Archive for September, 2010

Does Power have a prayer?

Posted in Uncategorized on September 23, 2010 by Sean Wrona

Will Power enters the IZOD IndyCar Series finale at Homestead-Miami Speedway with a twelve-point lead over Dario Franchitti. Despite the fact that Power has led the points standings for the entire season with the exception of the June race at Texas Motor Speedway, Franchitti has considerably more experience on ovals than the Aussie and has beaten Power all but twice on ovals since reentering the series in 2009, outscoring him by 95 points in those nine events.

I undertook a simple calculation to determine how many times Franchitti would have beaten Power by thirteen or more points based on their previous oval results (Power would win the championship if Franchitti gained only twelve points since he has more wins). Power scored 14, 16, 18, 22, 24, 30, 31, and 35 points in his eight oval starts for Penske (discounting this year’s Indianapolis 500 because it had bonus points for qualifying which other races do not have), while Franchitti scored 14, 26, 28, 30, 30, 40, 40, and 50 points in his eight oval starts for Ganassi. Franchitti thus outscored Power’s 14 points by thirteen or more points six times, beat Power’s 16 points five times, exceeded his 18, 22, and 24 points three times, and surpassed his 30, 31, and 35 points exactly once.

Combining all these possibilities in this very rudimentary simulation, Franchitti beat Power by more than twelve points in only 23 of 64 situations, indicating that despite Franchitti’s greater oval experience and despite Franchitti winning the last two championships he has contested, the upstart Power remains a slight favorite based on their past results, but it remains very clear that the IndyCar championship could go either way just like in 2006, 2007, and 2009.

However, what happens if you limit the discussion to be cookie-cutter ovals only (here defined as Chicagoland, Homestead, Kansas, Kentucky, and Texas?)  This excludes the 2009 Indianapolis 500 and 2010 Iowa race, the only two races where Power beat Franchitti straight-up, and last weekend’s race at Motegi (which given its egg shape rather than the standard quad-oval shape should probably be disqualified from the discussion, as it’s probably closer to a track like Darlington or Gateway in style than it is to the other five 1.5-milers).   This is probably more accurate than the above analysis because using Indianapolis, Iowa, or Motegi results to predict Homestead may be a bit disingenuous.

Power’s one clear weakness this season has been the cookie-cutter ovals, where he has scored four of his five finishes outside the top five this season.   He has only competed on this type of track in competitive equipment five times – last season’s Kentucky race and this year’s Chicagoland, Kansas, Kentucky, and Texas races, and all five times he has been outpaced by Franchitti; in fact Power’s best performance of 24 points in the five cookie-cutter races they’ve competed against each other is worse than Franchitti’s worst performance of 28 points.   Power’s five points totals were 14, 16, 18, 22, and 24 points, while Franchitti’s were 28, 30, 30, 40, and 50 points.   Using the same kind of pseudo-simulation as before, Franchitti would beat Power by 13 or more points in 15 of 25 situations, giving him a 60% probability of winning the championship.  Since predicting Homestead by past cookie-cutters is more sensible than by ovals in general, I hereby dub Franchitti the favorite despite Power’s points lead.  That may overstate Franchitti’s odds because several of Power’s bad runs (especially Chicagoland) had to do with pit miscues and little to do with Power himself, unlike last season, when Ryan Briscoe’s crash at the end of pit road caused him to lose the championship.  It also helps Power that he has two teammates to play defense against Franchitti (just as Castroneves did at Motegi by keeping Franchitti from winning), while Franchitti only has one.  In spite of that, I think Power’s inexperience on ovals will lead to an impressive third consecutive championship for Franchitti (obviously not counting 2008 which he did not contest due to his ill-fated NASCAR experiment).

Who the chasers should have been

Posted in Uncategorized on September 15, 2010 by Sean Wrona

Of course I think the Chase is misguided and is one of the things that makes NASCAR very hard to take seriously at times.  But even disregarding that, the traditional NASCAR points system was also probably the worst in motorsports.  Unlike most other major points systems (F1, WRC, CART, MotoGP, ALMS) everyone scores points, including those who fail to finish the race.  Despite all the talk about how important victories are, the percentage gap between the top positions was never as much as most other racing series, and until 2004, first and second place could and frequently did tie in points, when the second-place finisher led the most laps, thus making “points racing” to collect easy top ten finishes an almost strictly NASCAR phenomenon (although other series with the same general points structure, such as NASCAR-owned Grand-Am operate the same way).  Although NASCAR hypes its ten-point bonus for winning a race before the chase, anyone who really knows anything knows that is not nearly enough incentive to get people to fight for wins, when losing a top ten position is very costly.  Once I began inputting NASCAR races into my database, I experimented with a superior points system that would reflect more sizable gaps between positions like in other racing series, not award those who finish in lower positions, and make it harder to collect lap-leader bonuses, since staying out of the pits for one lap to gain 5 bonus points is far too easy (look at some of the drivers who have led laps in recent seasons).

