strong lead change percentages or clutch passing records. Underrated drivers will have the reverse of these statistics, indicating they put themselves into the lead more than their teams do. Drivers are sorted based on the number of lead shares they had, and then sorted by cumulative races led for those drivers who have no lead shares. While I do have separate tables for both CART/Champ Car and IRL/IndyCar I chose to group all the races for both series together, even though I acknowledge to some degree it is apples and oranges. I used IndyCar's decision to include all races from either side of the split equally in their summary statistics as a precedent for this decision.

Just as before, if there were no on-track passes for the lead, which was far more common in IndyCar than it is in NASCAR, I give the polesitter an entire lead share and TNL and race led naturally. This is why most of the top drivers (particularly those who are strong on road courses, where there are fewer lead changes) have significant differences between their number of TNL passes and their number of TNL in general (which includes races where each driver won the pole but were never passed on track.)

Michael Andretti in IndyCar, much like Jeff Gordon in NASCAR, was the most dominant driver by all metrics in terms of leading IndyCar races. Although he has more wins than any other driver since 1982 (albeit barely since now Scott Dixon is only a single win behind), this actually underrates the level of his dominance by a significant margin. While there are other drivers who are close to him in wins, no one is remotely close to him in any of the other categories. He has 13.5 more lead shares than any other driver, 14.4 more cumulative races led than any other driver, 13 more TNL than any other driver, 14 more races where he led the most laps than any other driver, and 13 more races with the most lead shares than any other driver. These are extremely large differences that indicate that despite the fact that Andretti is still (for the moment, but probably not for long) the third-winningest driver in IndyCar history, he may in a sense actually still be underrated. His three more lead shares than CRL and six more TNL than wins would seem to indicate this. Some of the drivers close behind him in wins clearly benefited from the strength of their teams and their pit crews, while we all know Andretti was let down a lot by his cars' mechanical breakdowns. This is usually blamed on Andretti's aggression being too much for his equipment to handle, while more conservative drivers like Rick Mears, Bobby Rahal, Al Unser, and Al Unser, Jr. capitalized repeatedly as a result. Additionally, some might blame Andretti's greater propensity to crash for losing several wins. I calculated the crash DNF percentage for Andretti and each of these more conservative drivers because I thought it might be illustrative. I counted races marked as 'spun out' but did not count races where a driver was listed as having wrecked in practice or qualifying. Michael Andretti had a crash DNF percentage of 33/317 (10.4%), indeed higher than these other drivers, but not by too much for the most part. Al Unser, Jr.'s was 32/329 (9.7%), Al Unser's was 28/321 (8.7%), Rahal's was 21/265 (7.9%), and Mears's was incredible: 11/203 (5.4%). For the most part, I don't think Andretti's crash DNF percentage was so much higher to make up most of this difference. It really was the mechanical breakdowns, and I tend to be of the school where I don't blame drivers as much for those. One of the key examples of this in IndyCar racing was the 1987 Indy 500 where Mario Andretti was so dominant he decided to intentionally slow down to conserve his equipment and that is actually what caused it to break. I think from the driver's perspective mechanical failures are more luck than skill, and Michael had pretty bad luck on a race-by-race basis. F1metrics argued convincingly that only Alain Prost likely had an impact on the reliability of his teams so I think any criticism of the Andrettis on that basis is unwarranted. Did Mears and the Unsers benefit because they were more reliable or because their teams (like Galles and Penske) simply did a better job of conserving the equipment and the drivers got the credit? It's not like Newman-Haas wasn't known for mechanical breakdowns before Michael arrived (although not as much after he left, as mechanical reliability improved markedly for everyone in almost all series in the 2000s.) However, I do think Andretti's equipment advantage at times was so absurd that he still may not have been the best driver of the era (like in 1986 when he had future F1 engineering titan Adrian Newey as his chief engineer in 1986 and lost the title to Rahal.) However, I probably now like his career more than I did a couple years ago. And unlike his NASCAR counterpart Gordon, Andretti was definitely one of the top duelists in his era too, although not the absolute best one. Curiously, despite Andretti's reputation for choking races due to equipment failures, he actually had a higher TNL lead change percentage than overall lead change percentage, indicating he was in fact better at the end of races. Just not at Indy maybe, which is all some people seem to care about.

