Racermetrics race-database.com

Season Scores and Weighted Career Scores

by Sean Wrona

Last month, all three of my domain names (Racermetrics, race-database, and seanwrona.com) were all deleted by hackers and it took a while for me to restore everything, but now I am back. While I was away, I was working on many ideas for future columns I will write over the next few months, although I admit with a day job working in a TTY call center for the deaf and hard of hearing, it is very unlikely that I will ever be as ambitious as I originally intended. In this column, I introduce a new method for evaluating season performance and career performance at any moment in time, which has numerous applications that can allow me to more deeply analyze many aspects of auto racing in any series that provides the necessary data. These ideas are a bit more simplified than my earlier statistics but they are much easier to calculate and would likely form the basis for my ideas for any future World Driver Rankings lists I may compile.

There are three main schools of thought in what is most important in evaluating auto racing talent. One of these is that winning races and maximizing the ability to win an individual race is the only thing that matters. Formula One for most of its history primarily awarded winning and nothing else but has steadily been moving to awarding consistency more and more like CART did for most of its history. Another is that consistency is what matters most - finishing well week in and week out should matter the most regardless of whether certain drivers win or dominate more - maximizing the ability to win a championship matters more than maximizing the ability to win an individual race. Most point systems NASCAR has had have primarily rewarded consistency over winning. A third view is that dominance is what truly matters, regardless of the actual finish. Many would argue that the drivers who are most feared are the usually the drivers who consistently dominate regardless of who consistently finishes. As more and more racing series (primarily in the US but also around the world to some extent) are moving in a direction of equalizing competition and as reliability also increases in almost every series, races become crapshoots to a much greater degree. For instance, in last weekend's IndyCar race at Mid-Ohio, there was extremely minimal passing for position meaning that the two cautions during green-flag pit cycles shuffled the field around, and Graham Rahal's win and the other top finishers' strong finishes had little do with how they were running in the race or how well any of the drivers did themselves - it was all pit shuffling. Since the driver is merely one of many factors in whether a team gets a top finish or not - pit strategy and the quality of equipment arguably mattering much more - there is a strong argument to be made that dominance matters more than the final result, in the sense that it is the best prediction of FUTURE success, especially when more and more races seem to be decided by non-driver factors as more series move towards a crapshoot direction for entertainment reasons, but again, I am primarily thinking of NASCAR and IndyCar here, but not exclusively. While I won't necessarily argue with you if you believe that winning and/or consistent finishes are the only thing that matters, I personally think dominating races and leading laps is the best prediction for future strong performance in the long run. There are obviously many other things that matter, but some of them like average speed have not been adequately reported or archived for most series for a very long period of time, making historical speed comparisons extremely difficult to make (and some series still do not make their speed data public), and there are other things like versatility, equipment strength, how drivers took the lead (on-track or off-track), that certainly matter also, but calculating these can be rather nebulous or arbitrary guesswork to a degree as for instance precisely measuring a team's caliber of equipment is especially difficult when speed data hardly exists for most series prior to the 2000s.

One of my principal goals for Racermetrics was to come up with a simplified scale for measuring performance throughout a season. My personal perspective is that a race is a set of laps (data points) where the last lap is easily the most important but hardly the only thing that matters when there are too many things that can go wrong in a race that have nothing to do with anything the driver (or team) did wrong, so therefore in the long run, I think dominance wins out, but the long run is likely longer than a single season. However, I simultaneously acknowledge that winning and consistency DO matter, and think that the stereotypical F1 fan who thinks winning is the only thing that matters regardless of how the races are won, or the stereotypical NASCAR fan who thinks consistent finishes are the only thing that matters regardless of whether the driver was competitive in the race at all (although I don't know many people who actually follow the stereotypes of their preferred series' point systems) are both a little misguided. Objectively, a method for evaluating racing talent should consider winning, consistency, and dominance equally. Since I'm not sure there's an objective answer to this question although we will all have our preferences, this is what I have done.

Drivers' season scores are a composite of the sum of three categories (winning, consistency, and dominance) with each category judged on a 0-5 scale for each driver for each season, hence the overall scores will range from 0-15, with 15 representing a 'perfect season'. Just like with the NFL quarterback ratings or the NASCAR driver ratings, 15 does not actually mean perfect. You don't need to win every race in a season to get a 15 score. All you need to do is demonstrate what I consider to be historically elite performance in all three categories.