The points system I came up with was as follows:

1. 200
2. 160
3. 130
4. 115
5. 100
6. 90
7. 80
8. 70
9. 60
10. 50
11. 45
12. 40
13. 35
14. 30
15. 25
16. 22
17. 19
18. 16
19. 13
20. 10
21. 8
22. 6
23. 4
24. 2
25. 1
Pole. 10
2 bonus points for each 10% of the race you lead.

Rewarding only the top 25 positions would probably prevent teams from bringing wrecked race cars back onto the track (especially critical at places like restrictor plate tracks where this kind of business can only lead to even bigger messes).  This would also prevent many payback situations, like the Carl Edwards wreck of Brad Keselowski in the spring Atlanta race.  The 99 team would have had no reason for Edwards to go back onto the track if 26th-43rd scored the same number of points.  Rewarding the top five and top ten positions much more would reward drivers who ran up front and fought for wins, rather than stroking (like most of Roush Fenway Racing has done this year).   I made sure to have the points totals for 5th, 10th, 15th, 20th, and 25th all be memorable because it might be less confusing to fans than a points system where those prominent positions are worth 155, 134, 118, 103, and 88 points respectively.   I felt winning the pole should be worth some small amount and added a ten point bonus for this.  I felt dominance in the race should be rewarded more than it presently is also.  The average percentage led in a race is as I have noticed the best measure of dominance in a race, better than laps led, miles led, and even wins (because it is easy to win without leading many laps if you have a good pit strategy for instance).   Given that, instead of awarding 5 points to all leader I decided in my mythical points system to reward drivers who led different percentages of the race differentially.  Someone who led 10-19.99% of a race would score 2 points, someone who led 20-29.99% would score 4, going all the way up to 20 bonus points for someone who led the entire race.  This would prevent people from scoring bonus points simply by leading a lap before entering the pits (or worse, taking the lead in the pits based on where the start-finish line is in relation to the leader’s pit stall, a more frequent occurrence in recent years).   It would also mean if two drivers dominated a race, the driver who led the second most laps would not be short-changed by scoring only 5 points (the same as all other leaders).  I’ve been calculating what the points would be using this system all season, and did so for past seasons as well, and in virtually every case, I think that the points system I’ve come up with is a better measure of performance.  It would also be more exciting because with a maximum of 230 points available, and 0 points possible if you finished worse than 25th, it would be much easier to make up points if you are behind if you start winning.

Like I said, the Chase is garbage, but if you still wanted to continue with the chase, I think the 12 drivers selected by my system are more reflective of the top 12 drivers in the Sprint Cup season to date than the 12 drivers the actual Chase selected.  Harvick still leads by a sizable margin as he does in the classic points, proving that consistency does still mean something, but Johnson and Hamlin, by far the top two winners, and the two biggest lap leaders on the season as well, move to second and third, more reflecting their actual performance this year in my opinion than their DNFs which left their seasons very underrated points-wise.  Jeff Gordon, who didn’t get the wins was thereby penalized and is in a more representative position with regard to his performance.  Finally and most notably, the Earnhardt-Ganassi entries of Jamie McMurray and Juan Pablo Montoya would have made the chase while the Roush-Fenway entries of Greg Biffle and Matt Kenseth would have missed.  If you look at McMurray and Montoya’s wins, top 5s, top 10s, and laps led it does seem somewhat shocking that neither of them made the chase.  McMurray had more wins than seven chasers, more poles than all of them, and more top fives than five of them, which would seem to indicate he deserves a spot in the chase, despite his relatively paltry 9 top tens (but Matt Kenseth only had one top ten and was very mediocre for several months entering the chase and still made it).  Montoya meanwhile had 13 top tens (more than only two chasers, but setting him WELL beyond all the other bubble drivers) but his DNFs proved costly.  While I acknowledge Montoya probably should have known better than to fight for the lost-cause wins at Loudon and Indianapolis which ultimately led to his wrecks (and had he driven cautiously in those two races, he probably would have made the chase), it does seem perverse that losing hurts more than winning helps, and Montoya had much more bad luck in the first third of the season than most other drivers.  It seems wrong that Montoya, a guy who led almost twice as many laps as Kevin Harvick and led 14 races (tied for fifth in the league with Jeff Gordon) isn’t making the chase while a guy like Kenseth, who led barely a lap a race, is.  I think my points system more accurately reflects who ran well in general.  Obviously, I do not have pull with NASCAR, but if they are serious about wanting to reward winning, with or without the chase, the base points system needs to be totally revamped.  If the goal is simply to reward consistency (which it has been in the past), they should just admit it and market that.