Paul Tracy and his rough modern-day counterpart Will Power also both seem to be extremely underrated by this as each had 3-4 more lead shares than cumulative races led, with substantially more TNL than wins and races with the most lead shares than races with the most laps led. However, most of that difference in Tracy's case probably is a result of his crashing. He had a reputation as one of the worst crashers among major drivers in IndyCar, and that does seem to be justified. With 62 crash DNFs in 281 starts (22.1%) he exceeds all the drivers in the previous paragraph by far, and it is easily enough to make him look underrated. I do not count passing a wrecked car as an on-track pass; if I did so drivers like Tracy and Greg Ray (16/73; 21.9%) would not necessarily look underrated here. If I forfeited TNL status to drivers who crashed while leading (as many others would likely do), drivers like them would look significantly worse. Power, who also has a reputation for crashing after crashing in three straight season finales from 2010-12 losing the points lead each time, is actually significantly better than those drivers at avoiding crashes (18/187, 9.6%) It's just that his crashes tended to come at inopportune times so they were more noticeable. While I previously more or less equated Tracy and Power in my mind for costing themselves a lot of big results by crashing, Power looks significantly better and significantly more underrated. He has seven more TNL than wins while Tracy only has four, and Power's tend to be being beaten out of the pits or losing on strategy (I don't think most of his TNLs where he did not win were in races he crashed, while the same can probably not be said for Tracy.) Additionally, Power for his reputation of not being clutch due to the crashes in late-season finales (even though 2011 was not his fault) actually is extremely clutch in taking the lead on track later more than he does in the beginning. Despite a losing overall lead change record, his 21-10 TNL lead change record is quite impressive, making him one of the most impressive closers from the perspective of an individual race, even if he is not necessarily one of the most impressive closers on a season-by-season level, much like Carl Edwards in NASCAR.

The most underrated driver by far appears to be Kenny Bräck. By all rights, he should have won way more than the nine races he actually did win. He placed himself in the lead far, far more often than he led, indicating that he was doing most of the work and making his teams look better than they were. His nine wins to 16 TNL is a huge, huge difference (that's even a greater ratio than Sterling Marlin's 10-17 jump), and his differences are just as absurd in the other categories (15 races with the most lead shares to 8 races leading the most laps, and 14.4 lead shares to 8.2 CRL, which is actually the largest absolute difference of anyone (not merely percentage difference), even the drivers with substantially more wins.) Some people may criticize Bräck for his participation in the early IRL, which did not have the greatest field depth. However, Bräck's stats actually improved in CART, despite facing some of that series's deepest fields ever (such as when he led the points for most of 2001, a season that may have had more good to great drivers competing regularly than any other season in CART history... sure, there were no Formula One champions full-time by that time, but there were also almost no bad drivers and the 1993-1994 seasons had several.) Bräck had 11 TNL and races with the most lead shares in CART to 5 wins and races leading the most laps, so his CART years actually made up most of the difference. He was an exceptionally strong duelist in either series and clutch as well with a 5-3 TNL record in his IRL years and a 6-1 TNL record in his CART years. He actually improved even relative to the greatly improved competition level in his CART years, and had comical bad luck, particularly in his 2002. Although he was hyped as an oval-only driver, he was actually TNL in several road course races where he won the pole, but this tends to be forgotten with regard to his career. He was much stronger than he is now regarded. He was the only driver to manage to give the Bobby Rahal team more than two wins in a season from 1992 to the present day, and the only driver to do so for A.J. Foyt in the IRL as well (despite being regarded as one of the premier IRL teams at the time, none of Foyt's other drivers came close to Bräck... his teammate Billy Boat's dueling record looks even worse than you'd expect.)