In the winning category I award 5 points for a winning percentage of 20% or more, 4 points for a winning percentage of 12-19%, 3 points for a winning percentage of 8-11%, 2 points for a winning percentage of 4-7%, and 1 point for a winning percentage of 2-3% (which will cover almost every winner in a series unless you have a *really* long season, which would only be typical of something like the World of Outlaws; there aren't many series that have 50 or more races in a year). I award 0 points to a driver with a lower winning percentage than 2%, but for all intents and purposes, a 0 score in this category will only include non-winning drivers.

In the domination category I award 5 points for an average percent led of 15% or higher, 4 points for an average percent led of 10-14%, 3 points for 5-9%, 2 points for 3-4%, and 1 point for 2-3%. I award 0 points for an average percent led of less than 2% a race because I consider it far too easy to lead a race or parts of several races solely due to pit strategy or other factors that had nothing to do with the driver, especially in series such as NASCAR and IndyCar that award points for even leading caution laps. Average percent led is my own statistic I introduced on race-database many years ago which equalizes dominance regardless of the number of laps in the race (hence, in NASCAR, leading 45 laps of a 90 lap Watkins Glen race, 94 laps of a 188 lap Talladega race, and 250 laps of a 500 lap Bristol race would all count the same). I don't doubt other people compiled this statistic as well privately, but I believe I was the first to do so publicly.

In the percent beat category I awarded 5 points for a percent beat > 80%, 4 points for 70-79%, 3 points for 60-69%, 2 points for 50-59%, and 1 point for 40-49%. Percent beat is another statistic I invented (which I have definitely seen nowhere else) which corrects for the bias in the average finish statistic from races historically having a different number of cars entered per race. To calculate this statistic, all you need to do is divide the total number of cars a driver beat throughout a season by the total number of cars that driver competed against (i.e. the number of drivers who competed throughout a season combined minus that driver's number of starts). For NASCAR Sprint Cup, this would be self-explanatory and would be defined as (43 - avg. finish)/42; in F1 and IndyCar, it would be a little less self-explanatory since I do not count drivers who fail to start the race in those series, not to mention that IndyCar has different numbers of competitors per race and F1 occasionally but increasingly rarely has DNQs of regular season entries. A percent beat of 80%, regardless of series, will usually be sufficient to be a championship favorite, while a percent beat of 70% will usually get you close if you're extremely dominant but somewhat consistent...obviously this varies somewhat based on how reliable the equipment is within a series, whether it is spec or not, and so on, but I do think having a percentage beat of 80% comes close enough to perfect for my liking. A percent beat of 50% would be average within a given field, and I think that should be worth slightly more than 1 point, especially considering most drivers in most series do not come close to dominating often. I like 40% as the absolute bottom threshold for a competitive season, because pretty much every driver who has any kind of longevity in a series should be able to do that once, while any drivers who fail to do even that did not contribute much at all (unless they had even worse equipment than that on a percentile basis, which does happen sometimes).

While I acknowledged above there are many other things like speed, equipment strength, versatility, on-track passing, and so on that are essential in measuring overall performance, the advantage of this compared to my earlier work is its simplicity. You can get a cut-and-dried sense of a driver's importance based on stats alone (which is truly objective, and a nice corrective to some of my more subjective work such as 'How the Races Were Won', which I still think is also important), and then once I come up with good ways of measuring equipment strength, versatility, and other factors, I can adjust my current basic formula accordingly. The advantages of this model are its simplicity, its ability to be applied to any driver in any series in any era as long as I have the finishing order and lap leader data which I will for many, many series, using average percent led instead of laps led to measure dominance which removes the bias from circuit to circuit, using percent beat instead of average finish to measure consistency, which removes the bias towards series that have fewer cars, and considering all three popular perspectives in evaluating racing talent all at once when summarizing a season, which I'm not sure has been done before. The drivers who are most feared are the drivers who can be consistent, can win, AND can dominate. For instance, Kyle Busch can win and dominate more or less to the degree Jimmie Johnson does, but he is nowhere near as consistent so that explains why he isn't feared as much. Graham Rahal has been good at winning this season, but both of his wins were pretty fortunate and I imagine people in the IndyCar paddock are just as worried about Sébastien Bourdais and Josef Newgarden, who have won just as much and dominated much more, even if they've been less consistent. It's all important, but too many people tend to focus on one metric at the expense of the others. The main drawback of course is the bias towards drivers in stronger cars, but winning races, earning consistent finishes, and dominating races are also all biased towards drivers who have stronger cars. Another drawback is that equating all races ignores that it may be more difficult to perform at some tracks or types of races than others, so this ignores versatility, but there's really no *objective* way to measure what the most difficult circuits are (although I admit I have my own personal biases in this regard), and I think this still does a quite good job at measuring performance regardless.

