2015 IndyCar Season Preview Data

I used 2 main types of data for the 2015 IndyCar season preview posts to assist with analyzing which drivers will do well in 2015.  One data type were each drivers' average starting position and average finishing position.  The other was a predicted average finish rank for each driver.

Average Starting Position and Finishing Position
After gathering start and finish position data for all drivers at each race going back to 2012, I classified each track as one of the following:

• Oval - a track with left turns only
• Road course - a track with left and right turns that is a race track all year
• Street course - a temporary track with left and right turns that spends the majority of the year as public roads or an airport 

It seems noteworthy that I classified the Edmonton Airport as a street track instead of a road course because Racing Reference seems to classify it as a road course.

I then removed all drivers who were not on the entry list for the 2015 season opener at St. Petersburg.  This means drivers (like Dario Franchitti) who are retired were not factored into the rankings.  Of course, Franchitti's many good qualifying and race performances were impressive and difficult to achieve, but they are not relevant to determining how 2015's drivers compare against one another.

Last, I calculated several averages based on the data.  I calculated average start and finish positions for each driver at each track type during each year (for example, Scott Dixon's average oval finish position in 2012, and Marco Andretti's average street course start position in 2014).  This data is summarized in the table at the beginning of each drivers' season preview.

I also calculated each drivers' average start and finish position across all relevant years for each track type (for example, Takuma Sato's average finish on ovals from 2012 to 2014) and across all track types (for example, Sebastien Bourdais' average starting position on all tracks from 2012 to 2014).  I then ranked each driver's average start and finish positions.  For these rankings, I did not factor in how many seasons a driver had competed.  Juan Pablo Montoya has the best average oval finish for all drivers since 2012 despite the fact that he only raced in 2014.  It should be noted that Simona de Silvestro's 2012 season with the dreadful Lotus engines was excluded from these calculations, but her 2013 season (with Chevrolet engines) was used.

Predicted 2015 Average Finish
I used a very simple method to attempt to predict each drivers' average finish in 2015.  I determined what percent of the scheduled races took place at each track type (39% oval, 33% road course, 28% street course).  I then weighted each drivers' average finish for each track type from 2012 to 2014 by the percent of scheduled races that will take place at the corresponding track type.

Consider Helio Castroneves, whose average finishing positions on ovals, road courses, and street courses are 6.5, 10.0, and 8.5 respectively.  Castroneves' predicted average finish in 2015 is (6.5*39%) + (10.0*33%) + (8.5*28%) = 8.1.

This method is admittedly quite crude.  I suspect that many other factors influence finishing position, like qualifying position and a driver's team.  I've decided to use a small model for simplicity's sake, but later on I'd love to investigate whether other factors influence a driver's finishing position.  Another limitation with this model is that it doesn't factor in a driver's improvement.  Will Power performed much better on ovals in 2014 than he did in 2012 or 2013, but this model weights each year's performance equally.

It's also important to note that the driver with the best average finish will not necessarily win the championship.  This is because the relationship between championship points awarded and finishing position is not linear.  A driver who finishes 2nd in 2 races earns 80 points, while a driver who wins and finishes 3rd in 2 races earns 85 points (50 points for the win and 35 points for 3rd).  Opinions may vary on how points should be awarded among the various finishing positions, but here I'm only concerned with how they are awarded.

Despite its limitations, I believe this model provides an easy way to compare drivers against one another.  It likely hasn't revealed anything new to an veteran IndyCar fan, but I believe that attempting to quantify how a driver will perform is a useful exercise.

If you want to get in touch, feel free to leave a comment or send an email to spottersstand@gmail.com.  Follow the blog on Twitter (@spotstandblog) for reminders about new posts.