Other drivers who came out as significantly underrated by these statistics included Tony Kanaan (with 6 more TNL than wins and 6 more races with the most lead shares than races leading the most laps), Juan Pablo Montoya (with 5 more TNL than wins and the second best TNL dueling record of any driver with more than ten races led at 14-4), Gil de Ferran (who despite being a weak duelist still managed to have 18 more TNL than wins because he tended to win poles and lose the lead either due to pit stop exchanges or pure dumb luck such as his shredded tire with a few laps to go at Long Beach), Rick Mears (whose differences aren't too large, but still larger than you'd think given his conservative reputation; although I must say the negative lead change record is a bit of a surprise, but he was certainly clutch in the end), Justin Wilson (not surprisingly given the weakness of his equipment, he had four more TNL than wins and more races with the most lead shares than races with the most laps led; he stands out nearly as much as Bräck, particularly given he had even worse cars), Bryan Herta (whose TNL lead change record 1-7 was indeed perhaps the worst duelist among semi-major drivers, as evidenced by the fact that he is probably more remembered for being passed by Alex Zanardi at Laguna Seca than for any of his wins, but he had a lot more speed than many people remember, and quite frequently won the pole without being passed before he had some dumb bad luck afterward), and even Marco Andretti (who has twice as many TNL and wins, and his number of races led and lead change record are very, very similar to his highly-touted and overrated teammate Ryan Hunter-Reay... they weren't very different as duelists at all; Hunter-Reay was just substantially more clutch (10.4 to -20.1) and substantially luckier on top of that). One can quibble about the early IRL, but Buddy Lazier looks especially impressive here and I almost regret not listing him on my top 100 list. Despite only leading the most laps in a race once (which I think would shock everybody to see considering how dominant he was regarded in those years) he won eight races, scored eight TNLs, and had the most lead shares seven times. His 8-1 TNL lead change record was even better than Montoya's and behind only Adrian Fernández among drivers with ten or more lead changes. Obviously, Lazier wasn't quite facing the same competition though but I believe it says something that Hemelgarn Racing was such a big factor with him, even against the gutted early IRL fields, when it almost never was without him.

I have repeatedly argued over the years that Hélio Castroneves is the most overrated IndyCar driver of my lifetime (and I believe second-most all time behind only Barney Oldfield), and nothing I have observed since has caused me to change my mind. Castroneves had the second largest numeric difference between his lead shares and cumulative races led. Although his wins and TNL were close, he still had one more win, indicating he was lucky to win as much as he had. In fact, the only reason his wins and TNL are close at all is because Castroneves was extremely good at winning pole positions at road course races, giving him a grand total of fourteen races where he won the pole for races where there were no on-track lead changes, but unlike his fellow teammate Will Power who was also dominant in qualifying, he was definitely not good at sealing the deal through his own efforts. His 130-149 is definitely one of the weakest dueling records of any major driver, although his fellow Penske drivers Gil de Ferran, Danny Sullivan, and Simon Pagenaud were actually worse. However, no one who led that often had anywhere close to as bad of a TNL lead change record as Castroneves's 15-30. Despite having a negative lead change record to begin with, he was so unclutch that his lead change percentage at the end of races was 13% worse, and that is extremely rare. Only Bryan Herta (regarded as the worst duelist), Bruno Junqueira (who looks much worse than I expected by this metric), and Marco Andretti (which shouldn't surprise many people) had worse percentages. This also includes the 2013 spring Texas race as a pass for Castroneves against Marco. In that race (as in Sébastien Bourdais's 2015 Milwaukee win) both cars were found illegal afterward but did not lose any points, so I counted the passes as legitimate. Had I not counted that, Castroneves's TNL lead change record would actually be worse than Marco's. Coupling that with him not winning a championship while five of his teammates did and you can begin to see why I think he is by far and away the most overrated driver of the era. I was criticized for ranking him so lowly on my top 100 list, and now at this point I actually think I maybe should have gone even lower. He was tremendously advantaged by being the longest-tenured Penske driver in history, even though he was far, far from the best one despite eventually accruing more wins for Penske solely due to longevity than anyone except his teammate Power who has now matched him. Indy-centrists will argue that his three Indy wins are an argument for his greatness, but I do not see it. First of all, his Indy TNL lead change record at 1-3 matches the career trends and is actually even worse! (He passed Scott Dixon in 2009, but was passed by Gil de Ferran in 2003, Ryan Hunter-Reay in 2014, and Takuma Sato in 2017 for the win, and if you think Paul Tracy's pass in 2002 counted, it would be worse yet still.) On top of this, both his 2001 and 2002 wins were controversial. Although nobody ever talks about this, he was penalized for cutting off Tony Stewart in the pits on a caution-flag pit stop in 2001 and came out of the pits 2nd behind his teammate de Ferran. However, bizarrely, even though de Ferran committed no violations, Castroneves was allowed to restart in front of de Ferran but behind Stewart; this ultimately decided the race after Stewart pitted. I don't have any problems necessarily with people winning races unnaturally, but both his first two wins have fishy elements that don't exist in most other unnatural wins, yet they are used as the centerpiece of his legacy and in many people's minds justify him not winning any championships because Indy 500 wins are more important. Considering Castroneves was driving for a car owner in Roger Penske who has outright dropped drivers (Tom Sneva in 1978 and Al Unser in 1985) after championship seasons, and also fired Paul Tracy in 1997 after he led the points for most of the first half of the season, he was extremely, extremely lucky to somehow stay at Penske longer than anyone else. It's hard for me to understand.