However, wouldn't this be even better if I attempted to calculate a driver's performance at a particular instant in time, rather than just a discrete measure of season performance? This would correct some of the issues like equipment strength (Fernando Alonso may be stuck in a miserable McLaren this year, but a weighted average over several years would help correct for this by including earlier seasons he had that were more successful; similarly some drivers who work their way up from overachieving in weaker equipment to driving stronger cars would have their seasons for multiple teams of differing strengths considered in a weighted average, which should likely remove a good portion of the bias based on differing equipment strength. Drivers who overachieve enough will end up in stronger cars (like Simon Pagenaud did this year) or help improve the strength of their fledgling team (like Josef Newgarden did this year), so a multiple-year weighted career score based on a set of season scores is the best relatively simple measure of a driver's ability in a particular season. Any driver can have a good random lucky season. Jimmy Vasser may have won the CART title in 1996, Bobby Labonte the NASCAR title in 2000, and Jenson Button the F1 title in 2009. I'm not going to necessarily say any of those drivers didn't have the best seasons those years, although I certainly think there's a lot of room to debate all three of them. However, regardless of whether they won the title or not, would you have gotten many people to say that they were actually the best driver in the sport overall over an extended period of time? A single season is not the long run, and a driver can get lucky in the short run while being unable to sustain it for an extended period. When measuring a driver's ability, it should be based on much more than merely the current season alone. For instance, Dale Earnhardt won three titles in the five year period 1986-1990 but four in the five year period 1990-1994. When was he better? Most people would say the championships trump everything else, but for most of the early '90s he relied on consistency rather than raw dominance. Coming off of his 1986-87 dominance, I would say he was much more feared in 1988-89 even though he didn't win the titles than in 1991-94 when he did but was starting to have other drivers surpass him in dominance. He was probably overall the dominant driver in NASCAR over the entire period, but I would say 1986-90 was his best half-decade, not 1990-94, when considering all three methods of assessing performance, and this model does acknowledge that. I prefer measuring who the best driver is overall, regardless of who the actual champion was, because the championship, just like an individual race, can be a bit of a crapshoot where an inferior driver gets lucky with a much more dominant car or superior pit strategy or better reliability (for which the driver may or may not play a significant role), but measuring performance over a five-year period to determine one's ability in a given season gives a much better idea of what drivers are actually most feared at a given time. Jimmy Vasser and Bobby Labonte were likely never the drivers most feared in American open wheel or NASCAR, as this would reflect. That doesn't necessarily mean they didn't deserve their titles though. It just means that there are other factors that might be more important than titles. Just as the Oscar for Best Picture does not always go to the best movie in a year, neither does a championship necessarily go to the best driver, particularly because it's very difficult to separate out the team factors. This system does about the best job of removing team-related factors as I think I can do with a simple system based purely on stats.

So how did I weight drivers' performance over multiple years? For drivers who had two seasons before or two seasons after given season n, I used the following formula: .4*(year n score) + .2*(year n + 1 score) + .2*(year n - 1 score) + .1*(year n + 2 score) + .1*(year = n - 2 score) to create a weighted measure of the driver's performance in time. For instance, Lewis Hamilton in 2013 had a season score of 9. His 2011 and 2012 scores were 12 and 13 respectively, while his 2014 and 2015 scores were both 15s (unfortunately, Nico Rosberg also got a 15 both seasons even though Hamilton clearly outperformed him...in non-spec series when one team has THAT dominant an advantage; unfortunately, this will happen, but I could improve on this by redefining the thresholds for each category depending on the nature of the series). Hamilton therefore earned .4*9 + .2*15 + .2*13 + .1*15 + .1*12 = 11.9. Hamilton's 2013 score of 11.9 reflects his performances for the entire 2011-15 period as a way of estimating his 2013 ability, which was just barely 2nd place in Formula One in that year behind Sebastian Vettel's score of 12.1, which sounds about right.

