Head to Head IndyCar Teammate Records

by Sean Wrona

Over the past couple months while working on my previous article on NASCAR teammate head-to-head records, I was simultaneously working on the same analysis for the history of IndyCar racing. Just as in my previous column, I wanted to include results for the entirety of IndyCar history, not merely the CART/Champ Car/IRL/IndyCar period, although admittedly multi-car teams were fairly rare in the decades prior to the formation of CART. Although I only ever went back to 1979 on race-database.com, I used racing-reference.info's listed car owners to determine which drivers were teammates. This seems to be highly accurate for the post-World War II era, but prior to that there are a ton of car owners who seem to be unknown, especially in the '10s and '20s. I know that especially in this era engine manufacturers entered IndyCar races with their own factory teams that I believe were owned by the manufacturers themselves, such as Frontenac and Duesenberg. However, it also appears that there were independent privateers for those manufacturers simultaneously and I couldn't really tell which entries were factory-backed and which were privateers, so I decided not to. From this period I only included car owners that were clearly stated in parentheses after the entry name or sponsor on racing-reference, even though I know there are likely several teammates I missed in this period, although a few drivers from that era like Harry Hartz and Tommy Milton are represented on this list. For the most part though this list primarily includes drivers who raced in the 1980s or later, much like the NASCAR list.

I actually finished determining the teammate records several weeks ago but did not write the summary until today. I completed calculating the recent years after the Mid-Ohio race and finished counting the entire data set after the Pocono race, so I've been sitting on this for a while. Over that time, there was one significant change in the results. Josef Newgarden had a 18-17 race lead over Will Power, but with Power's wins at Pocono and Portland, he has overtaken Newgarden to take a 19-18 lead in the results. Just as before, I required both teammates to complete five or more shared races for the same car owner. However, since IndyCar seasons are usually about half as long as NASCAR seasons and there was much higher attrition in IndyCar races historically, the effect is that relatively few teammates prior to the CART era appear on the list.

Only eleven drivers in IndyCar history had multiple teammates and were never beaten in a head-to-head teammate matchup. These are Sébastien Bourdais, Kenny Bräck, Robbie Buhl, Dan Clarke, Arie Luyendyk, Darren Manning, Greg Moore, Andrew Ranger, Alexander Rossi, Scott Sharp, Bobby Unser, and Rodger Ward. Most of the biggest names are surprisingly not on this list. Almost every driver who had a lot of teammates was at some point beaten by another, while the biggest legends, who tended to come from earlier decades had very few eligible teammates. For instance, A.J. Foyt only had one (George Snider) but he did have a perfect record against him. As a result, some of the names on this list are pretty surprising. The first thing I note is that all these drivers except for four (Luyendyk, Rossi, Unser, and Ward) had their best seasons during the CART/IRL split thereby suggesting that most of these results came from competing against relatively shallow competition. Bourdais, Clarke, and Ranger competed in the shallow late Champ Car period, while Buhl and Sharp primarily competed in the weak early IRL period and Luyendyk also crossed over then even if his best seasons (and the seasons where he most frequently had teammates) were in early '90s CART.

Among those drivers, only Bräck and Moore actually competed against strong teammates in the highly competitive CART seasons. Bräck dominated Max Papis 15-5 and also beat Bruno Junqueira and Scott Dixon 4-3. This urprised me because Junqueira had a much better 2002 season than Bräck, but that was mostly beacuse Br¨ack had so many crash DNFs. In the races both finished, they were much closer. Greg Moore might look the most impressive of the lot as he was the only teammate to beat Patrick Carpentier (even Paul Tracy merely tied him at 10-10 despite winning his championship in that period) and gave Tony Kanaan his worst defeat in his career, even though Moore had the inferior Mercedes engine and Kanaan had the superior Honda engine that year. Because Moore and Bräck's careers were ended by crashes they never really went into a period of decline, which explains why they were never beaten by a teammate. Bourdais is the only one of these drivers who had teammates for the entirety of a lengthy career and never lost to any of them, but he also had much worse teammates than any other champion did. Only three of his teammates - Bruno Junqueira, Graham Rahal, and Oriol Servià were particularly good, while his teammates in post-split IndyCar have for the most part been pretty weak, which makes him harder to evaluate than a lot of other drivers. The 11-0 sweep of Rahal in 2007, when Bourdais won the championship and Rahal was a rookie looks particularly awesome though when you consider Rahal finished 5th in points, beat fellow rookie Simon Pagenaud in the championship, and beat Bourdais's runnerup Justin Wilson in the very next season.