In a way, Castroneves is like the rich man's equivalent of Ryan Newman in NASCAR. Yes, I think Castroneves is a substantially better IndyCar driver than Newman was a NASCAR driver. However, there are numerous similarities. Both of them drove for Penske simultaneously. Both of them were definitely among the best qualifiers of their time but far, far better qualifiers than racers. Both of them gained a reputation for being hard to pass because of their notorious blocking, but that exactly explains why they are overrated and why media narratives sometimes create very false impressions of drivers' careers. Generally if a driver was actually hard to pass, that driver would be consistently faster at all points in the race and would pull away. Such drivers wouldn't battle for the lead as long. They would merely take the lead quickly and pull away. Castroneves and Newman both gained reputations for being hard to pass due to their blocking but their blocking clearly indicates that they were slower than the drivers who were trying to pass them. If in the future, you see a driver being hyped as being "hard to pass", be aware what this likely means. Drivers like that usually are in fact the easiest to pass (think of how Castroneves managed to get outdueled by Danica Patrick at Motegi when both were stretching their fuel mileage, then consider if Dario Franchitti or Scott Dixon would have lost that race in the same position; I think not.) The actual hardest drivers to pass are always those who pull away from the field, because those are the drivers who don't give you a chance to repass them.

Now that Castroneves is no longer a full-time driver, the most overrated drivers on the current scene would appear to be Simon Pagenaud and Ryan Hunter-Reay. Pagenaud at 13-21 has an even worse lead change percentage than Castroneves, and his 3-6 TNL lead change record is even worse than that. Pagenaud, like Castroneves, is very fast in terms of speed and a much better driver than a racer, and Pagenaud especially has one major advantage over all other Penske drivers. He is clearly one of the least crash-prone drivers in IndyCar history with an extremely impressive record of avoiding crashes. Pagenaud has only crashed three times in 118 races (2.5%), which might be one of the greatest percentages in IndyCar history. His speed along with Penske's at times brilliant strategies can place him in the lead often and he is fast enough to stay there. He does not have to rely on blocking faster cars as Castroneves usually does, since he is one of the faster drivers and at this time probably faster than Castroneves ever was. However, Pagenaud is passed far too often in situations where it seems like he shouldn't be. Juan Pablo Montoya in 2016 passed him at St. Petersburg to take control of the race (although it was for second place before he passed Conor Daly, who stayed out of the pits, shortly thereafter) but it was later revealed that Montoya's steering was coming apart and he still managed to hold Pagenaud off. Josef Newgarden managed to pass him at Iowa on the start and nearly lap a field less than a month after cracking his sternum. At St. Petersburg in 2017, Pagenaud was passed again by Sébastien Bourdais in a Dale Coyne car that was no doubt considerably slower than Pagenaud's Penske. I can't really understand why he lost most of those races in which he seemed to be in the driver's seat, so to speak.

As for Ryan Hunter-Reay, he seems to have been especially lucky because he has 16 wins but only 11 TNL. As I said, it stands out that he and Marco Andretti have nearly identical records in terms of races led and career lead change records despite Hunter-Reay being generally considered as great and Andretti being generally considered weak. They have been teammates for eight years, most of their entire career (and particularly most of the portion of their career when they were frequently leading races), which says something rather mediocre about Hunter-Reay's career generally, and when you consider how many times he has been beaten by teammates who are not especially great or weren't at that time (like Jimmy Vasser in 2003, Mario Domínguez in 2004, Marco in 2013, and Carlos Muñoz and then-rookie Alexander Rossi in 2016, who I now think is better) it does seem to indicate that Hunter-Reay's brief domination in 2012 and early 2013 was something of a fluke.