If a driver did not have two consecutive seasons before and after a given year, I adjusted the formula weighting the current year 4 times, weighting any years before or after twice, and weighting any years two years before or after a single time. Hence, if a driver only had four consecutive years to judge, I would simply divide by a denominator of 9 instead of 10 excluding the year on whichever end had missing data. If a driver only had three consecutive years to judge (such as 2013-2015), I would weight the current year 4 times, the past year 2 times, and the first year 1 time (or if I was evaluating 2013 instead of 2015 based on the same data, I would weight 2013 4 times and 2015 times only once...if I was instead evaluating 2014, I would weight that year twice and the surrounding years once each. I included any years a driver had any starts regardless of the number of starts. I decided there was too much data lost if I failed to consider any seasons, even those with only a start or two. For instance, especially historically, marquee drivers crossed over to appear in races in other series, and I wouldn't necessarily want to exclude Formula One drivers who appeared in the Indy 500 and no other IndyCar races in the '60s, CART drivers crossing over to race at Indy in the IRL era, versatile drivers like A.J. Foyt, Mario Andretti, and Dan Gurney doing one-offs in NASCAR, drivers like Rick Mears, Ernie Irvan, or James Hinchcliffe having injury-shortened seasons, and so on. It's very rare that a driver with few starts will actually win, so in most cases, drivers with extremely partial seasons actually do have season scores comparable to most of their surrounding season scores. There are a few outliers in this regard (such as James Hinchcliffe having a score of 11 this year because he won a flukish race in a five-race injury shortened season, when his performance isn't near that of most 11-caliber seasons), but there are surprisingly fewer flukes in this regard than you'd expect. I think there is too much information lost if I use only a minimum number of starts, although there are some occasional scores that will need to be taken with a grain of salt for this reason, and Hinchcliffe's is a good example. If a driver had a gap between seasons, I calculated all season scores as if they were based on separate careers, so Alex Zanardi's 1996-98 dominance was not considered in determining his 2001 score, nor was Michael Schumacher's Ferrari-era dominance used to evaluate his Mercedes years.

Now that I have determined a way to measure each driver's performance in one moment of time and remove most of the possible biases, it's worth noting that there are a lot of different applications of this that I haven't even mentioned here yet. One can view these weighted career scores as the contribution that a particular driver makes to a given race. Furthermore, it would thereby follow that the sum of all weighted career scores would thereby measure the strength of the field. I used these statistics to determine what the strongest and weakest F1, IndyCar, and NASCAR fields ever were, but I will save that for a future column. I can create an 'age line' as is seen in other sports by comparing drivers' career scores at certain ages and periods of times - I suspect as drivers have gotten more and more athletic in recent decades that the age line may shift, and that will be interesting to find out. I can also compare drivers who switched from one series to another and thereby begin the process of comparing series strength. This would especially be useful for comparing feeder series to the major league series in each discipline, but I could also compare F1 drivers who crossed over to CART and vice versa, or CART drivers who crossed over to IRL and vice versa. Since all three measures (winning, consistency, and dominance) are based on percentages, it is essentially unit-free so direct comparisons can be made across series. These ideas will likely form the foundation of any cross-series driver rankings I will introduce in future, although I will probably also include a speed factor if I do that, since I am only interested in doing overall driver rankings for current seasons (for which speed data actually will be available). Over the past month or two I've done a lot of this work already. I attempted to use these driver scores to determine how many times stronger CART was in 1995 than the IRL was in 1996. Just to give you a teaser for potential future columns, I have estimated based on these driver weighted score figures that CART 1995 was seven times stronger than IRL 1996 (i.e. 1 CART win would be worth about 7 early IRL wins), but CART itself was only about 10-15% weaker in 1996. I have a lot more work to do before I start doing columns on all of these things, but it's exciting that I've finally got a way to make the direct comparisons that people have been debating (with minimal proof) for years.