The biggest surprise and revelation here is probably Scott Sharp. Although his teammates were generally not great (four of his teammates combined for only one win) he still beat all these teammates including Mark Dismore, the one winner amongst them, by more than a 2-1 margin, and he was consistently the leader of his teams for over an entire decade from 1994-2005. Additionally, he was the only teammate to ever beat Al Unser, Jr. in his career. Now Unser had clearly significantly declined from his '90s peak without question (he had his DUI arrest while Sharp was his teammate) but actually Unser's early alcoholic period looked a lot better than I was expecting. Although Paul Tracy did beat him in 1997 it was closer than it looked and not by enough to override how badly Unser beat Tracy in 1994 and 1996. Then Unser beat his next two teammates André Ribeiro and Didier André by large 6-1 and 6-0 margins and André was his teammate the previous year, so much to my surprise I don't think I can really argue Unser was washed up then. Sharp was a lot better than I thought. This shouldn't be that surprising I guess since he was one of the few original IRL drivers who had real cred (two Trans-Am titles and a 24 Hours of Daytona win before his IRL co-championship season was even over) and he was also the first IRL driver to win in seven consecutive seasons and the only original IRL driver to survive and win after the CART drivers crossed over full-time.

The other undefeated drivers are probably more surprising. Darren Manning is the most notable of these given that he actually beat Scott Dixon 7-4 in 2004-05 at Ganassi in their lame duck Toyota years when their cars were not capable of winning. Despite outperforming Dixon, he was fired mid-season, making him one of the biggest what could have been drivers of the decade. He also beat Rodolfo Lavin 10-0, a greater margin than Paul Tracy did, and he even did better in the A.J. Foyt car in 2007-08 than Tracy and Ryan Hunter-Reay did the next year. Was Manning a lost potential champion? Based on all that, I think he's probably one of the best IndyCar drivers to never win a race at any rate. Robbie Buhl was Tony Stewart's teammate in the IRL in 1997-98 and for all Stewart's dominance to the point of leading the most laps in seven consecutive races, Buhl actually tied Stewart 3-3. That's probably just due to the extremely small sample size but it does make me reconsider how great a formula open wheel driver Stewart actually was since Buhl is not as well-remembered today. Although Dan Clarke garnered the nickname "Speedy Dan", most fans thought this was kind of a joke comparable to how Joe Nemechek got the nickname "Front Row Joe." However, maybe there was something to that. I was very surprised to observe that he ended up actually beating both Nelson Philippe and Robert Doornbos in years those drivers won races and finished in the top five in the points. When you consider Clarke had a mediocre reputation despite his nickname, never won a race, and never got a top ten points finish, it makes one wonder if he was actually a lost talent. Having said that, that was a very weak era in Champ Car and that was proven in the united IndyCar when both Philippe and especially Doornbos badly struggled compared to their Champ Car results. Andrew Ranger, who was the youngest driver to score a podium in Champ Car history, also beat Philippe as well as his other teammate Charles Zwolsman, Jr. I wonder if Clarke and Ranger might have been able to do anything in the merged IndyCar if they hadn't been lost in the shuffle, but I rather doubt it.