Other drivers who appear to be overrated are Scott Dixon and Dario Franchitti (both slightly, as they frequently benefited from beating people out of the pits on green-flag pit cycles, which Dixon in particular was especially good at doing... one could give him credit for being the ultimate master of this strategy, but I think with regard to pit stop exchanges the team is generally more responsible, particularly because both Dixon and Franchitti gained a lot from this consistently in their heyday while the Penske drivers all tended to lose a lot. Al Unser, Jr. looks to be slightly overrated, probably because he was a major benefactor of Michael Andretti's bad luck, and/or Paul Tracy's frequent crashing. Sébastien Bourdais looks somewhat overrated since he has five more wins than TNL and more races leading the most laps than earning the most lead shares, but he has more lead shares in general than cumulative races led, which indicates he was still pulling his own weight. Bobby Rahal also has five more wins than TNL and seems to have been another major benefactor of Andretti misfortune, as Emerson Fittipaldi, whose IndyCar career looks almost as overrated as Castroneves, but it's fairly understandable since he was almost 40 years old when his IndyCar career started, and still raced almost to the age of 50. Likewise, Mario Andretti's lead change percentage looks shockingly weak, but bear in mind that this is starting from the point he was 42 years old, and it's likely he might still end up with a positive record if you consider his earlier years in the 1960s (if it were even possible to include these, which it might be since most of the oval races of those days did not have many lead changes.) Both Andretti and Fittipaldi are probably hurt by being in decline when their careers started. If you stop in the early to mid '80s for Andretti, his record is actually positive, and probably more like his onetime teammate Nigel Mansell, who raced at age 40 and 41 and came out as underrated, but likely would not have if he had continued as late as Andretti did. Alex Zanardi continues to look very overrated as he always has to me, but only slightly based on the passing statistics (mostly it's because of the absurd dominance of the Ganassi cars that made drivers who were very good but not great like Zanardi and Jimmy Vasser look like superstars; Zanardi was still a fantastic duelist as he is always hyped, but his cars clearly gave him an advantage as far as that goes.) Dan Wheldon, Cristiano da Matta, Bruno Junqueira, Tony Stewart's IndyCar career (but not - not his Cup career), Tomas Scheckter, and Scott Sharp look overrated too.