Below I provide a list of each driver's 'superstar period', weighted career peak score, and overall peak season (regardless of performance in the surrounding seasons). Another neat thing I can do with this formula is come up with an objective definition for what a superstar is. I have decided that in F1, IndyCar, AND Cup a superstar is a driver who sustains a weighted average of 5 in two or more seasons (even if they are non-consecutive). It is pretty rare that a driver will exceed a weighted average score of 5 over multiple seasons unless they are pretty special. I decided one season is not enough because it would be too easy for someone to for instance beat 80% of the cars in a single race and compete in no other races to obtain a 5 score, particularly if somebody entered only a hometown or home country race in F1 or NASCAR many decades ago when few cars finished. Single fluke seasons can happen now and then not infrequently and I want to be sure that a driver has sustained superstar status before listing them as a superstar. You may quibble with a few drivers on each of the following lists and say that they aren't quite superstars (especially the IRL, when the real superstars were in CART at the time) but for the most part, I believe my ideas are right. That does mean that it takes a while for emerging drivers to appear on a list like this. Josef Newgarden and Graham Rahal both crossed over the 5 boundary necessary for superstardom this season, but need to continue to do well enough in 2016 in order to confirm that they have indeed entered superstar status. Considering their earlier seasons had shown mere potential that hadn't been fully realized, this is understandable. A similar case also exists in Formula One where Daniel Ricciardo entered superstar status in 2014 before dropping out again this year. I do think Ricciardo, Newgarden, and Rahal *will* eventually be on these lists, but I want to remain conservative in determining which drivers are true superstars, and which years they maintained superstar status. Unfortunately, there are some obvious flaws. For instance, Alain Prost took a one-year hiatus in 1992 before having a dominant year in a dominant Williams in 1993, which left him as the only driver with a career peak of 15, which was only based on one season. Similarly, drivers who had *extremely* dominant cars, which is most important in F1, such as Nico Rosberg, will likely be overrated by this. Remember however that this is purely based on stats. Given that, I think it should be easily understandable why each driver is where they are. Adding an equipment strength factor to this if I do so will come later.

Formula One

DriverSuperstar PeriodCareer PeakSingle-Season Peak
Alain Prost80-9315.0 (93)15 (85-86, 88-89, 93)
Michael Schumacher92-0614.9 (02)15 (94, 98, 01-04, 06)
Jackie Stewart65-7314.71 (73)15 (71, 73)
Juan Manuel Fangio50-51, 53-5814.6 (55)15 (51, 54-57)
Sebastian Vettel08-1514.6 (11)15 (11-13)
Lewis Hamilton07-1514.14 (15)15 (07, 14-15)
Nico Rosberg12-1514.14 (15)15 (14-15)
Stirling Moss54-6113.78 (60)15 (56, 60)
Jim Clark61-6813.71 (68)15 (63, 68)
Ayrton Senna84-9413.7 (89)15 (88, 91)
Mika Hakkinen97-0113.67 (00)15 (98, 00)
Fernando Alonso03-1413.5 (06)15 (05-07)
Niki Lauda74-78, 82-8513.5 (76)15 (75)
Nigel Mansell84-92, 9413.43 (92)15 (92)
Damon Hill92-9813.2 (95)15 (94, 96)
Nino Farina50-5513.0 (50)15 (50)
Alan Jones78-8113.0 (81)15 (80)
Alberto Ascari50-5512.8 (52)15 (51-53)
Jacques Villeneuve96-9812.43 (96)14 (96-97)
Emerson Fittipaldi70-7511.7 (73)14 (72)
Graham Hill62-6911.6 (64)15 (62, 65)
Nelson Piquet80-9111.5 (86)15 (86-87)
Kimi Raikkonen03-09, 12-1311.4 (07)15 (07)
Rubens Barrichello00-04, 0911.1 (02)14 (02)
Jenson Button09-1311.0 (11)15 (09)
David Coulthard94-0310.8 (99)12 (99)
Felipe Massa06-0910.5 (07)14 (08)
Jacky Ickx67-7210.5 (70)12 (70)
Mark Webber09-1310.3 (10)14 (10)
Jose Froilan Gonzalez50-5510.3 (53)13 (51, 54)
Jochen Rindt69-7010.29 (70)13 (70)
Mike Hawthorn52-589.86 (58)13 (58)
James Hunt75-779.8 (76)14 (76)
Peter Collins56-589.57 (58)13 (56)
Didier Pironi80-829.57 (82)13 (82)
Jack Brabham59-61, 65-68, 709.5 (60)15 (60)
Juan Pablo Montoya01-069.5 (03)12 (03, 05)
Jody Scheckter72-809.3 (77)15 (79)
Rene Arnoux80-849.1 (83)14 (83)
John Surtees63-679.1 (64)13 (64)
Mario Andretti76-789.1 (77)13 (78)
Denny Hulme66-749.0 (68)14 (67)
Ronnie Peterson71-75, 789.0 (73)13 (73)
Phil Hill58-618.8 (61)15 (61)
Tony Brooks57-598.8 (58)12 (58-59)
Riccardo Patrese89-938.8 (91)12 (91)
Gilles Villeneuve78-828.7 (79)15 (79)
Luigi Fagioli50-518.67 (51)10 (51)
Piero Taruffi51-528.57 (52)13 (52)
Ralf Schumacher01-048.4 (03)12 (03)
Keke Rosberg82-868.0 (85)11 (85)
Carlos Reutemann73-827.8 (74)13 (74, 78)
Clay Regazzoni70, 74-767.71 (70)12 (70)
Gerhard Berger86-92, 94, 96-977.7 (87)11 (87)
Jacques Laffite79-817.6 (79)10 (79, 81)
Richie Ginther60-61, 657.43 (60)9 (60)
Patrick Depailler78-797.33 (79)13 (79)
Patrick Tambay82-847.22 (82)11 (82)
Bruce McLaren58-617.0 (60)10 (60)
Eddie Irvine98-996.9 (99)14 (99)
Dan Gurney62-646.7 (64)12 (64)
Thierry Boutsen88-906.7 (89)10 (89)
Michele Alboreto84-856.3 (85)10 (85)
Heinz-Harald Frentzen97, 995.9 (99)11 (99)
Jean Alesi95-965.9 (95)8 (95)
John Watson82-835.89 (82)8 (82)
Francois Cevert71, 735.85 (73)9 (71)
Giancarlo Fisichella05-065.8 (05-06)9 (06)
Pedro Rodriguez70-715.29 (71)8 (70)
Lorenzo Bandini64, 665.1 (64)9 (64)