Among current drivers, Alexander Rossi is the only driver besides Bourdais to score a winning head-to-head against all his teammates. This isn't a big surprise as he has pretty much been unambiguously the team leader at Andretti. Since the start of the 2017 season, and he only finished four points behind Carlos Muñoz and two behind Ryan Hunter-Reay even in his rookie season (they were the only teammates to ever beat him in the championship and then only barely.) Considering that, I'm actually surprised his records aren't even stronger. Although he finished behind Muñoz in the championship, he did beat him in finishes 8-7, although that would have been reversed had Muñoz pitted only two laps earlier in the Indy 500. Considering Hunter-Reay has been in decline in recent years, I expected Rossi to be leading that one by more too. And although he did manage to beat Takuma Sato by a 2-1 ratio, Graham Rahal is actually (and very surprisingly) beating Sato by a substantially greater margin right now. Although Rossi is certainly one of the best drivers in the series, I expected more.

Two currently active champions beat every one of their teammates for over a decade, with one of them only being beaten by the other. Nobody should be surprised that the first of these is Scott Dixon. Although he lost to Bruno Junqueira and Kenny Bräck in 2002 and surprisingly lost to Darren Manning in 2004-05, he has beaten every teammate he's ever had in the fourteen years since, which is nearly as long a streak as Kevin Harvick's in NASCAR. Besides Bräck, Dixon beat all three of his championship-winning teammates: Tony Kanaan, Dan Wheldon, and even Dario Franchitti. While Franchitti might have stolen in the spotlight in that era with three championships to Dixon's one and two Indy 500s to Dixon's zero, he didn't outperform him by nearly as much as that implies and may not have even outperformed him at all. They were essentially equal in 2009 and only a fuel-mileage season finale decided that title in Franchitti's favor. Although Franchitti did outperform Dixon in 2010-11 overall (but only by one race by this metric), after the DW12 chassis replaced the IR03, Dixon outperformed Franchitti in 2012-13 by a greater margin (13-9) giving him the edge in the matchup, despite Franchitti seeming to have greater objective stats on the surface. Dixon's 30-19 edge over Kanaan may be even better when you consider Dixon was the only teammate who beat Kanaan in this millennium. Even Dixon himself lost to more teammates over that same period.

Kanaan was perhaps the biggest revelation in all these data. Although he seems to have the reputation of a second-tier legend now primarily because he wasn't as strong on road courses as on ovals, he seems to have been far better than people acknowledge these days. Not only did he beat seventeen different teammates, more than any other driver, while losing to only three, he also was one of only two drivers along with Al Unser, Jr. to beat four different championship-winning teammates: Dario Franchitti, Ryan Hunter-Reay, Dan Wheldon, and Alex Zanardi. He also beat Rubens Barrichello in his IndyCar season as well, while all of Barrichello's Formula One teammates except Michael Schumacher and Jenson Button did not. Kanaan's domination of Franchitti and Zanardi was particularly staggering. His 34-14 record over Franchitti is astounding when you consider that Franchitti won four titles in a row (minus his 2008 gap year in NASCAR) immediately afterward, including one of them when he and Kanaan were Andretti Green Racing teammates. Even in that year when Franchitti won the title in 2007, Kanaan won more races and tied him in the head-to-head. What's amazing is that this result should arguably be even more extreme than that. Kanaan let off the gas on the last lap of the 2005 Fontana race to let Franchitti win, utterly dominated the 2007 Indy 500 before pitting immediately before rain ended the race handing Franchitti the win, and decided to sacrifice his championship bid in 2007 to block for Franchitti at Sonoma when he had a damaged wing. If Kanaan had finished higher in those three races, which most would argue he should have, the result would be even more extreme: 37-11.