Two drivers appear overrated from these metrics but I would argue otherwise because they were such fantastic duelists. Josef Newgarden deserves a major shout-out because despite a relatively short career so far, he has the best lead change percentage of any driver with ten or more races led at 14-8 (63.64%.) He certainly benefits because he has improved from season to season and has never declined, while many of the listed drivers fell off from their earlier peak. However, I'm not sure people have noticed just how great a duelist Newgarden is. In all three seasons in which he has led a race naturally (2015-17) he led IndyCar in lead change percentage among drivers who led three or more races. No other driver has managed to be the top duelist in three consecutive seasons, and in the first two of those, he clearly had second-tier equipment for the Ed Carpenter operation (Carpenter himself even after a brief reign of semi-dominance on the superspeedways from 2011-14, Spencer Pigot, and J.R. Hildebrand didn't come close to matching him, or even contending, except that Hildebrand seemed to continue somewhat to a lesser degree with Newgarden's Iowa dominance, which makes some sense since he did set up the car where Newgarden annihilated the field at the 2016 Iowa race when Newgarden was injured.) In the IRL/IndyCar years only, no other driver besides Newgarden has even led in lead change percentage three or more times, let alone consecutively as Newgarden has done. Even if you exclude the 2015 Barber race, where I counted Newgarden's pass of Castroneves as a pass for the win even though it was a pass in midfield that became the pass for the lead after a bunch of other cars pitted, his 13-8 record would still be the best for a career with ten or more races, and his 3-2 record would still be the best for 2015. His 2017 in particular was awesome from a dueling perspective as he went 4-1, with an undefeated 4-0 lead change record against his Penske teammates (scoring passes against all three of them, some of which were extremely slick, such as his pass of Will Power at Mid-Ohio and his pass of Simon Pagenaud at Gateway) and only being passed by Scott Dixon, who had red tires on a restart at Road America while Newgarden had black tires, making that pass fairly unavoidable. Newgarden appears slightly overrated because he has more TNL than wins, more CRL than lead shares, and more races where he led the most laps than led in lead shares. He has been lucky, inheriting the lead at Barber in 2017 when Will Power cut a tire and inheriting the lead at both the 2015 and 2017 Toronto races when a caution came out after he pitted but before all the other contenders had pitted, which I admit is fairly cheap. Having said that, I think criticizing him or calling him overrated because he has to this point been lucky is wrong. It would be hard for him to be a better duelist than he already has been, and even if he has backed into some wins, he still seems to do more through his own efforts than just about anyone else now. In other words, lucky or not, I still think he deserved the championship. Adrián Fernández looks especially overrated on the surface because he won 11 races but only had 6 TNL and 5 races where he led the most laps or led in lead shares, along with more CRL (5.0) than lead shares (4.0), none of which even approach his win total. While Fernández specialized in fuel mileage races, largely because he had Jim McGee, one of the all-time great strategists in IndyCar history and the all-time winning crew chief who dated all the way back to Mario Andretti's peak in the late '60s, on his pit box in his Patrick Racing years, he still was an exceptional duelist in his own right, with a 21-13 lead change record and especially a 5-0 record on TNL lead changes. When he got in the lead in other words, he was passed far less often than most other drivers were, even though he did admittedly back into wins and leading a lot. Both Newgarden and Fernández appear overrated on the surface statistics, but when you consider their great dueling abilities, I would argue otherwise.

Although I will need to make many adjustments in terms of which passes I count and which passes I do not, it's clear that these data do have meaning in evaluating drivers, particularly those who seem to deserve more or less attention than their results or hype would indicate, and they generally did match my earlier perceptions, with some bizarrely inexplicable results (like Greg Ray actually coming out as extremely clutch because he was really good at being the last driver to take the lead in a race before crashing later). For the most part, there was a strong correlation between which drivers put themselves into the lead - and into the battle for the win - more than you would expect based on the level of their dominance, and those who did so rather less. The biggest revelation I think that has emerged from all these analyses is the idea that the drivers who are hyped as the hardest to pass (such as Ryan Newman and Hélio Castroneves) really are actually the easiest major drivers to pass because the hardest drivers to pass are those who pull away and therefore won't give you an opportunity to repass them. This definitely defies the conventional wisdom, but it is I think correct. The drivers hype drivers for being hard to pass based on how much they block or hold up traffic behind them, but the fact that they are blocking or holding up traffic by definition means that they aren't as fast as the drivers behind them, and eventually will be passed. Most everything else did match my earlier opinions, including just from my observations watching the races among the most recent drivers. I could already tell that while Simon Pagenaud may be a better driver than Josef Newgarden (because he crashes less, and Pagenaud does crash less than probably any other driver in the field), Newgarden is clearly by far the better racer. Would almost anyone else have managed to pass Pagenaud at Gateway when it was so impossible to pass? Would Newgarden have been able to pass someone in a race like that who was a positive duelist? The same can be said for Alex Zanardi's famous pass of Bryan Herta at Laguna Seca in 1996. Would that have happened if Herta was an average duelist rather than one of the weakest ones? Or if Zanardi had not been a quite good duelist? I think these results ultimately have meaning and do for the most part reflect people's perception of drivers based on what happened while actually watching the races. As I believe I've said earlier, I used to have trouble comprehending the difference between the best drivers and the best racers, but now I definitely get it. It's an open question which you value more - ability to battle on track or ability to be fast while avoiding mistakes. Likely both are important, but I tend to be on the side of the racers, as I think drivers who get better finishing results than their passing/racing abilities would indicate too often benefit from the strength of their teams. To be sure, passing ability comes down to equipment strength as well, but ultimately passing ability comes down only to the driver to the car, not the strength of the pit crew or the strength of the strategists, which makes it more valuable to me. I think analyses like these do a good job of canceling out team factors. A passing analysis that managed to adjust for team quality as well might do the best job.

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.

---