Champ Car

DriverSuperstar PeriodCareer PeakSingle-Season Peak
Sebastien Bourdais03-0715.0 (06-07)15 (04-07)
Alex Zanardi96-9814.43 (98)15 (97-98)
Michael Andretti85-92, 94-0214.0 (92)14 (87, 90-92)
Rick Mears79-9213.71 (79)15 (81)
Bobby Unser79-8113.57 (79)15 (79)
Bobby Rahal82-9313.1 (86)14 (86-87)
Juan Pablo Montoya99-0013.0 (99)14 (99)
Al Unser, Jr.82-9612.7 (90)15 (90, 94)
Mario Andretti79-90, 9312.2 (84)14 (84)
Emerson Fittipaldi85-9512.0 (89)15 (89)
Johnny Rutherford79-8111.86 (79)15 (80)
Cristiano da Matta00-0211.43 (02)14 (02)
Jacques Villeneuve94-9511.33 (95)14 (95)
Bruno Junqueira01-0711.3 (04)14 (05)
Nigel Mansell93-9410.67 (93)14 (93)
Gil de Ferran95-0110.57 (01)12 (01)
Paul Tracy91-0710.4 (03)14 (03)
Danny Sullivan84-9110.3 (88)14 (88)
Tom Sneva79-8510.3 (82)13 (84)
Helio Castroneves99-0110.29 (01)12 (01)
A.J. Allmendinger05-069.43 (06)13 (06)
Kenny Brack00-029.25 (01)13 (01)
Teo Fabi83-849.0 (83)13 (83)
Justin Wilson04-079.0 (07)11 (05)
Gordon Johncock79-848.71 (79)12 (79)
Jimmy Vasser96-008.6 (97)12 (96)
Al Unser79-81, 83-85, 87-888.29 (79)10 (79, 83, 85)
Dario Franchitti97-028.22 (98)12 (98-99)
Adrian Fernandez98-008.11 (99)10 (99)
Patrick Carpentier02-048.0 (04)8 (02-04)
Greg Moore96-997.44 (98)9 (98)
Bryan Herta98-995.7 (98)7 (98-99)
Roberto Moreno00-015.57 (01)8 (00)
Andre Ribeiro95-965.11 (96)8 (96)