Kanaan's record against Alex Zanardi was even more extreme: 5-1. One of the main reasons I ranked Franchitti and Zanardi lower than most fans would expect is because of how badly Kanaan beat them (and even I didn't realize Kanaan beat Franchitti that badly.) At the time, I myself was following the traditional view of Kanaan as a second-tier oval specialist and considered Franchitti and Zanardi to be massively overrated as a result for losing to a second-tier oval specialist, but I think I was wrong. For one thing, Kanaan actually beat Franchitti and Zanardi on the road/street courses in those years. Kanaan won two road/street course races from 2005-2007 while Franchitti won none, and Kanaan had the edge on them, outfinishing Franchitti 6-4, although that was certainly a lot closer than their oval record. He also beat Zanardi 2-1, which was again closer than their overall mark of 5-1 with Kanaan beating Zanardi in all three oval races. Both Franchitti and Zanardi were dominant road course racers in their Ganassi years and perhaps the only reason Kanaan wasn't was because there simply weren't many road course races in his peak years. He was offered the Ganassi #10 car in 2009 but turned it down and then Franchitti won three consecutive titles in it. Would Kanaan have done the same? I kind of doubt it but it does seem like a possibility. Even if he wasn't the greatest road racer, he's definitely one of the best oval drivers in IndyCar history and has a case for being the best ever. It is not a case I think that I would make, but he does have the most races led and most natural races led in Indy 500 history along with the most races led naturally on superspeedways. At the very least, he is the best superspeedway driver of his generation and I'm ashamed of myself for only rating him 59th before. He now seems like he should be in the top 20.

One of the things I've noticed is how unpredictable a lot of these results are. Often they have little to do with drivers' finishes in the points standings. For one example, Adrian Fernández beat Roberto Moreno by one position in the CART points standings in 2000 before leaving Patrick Racing to start his own team. The next year, Jimmy Vasser replaced him and also beat Moreno by a single position in the points. Fernández beat Moreno 2-1 in wins in 2000 while Moreno beat Vasser 1-0 in 2001. Based on that along with Fernández's greater win total and the fact that he was able to win as an owner-driver while Vasser seemed to need great equipment to win, my instinct told me that Fernández was better but looking at these head-to-head data it seems like I was clearly wrong there. I rated Vasser so lowly because I rated Zanardi relatively lowly and Zanardi dominated him (even outfinishing him 8-2 in Vasser's championship season!) However, particularly in extremely attrition-heavy and competitive eras like that one, the points standings were pretty inaccurate in reflecting how people actually ran in the races, most notably in the 2000 CART season when Juan Pablo Montoya clearly outran everyone else on a race-by-race level but finished 8th in the championship because he had 12 DNFs, which were almost all mechanical. Since I very rarely consider mechanical failures to be the driver's fault, in a case like Montoya's 2000 or Michael Andretti's 1992, I don't really hold the lack of championships against them. These sort of things happened to a degree in NASCAR too, like when Jeff Gordon actually outfinished Terry Labonte in 1996 and Tony Stewart did the same against Bobby Labonte in 2000, but it could often be a lot more random in CART due to the greater attrition and also due to the fact that the entire field did not score points. Hence, despite Fernández and Vasser looking similar at first glance, Vasser had a way stronger teammate record. He beat Moreno by a 6-1 margin (the same exact margin Michael Andretti had against Moreno) despite running midpack, while Fernández tied Moreno at 6-6. Additionally, Fernández also tied Scott Pruett and lost to André Ribeiro, who immediately went to Penske after that and had arguably the worst IndyCar season there ever, losing to late-period Al Unser, Jr. 6-1. Despite winning eleven races, Fernández tied two drivers who only won twice and lost to a three-time winner even in his peak heyday. Clearly he was one of the most overrated drivers at the time, and that does make sense. Fernández only had about 5 cumulative races led making the 11 wins seem like an anomaly. He also had Jim McGee, the winningest crew chiefchief mechanic in IndyCar history (whose success went all the way back to Mario Andretti in the 1960s!), which was probably the main reason for his success. Despite how often Fernández was winning I was watching enough CART races back then to know he was never really hyped as one of the biggest stars and most people have now entirely forgotten that he actually finished second in points once.