Below are the top 25 in my points system for 2010 and where they would be in the classic points system after 26 races in parentheses:

1. (1) Kevin Harvick – 2347
2. (6) Jimmie Johnson – 2075
3. (9) Denny Hamlin – 1987
4. (2) Kyle Busch – 1934
5. (10) Kurt Busch – 1834
6. (3) Jeff Gordon – 1784
7. (4) Carl Edwards – 1682
8. (5) Tony Stewart – 1659
9. (7) Jeff Burton – 1453
10. (14) Jamie McMurray – 1447
11. (11) Clint Bowyer – 1366
12. (16) Juan Pablo Montoya – 1353
13. (12) Greg Biffle – 1347
14. (8) Matt Kenseth – 1320
15. (18) Kasey Kahne – 1126
16. (13) Ryan Newman – 1108
17. (17) David Reutimann – 1033
18. (15) Mark Martin – 1001
19. (21) Joey Logano – 927
20. (19) Dale Earnhardt, Jr. – 872
21. (20) Martin Truex, Jr. – 797
22. (22) A.J. Allmendinger – 727
23. (24) Marcos Ambrose – 724
24. (23) Paul Menard – 485
25. (25) David Ragan – 368

Doesn’t that look much more accurate with regard to who’s running up front?  And also, given the small gap between 11th and 14th after 26 races, wouldn’t that have been a WAY better battle for the final transfer spot?

Finally, albeit a bit too late…

Posted in Uncategorized on September 15, 2010 by Sean Wrona

I’ve joined the racing blogosphere.  I’ll probably be opinionated to some degree as is my nature.  But I am mainly going to try and focus on racing statistics which my site covers.

I took a break for a while from adding new racing series to the main site.  I needed a break after entering 12 NASCAR races a day most days for several months, but I am in the process of (somewhat slowly) adding MotoGP, which is among the most prestigious racing series I have not yet archived.  I also intend to add the European Le Mans Series shortly thereafter.  If you tried to access one of the 24 Hours of Le Mans pages in the past few days, I apologize.  I made an error in code when attempting to edit my results page to accommodate MotoGP results that caused Le Mans results to go blank; this has been fixed.

If you’re wondering how I decide which series to cover, there are basically three rules:

1. I generally do not cover series that are defunct, because the data would generally be too difficult to obtain.  So I am unlikely to cover series such as Can-Am or historic Trans-Am (before this year’s series revival) or probably IMSA any time soon.  CART and Champ Car are exceptions as they are probably the most prominent defunct series, and historical IndyCar data is much more available than most other forms of racing.  I am likely to at some point venture into archiving AAA and USAC (IndyCar racing prior to 1979).

2. I am not interested in archiving minor-league feeder series for the most part, and would rather cover all the active major racing leagues in the world than the minor leagues in specific disciplines.  So a site that specializes in NASCAR history would be better for Nationwide/Craftsman Truck statistics and a site that specializes in IndyCar history would be better for Atlantics/Indy Lights statistics.  I do not plan on entering any feeder series in the near future.

3. I have a strong preference for series with a short history or small number of races, because I can enter them faster, so I may have a bias towards entering series that started in the past ten years; also, since such series started since the Internet went mainstream, statistics will be more readily available.

MotoGP and the European Le Mans Series seemed to be the best choices given these general guidelines.  Both have short histories, both are among the most prominent series I have not covered, and will probably be popular with my audience.

Although I am not planning to archive anything after LMS for a while, what other series do you think I should consider?