DriverSuperstar PeriodCareer PeakSingle-Season Peak
Juan Pablo Montoya00, 14-1515.0 (00)15 (00)
Dario Franchitti04-07, 09-1314.71 (09)15 (07, 09, 11)
Scott Dixon03-1514.1 (08)15 (08-09)
Will Power08-1513.5 (11)14 (09-11)
Helio Castroneves01-1512.71 (01)14 (06)
Dan Wheldon03-1112.5 (06)15 (05)
Gil de Ferran01-0311.86 (03)13 (02-03)
Sam Hornish, Jr.00-0711.6 (02)15 (01)
Arie Luyendyk96-9911.14 (96)13 (96)
Tony Kanaan02-10, 13-1511.1 (05)14 (04)
Richie Hearn96-9710.33 (97)14 (97)
Ryan Briscoe07-1110.0 (09)13 (09)
Tony Stewart96-99, 019.56 (97)12 (98)
Kenny Brack97-999.43 (99)12 (98)
Buddy Lazier96-019.4 (00)14 (01)
Ryan Hunter-Reay10-159.1 (12)11 (12)
Scott Sharp96-028.9 (98)10 (98)
Adrian Fernandez04-058.67 (04)12 (04)
James Hinchcliffe13-158.43 (15)11 (15)
Sebastien Bourdais14-158.29 (15)10 (15)
Scott Goodyear97-007.9 (99)11 (99)
Eddie Cheever97-017.9 (99)10 (00)
Greg Ray99-017.8 (00)12 (99)
Mike Conway13-147.43 (14)12 (13)
Al Unser, Jr.00-027.43 (00)9 (00)
Robby Gordon97, 99-007.14 (99)9 (99)
Michael Andretti01-03, 066.71 (01)8 (01)
Simon Pagenaud12-156.67 (14)8 (14)
Tomas Scheckter02-036.57 (02)8 (02)
Eliseo Salazar96-975.86 (96)10 (97)
Robbie Buhl96-975.67 (97)8 (97)
Billy Boat97-985.43 (97)8 (98)
Bruno Junqueira01-02, 045.33 (04)8 (04)
Buddy Rice02, 045.29 (02)11 (04)