Vasser was sort of a revelation in the other direction. While he lost to Juan Pablo Montoya and Alex Zanardi badly, he beat all his other teammates by a 1.5-1 margin or greater and four out of the five drivers did have multiple wins even if few would call Bryan Herta, Michel Jourdain, or Roberto Moreno great. He was also the only driver to beat 2002 CART champion Cristiano da Matta, and he beat him 4-2 upon da Matta's return from F1, where he didn't do badly as he beat Olivier Panis before he was fired. Vasser seems to be similar to Heinz-Harald Frentzen or Jamie McMurray in that he massively underachieved in championship-caliber equipment but overachieved in mid-pack cars. Vasser is a rare champion whose best years may have been in the years he was no longer winning. Since I now appear to have underrated him that also means I likely underrated Zanardi, who dominated him, and also Kanaan, who dominated Zanardi. It seems like I might have been unfair to all of them mainly because Zanardi struggled in F1 and lost badly to Kanaan in CART.

There are many other examples of instances when the points standings did not match how drivers actually ran in the races. One of the reasons I never rated Scott Sharp as highly as I should have was because in his only full CART season in 1994 he got beaten by his teammate Dominic Dobson who was considered one of the weaker drivers of that era. However, looking at it this way, Sharp beat Dobson 6-1 while he was a rookie and Dobson was a veteran, so it was actually impressive. Dobson only finished higher in points because he got a 3rd place finish due to attrition at the Michigan 500 but Sharp regularly outran him when they were both running mid-pack and not scoring points. A similar situation came with the former F1 veterans turned PacWest Racing teammates Mauricio Gugelmin and Mark Blundell. Although Gugelmin beat Blundell every single year in the championship from 1996-2000 Blundell actually won the head-to-head 18-16, which was similarly striking. Their win totals were consistent with this as Blundell won three times while Gugelmin only won once in 1997, which is the only year they really had equipment capable of winning, but Gugelmin finished higher in the points that year too. Attrition made a lot of these results come out rather weird.

There is even a situation right now that is very similar to this. After winning three races over the past two years and getting a 3rd place finish at the Indy 500 while his teammate Graham Rahal has gone winless, conventional wisdom and the media have been saying Sato is one of the hottest drivers on the circuit while Rahal is slumping, but looking more closely at the data here that doesn't seem to be true at all. Since they became teammates in 2018, Rahal has been dominating Sato and has a 15-4 edge in head-to-head finishes. This is a substantially greater advantage than even Alexander Rossi's 8-4 edge against Sato in 2017. Rossi is now widely considered the best driver in the field. I'm definitely not saying Rahal is better than Rossi today (for one thing Rossi is certainly a lot better in 2019 than he was in 2017) and it's probably hard to argue Rahal is elite because he hasn't been winning or fighting for wins in a while when his teammate has. However, this result is wildly unexpected nonetheless when considering the win difference and the media narrative. It's worth pointing out that two of Sato's three wins came in flukish strategy races where he won solely because he pitted (at Portland) or didn't pit (at Gateway) before a caution came out. While it's impressive Sato called the Portland strategy himself he still was running nowhere near the lead in those races before winning on strategy. It says nothing about how he ran. He did impressively win from the pole at Barber, but aside from that he's actually been declining. Despite beating Rahal in points this year solely because he won two races and finished 3rd in the double points Indy 500 while Rahal DNFed in all three of those races, Rahal has actually beaten him 7-1 in shared finishes this year. The results are clearly skewed by the fact that Rahal did not finish any of the races Sato won, but even counting those Rahal would still have over a 2-1 ratio over Sato, which is definitely not what anyone would expect. Based on that, it seems clear that Rahal is the most underrated driver in the field today and Sato the most overrated.

Rahal's actually had a very underrated career in general. He has beaten six of his ten teammates and only lost to four, three of which were Sébastien Bourdais, Scott Dixon, and Dario Franchitti in their championship-caliber periods. A lot of people marked Rahal down severely for his inexplicable 2013-2014 when he did not come close to matching what Sato did in 2012 and surprisingly lost to the ride-buyer James Jakes, but the rest of his career looks quite all right. In addition to his stellar record against Sato, he had a very strong record against most of the other late-period Champ Car drivers. Despite getting shut out by Bourdais as a rookie in 2003, Rahal went on to thrive substantially more in the merged IndyCar as he beat the 2nd and 3rd place finishers from that year behind Bourdais, Justin Wilson and Robert Doornbos, the next two years and he also beat Oriol Servià 6-1 over multiple different stints as teammates. When you consider that Servià beat Will Power, tied Paul Tracy, and came closer to Bourdais than any other teammate, that's quite good. Rahal also beat 2008 IRL Rookie of the Year Hideki Mutoh 5-1, which was about the same ratio that Tony Kanaan did immediately before. Rahal may have never quite been elite or great but it seems apart from his 2013-14 period, he's probably been a top ten driver almost every other season in his career, and his sudden improvement in 2015 shouldn't have been as surprising as it seemed at the time.

It's harder to evaluate the drivers prior to the CART/IRL split as most teams other than Penske did not have multiple cars most of the time in those years. However, many legends did drive for Penske so some very real comparisons can be made there. Probably the driver most closely linked to Penske in IndyCar racing was Rick Mears. One of the biggest surprises of all in this place is that Mears looks more like a second-tier legend than an all-time great. Mears certainly had an auspicious and legendary debut season for Penske in 1978. He was hired to drive Penske's second car part-time alongside the defending champion Tom Sneva. Mears entered the races when Mario Andretti was not available due to his Formula One commitments. Even though Sneva defended his championship and Andretti won the Formula 1 title, Mears actually outperformed both in IndyCar that year. Mears swept Sneva at 5-0 in shared finishes while both Mears and Sneva beat Andretti. Mears also won 3 times to Andretti's 1 and Sneva's 0. It's no wonder Mears was immediately elevated to team leader for 1979. However, for the rest of his career, Mears was good but not as overwhelming as you might expect. Mears's teammate from 1979-81 was Bobby Unser. They tied at 13-13, which doesn't sound bad until you realize that this was Mears at his absolute career peak while Unser was 45-47 years old and probably thereby not at his peak. After Unser retired in 1981, he was replaced by Kevin Cogan, who Mears did sweep 5-0 in 1982, which might have been his best season. However, Cogan is frequently cited as Penske's worst full-time driver so it's hard to tell how impressive that is. To be fair, Mears beat Cogan by a far larger margin than Emerson Fittipaldi did later. Cogan was replaced by Al Unser from 1983-85 and Unser instantly won two out of three championships while Mears wasn't winning as much (even before his crash at Sanair, Unser had a winning head-to-head record against Mears, but Mears slightly improved that to 9-8 even though Unser won the championship in 1985.) Considering both Unsers were in their mid-40s while Mears was in his prime, it seems they were both greatly superior, even though many try to rate Mears in the same league because of his Indy record, but even there he has fewer laps led than Michael Andretti for example. Mears did end up beating Al Unser by quite a bit because Al continued to do one-offs for Penske for much of the tail end of his career, but when they competed full-time against each other it was pretty even and Unser was stronger in the championships. Fittipaldi and Danny Sullivan both also beat Mears in his later years at Penske. I think these results reveal Mears to be rather overrated, but his 1978, 1981, and 1982 are still great though.