DriverSuperstar PeriodCareer PeakSingle-Season Peak
Richard Petty60-8414.8 (71-72)15 (67, 70-72, 74-75)
David Pearson61-8014.6 (74)15 (66, 68-69, 72-74, 76)
Darrell Waltrip75-9214.2 (81-83)15 (79, 81, 83)
Jeff Gordon94-1514.2 (97)15 (98)
Cale Yarborough66-8514.1 (78)15 (74, 77-78)
Dan Gurney62-7013.9 (65)15 (63, 65-66, 68)
Herb Thomas50-5613.7 (53)15 (53)
Dale Earnhardt79-0113.6 (87)15 (87, 90, 93)
Fred Lorenzen60-6713.3 (64)14 (63-64)
Paul Goldsmith56-58, 66-6713.14 (58)15 (58)
Bobby Allison66-8813.1 (82)15 (83)
Bobby Isaac67-7212.9 (70)15 (70)
Jimmie Johnson02-1512.7 (08)13 (04, 07-09, 12-13, 15)
Fireball Roberts50, 56-6412.6 (58)15 (58)
Ned Jarrett59-6612.3 (64)15 (65)
Buck Baker52-5912.3 (56)14 (56-57)
Kevin Harvick01-1512.29 (15)13 (14-15)
Bill Elliott83-93, 02-0312.2 (87)14 (85, 87-88)
Mark Martin89-00, 04-05, 08-0912.0 (98)15 (98)
Rusty Wallace86-0211.9 (94)14 (88-89, 93-94)
Lee Petty49-6211.7 (59)14 (54)
Dale Jarrett95-0211.7 (97)14 (97)
Tim Flock49-5711.5 (52)15 (55)
Rex White56-6311.5 (59-60)13 (59)
Parnelli Jones57-60, 67-7011.43 (67)15 (67)
Marshall Teague51-5211.43 (52)14 (52)
Fonty Flock49-5711.4 (51)14 (51)
Tim Richmond82-8711.29 (87)13 (86)
Marvin Panch55-58, 61-6611.22 (65)14 (65)
A.J. Foyt63-65, 69-7411.2 (71)15 (72)
Dick Hutcherson64-6711.11 (66)13 (65)
Johnny Mantz50-5111.0 (50)15 (50)
Tony Stewart99-1311.0 (05)13 (05)
Junior Johnson58-6610.89 (65)13 (65)
Davey Allison87-9310.78 (92)13 (91-92)
Jack Smith56-6210.6 (60)14 (60)
Joe Weatherly58-6410.6 (61)13 (61)
Red Byron49-5110.57 (49)13 (49)
Kyle Busch06-1510.5 (08)14 (08)
Denny Hamlin05-1510.5 (10)13 (10)
Bob Welborn56-60, 6410.4 (58)15 (57)
Curtis Turner49-61, 6510.29 (49)13 (59)
Benny Parsons75-8410.2 (77)13 (77)
Matt Kenseth02-1510.2 (13)13 (13)
Ernie Irvan90-9610.1 (94)14 (94)
Donnie Allison68-71, 76-7910.1 (70)13 (70)
Buddy Baker68-8110.1 (71)12 (75)
Jeff Burton97-0110.1 (99)11 (99-00)
Chuck Stevenson55-5610.0 (56)15 (56)
Glen Wood58-61, 63-6410.0 (63)14 (60, 63)
Brad Keselowski11-159.89 (14)12 (14)
Joey Logano13-159.71 (15)11 (14)
Harold Kite50-519.67 (50)13 (50)
Bobby Labonte96-029.6 (99)13 (99)
Eddie Gray57-619.57 (61)15 (61)
Dick Rathman51-559.56 (54)12 (54)
Danny Letner54-55, 61, 639.43 (54)11 (54-55)
Jack White49-509.29 (49)15 (49)
Jim Reed56-609.2 (58)14 (58)
Dale Earnhardt, Jr.00-06, 12-159.14 (15)11 (04)
Ray Elder70-739.1 (72)15 (72)
Speedy Thompson53-609.1 (56)13 (53)
Lee Roy Yarbrough66-709.0 (69)13 (69)
Johnny Beauchamp57, 59-618.86 (59)12 (59)
Kurt Busch02-11, 13-158.8 (04)11 (15)
Neil Bonnett78-868.7 (81)10 (81)
Frankie Schneider57-588.67 (58)12 (58)
Carl Edwards05-158.6 (08)13 (08)
Harry Gant82-86, 90-928.6 (84)12 (91)
Terry Labonte82-88, 94-988.6 (96)10 (96)
Ryan Newman02-058.5 (03)12 (03)
Jim Paschal54-678.2 (62)13 (58)
Cotton Owens58-61, 63-648.14 (64)13 (64)
Greg Biffle04-128.0 (05)12 (05)
Geoff Bodine84-948.0 (86)10 (86, 90)
Bill Blair49-527.71 (49)9 (49)
Dave MacDonald63-647.67 (63)10 (63)
Lloyd Dane54-617.57 (61)9 (56)
Hershel McGriff53-547.57 (54)9 (54)
Art Watts57-587.5 (57)13 (57)
Pete Hamilton69-717.4 (70)12 (70)
Darel Dieringer63-677.4 (66)9 (63, 66-67)
Charlie Glotzbach68-717.4 (70)9 (70)
Gwyn Staley57-587.33 (57)12 (57)
Kyle Petty90-937.3 (92)9 (92)
Bob Flock49-527.29 (49)9 (49)
Ralph Moody56-57, 597.0 (56-57)7 (56-57)
Ricky Rudd83-95, 00-026.9 (86-87)9 (01)
Clint Bowyer10-136.8 (12)9 (12)
Nelson Stacy61-626.71 (61)10 (62)
Sterling Marlin94-96, 01-026.7 (95)10 (95)
Martin Truex, Jr.13, 156.57 (15)9 (15)
Mario Andretti67-696.56 (67)12 (67)
Kasey Kahne04-06, 08-146.5 (06, 12-13)10 (06)
Alan Kulwicki90-936.44 (92)10 (92)
Gober Sosebee49, 526.29 (49)9 (49)
Neil Cole50-516.22 (51)11 (51)
Dave Marcis75-766.1 (76)9 (76)
Eddie Pagan56-586.0 (57)9 (57)
Dick Meyer51-535.86 (51)7 (51)
Ken Schrader89-915.8 (89)8 (91)
Tiny Lund69, 715.4 (69)10 (71)
Allen Adkins54-555.29 (54)7 (55)
Bobby Johns60, 625.11 (60)8 (60, 62)
Don White54-555.0 (54-55)5 (54-55)
Jack Radtke55-565.0 (55-56)5 (55-56)
Al Unser68-695.0 (68-69)5 (68-69)

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.