Of the all-time legends of the '60s and '70s Mario Andretti seemed to have the overall best record. Although he lost to Sneva while he was making IndyCar one-offs as a full-time F1 driver, he was a lot more effective as a full-time IndyCar driver, beating three champions by very large margins (6-2 against Gordon Johncock, 6-3 against Joe Leonard, and an astounding 7-1 against Al Unser.) The three-car Parnelli ensemble of Andretti, Leonard, and Unser was one of the rare instances prior to Team Penske where every driver on a full-time multi-car team won a championship but much to my surprise (since Unser and Leonard won the championships for them from 1970-72) Andretti came out on top, although Unser did beat Leonard by a larger magin than Andretti did. Clearly Andretti was the fastest of the three drivers but was just unlucky in terms of finishing races, and he was also unlucky that he signed with Parnelli as they were fading from their 1970-71 peak.

None of the other USAC-era drivers dominated championship-caliber teammates in their prime to such a degree as this. Of course A.J. Foyt never had any championship-caliber teammates for more than a race or two throughout his heyday. Al Unser only lost to Andretti and Mears but beat Danny Sullivan even though Unser was in his mid-40s and Sullivan was at his peak. It says a lot for Unser that he was basically at least matching Mears and Sullivan in his very late career. Most of the other legends of this period had few major teammates. Bobby Unser had only Mears, Gordon Johncock had only Andretti, and Johnny Rutherford had no full-time teammates at all but this definitely does make Mario Andretti's '70s in IndyCar look a lot better than I thought it would.

As for the major legends who debuted in the pre-split CART period, I expected Michael Andretti to come out the best because he only finished behind a full-season teammate in the championship once ever in his career (Dario Franchitti in 2002) but he didn't look nearly as impressive doing teammate comparisons this way, and Al Unser, Jr. proved to be the standout from that generation instead. Michael had only three teammates who could be remotely considered legendary. While he did beat his father Mario by a staggering 30-3 margin from 1989-92, Mario was also 49-52 years old in those years (although it's impressive that Michael did outperform Nigel Mansell's record against Mario, which was only 11-3; when you consider that Mansell was an F1 champion and Andretti was a bust there.) Michael's other two major teammates were Franchitti and Paul Tracy at Team Green and he ended up losing to both of them narrowly even though Andretti beat Tracy all three years they were teammates in the championship (including 1995 for Newman-Haas as well as the Team Green years.) By comparison, Unser defeated four championship-winning teammates including Emerson Fittipaldi, Bobby Rahal, Danny Sullivan, and Paul Tracy (and Unser gave all of them except Sullivan their worst defeats ever.) Trying to decide between Andretti and Unser based on shared teammates is pretty murky. While Andretti lost to Tracy and Unser beat him, Andretti also beat Mauricio Gugelmin 8-0 at Ganassi in 1994 before Gugelmin went on to dominate Danny Sullivan 6-1 the very next year at PacWest, an even worse defeat than Unser himself gave to Sullivan. Evidence could be argued for both sides in that debate. Bobby Rahal is harder to compare because in most of his peak seasons in the '80s and early '90s he didn't have teammates. Rahal's only major teammate was Unser but he only lost to him 11-10 at Galles Racing in 1990-91. Bizarrely, Unser beat Rahal 9-2 in 1990 in a year he won the championship and Rahal went winless, but the next year Rahal turned the tables and beat Unser 8-2 in finishes despite Unser winning 2 races to Rahal's 1. I'd probably say based on their head-to-heads in wins that Unser was better, but Rahal still won more races than the other two when you start at the age Rahal was a rookie, and I don't think his teams were as dominant either, so it seems likely they are all in the same tier. The other CART champion I did not address was Jacques Villeneuve, but he never had any full-season teammates in CART so his F1 career will be much more illustrative there.

Among the more recent generation of drivers, Dario Franchitti certainly came out fine except for his shocking defeat by Tony Kanaan (he also lost to Dan Wheldon fairly badly but came very close to Scott Dixon, and beat all other teammates.) Paul Tracy is pretty inexplicable: although he lost to Dario Franchitti in 4 of 5 years he only lost to him by one total race, and he beat Michael Andretti here despite finishing lower all three years in the championship. Even weirder is that Tracy was basically neck-and-neck with Franchitti and Andretti but also tied Patrick Carpentier and Oriol Servià, who are generally considered more mediocre. It's bizarre for him to have roughly .500 records against both legends and slightly above average talents. Clearly it seems Tracy's record is badly skewed by his crashing: he could run as well as anyone but it made his finishing record so unpredictable even in races he finished that it's hard to draw any conclusions. Will Power was pretty weak here in his earlier seasons but has beaten every Penske teammate (although Josef Newgarden was ahead of him a few races ago.) Hélio Castroneves lost to every single Penske teammate in his career except Juan Pablo Montoya, whom he beat only narrowly (the only other teammate Montoya lost to Will Power.) Simon Pagenaud has lost to every Penske teammate except Castroneves. Along with those two, Ryan Hunter-Reay stands out as not especially impressive, especially with regard to his early-career losses to Mario Dominguez and Timo Glock (who was a rookie while Hunter-Reay was in his third season), but RHR did beat all his Andretti teammates except for Kanaan and Rossi (admittedly, one can make the case that those were his only teammates who were very good, so his record may not be very impressive at all.) Dan Wheldon was very evenly matched with Tony Kanaan at both of their absolute peaks and although Kanaan edged him out narrowly by two races Wheldon won more races and also came closer to matching Scott Dixon at Ganassi. For the most part, the modern drivers tended to do largely what you'd expect since there are a lot fewer mechanical DNFs nowadays than they used to be so the championship results are more predictive of which driver outran the other than they used to be, but you still get anomalies like Rahal and Sato where the conventional wisdom is totally wrong. These sort of things generally don't hold for the teammates in championship-caliber equipment though for the most part.

Overall I found the results of this analysis to be a lot more interesting and a lot more surprising than on the NASCAR list. For one thing multi-car teams in IndyCar were a bit more common in the '70s and '80s than they were in NASCAR, even though I'd say they were actually a bit less common in the '50s and '60s for the most part. However, the main reason is simply that IndyCar races until very recently had extreme attrition that can totally mask how well each driver is running, and that's not even to mention the pit strategy roulettes that have become particularly common in recent IndyCar that you don't find as much in NASCAR. You'd almost do better looking at qualifying speeds to determine who had pace, and I know that's considered especially important in Formula One, but I still prefer to look at finishes. Looking at qualifying clearly underrates drivers who were far better in the races than they were in qualifying (such as Alain Prost in Formula One, Al Unser, Jr. in IndyCar, and Dale Earnhardt in NASCAR) but it might still be better than looking at the championships. When you consider how big a factor attrition played in individual races (especially in decades prior to the '80s when IndyCar races in general were all weighted according to their length and the Indy 500 was worth substantially more than it even is today) the championship can actually give you a really bad idea how everyone was running, like how Mario Andretti actually dominated Al Unser and Joe Leonard on the multi-car Parnelli powerhouse in the '70s when you wouldn't know that if you looked at the results. I think this sort of analysis (particularly in attrition-heavy eras like the '70s and '90s) provides a needed additional perspective to complement both the advanced statistics I've invented basic finishing results (which are heavily luck-dependent and the idea that drivers have impact on their mechanical DNFs is largely disproven.) At some point in the next couple months I will proceed by doing this for Formula One as well (I've already done a little of this - would you believe Juan Pablo Montoya actually outfinished Kimi Raikkönen in their shared finishes? It shocked me too...) After that, I will eventually move on to a new subjective method to evaluate individual seasons, discuss a methodology for a book ranking the top 1000 drivers in motorsports history that I've mentioned several times here before, and likely do more historical year-by-year leader statistics for CART, USAC, and Formula One. However, at the moment I'm primarily focused on the book I'm writing about the history of competitive typing so I'll probably be taking these columns slow, but I'll keep them coming.

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.

---