Wednesday, February 27, 2013

Measuring Danica Patrick after 59 Races

Question & Data
Let's try to assess Danica's Sprint Cup prospects by comparing the start of her career with other notable drivers. Which drivers have had performances that most closely resemble hers?

She has run very few Cup races, so let's use her much larger Nationwide dataset to form our predictions. Danica has competed in 59 Nationwide races. We can compare that to the first 59 Nationwide races run by several other drivers. This will help us forecast what to expect from her.

As a comparison set, let's consider the entire top 15 in 2012 Sprint Cup points, plus 5 other drivers with varying success (such as Regan Smith and Josh Wise) and IndyCar background (like Sam Hornish).

The Chart
Figure 1
What Does This Chart Mean?
Figure 1 is a box-and-whisker plot of the performance of 21 drivers after their first 59 Nationwide races. Each driver's median finish is the little red line. Half of a driver's finishes fall inside the blue box. A driver's best finish is the right-most point on the chart.

The chart is sorted so the drivers with the best median finish are at top. Notice how the little red line moves to the right as we go down the driver list. Out of 59 races, that red line is their 30th best finish (right in the middle of 59 races).

The Good Sign for Danica
Danica's median finish is 19th, which compares poorly with the group overall, but is in similar territory with Kasey Kahne and Tony Stewart, who became stars.

Remember that sometimes the median finish can be misleading, as you see Sam Hornish and David Stremme up reasonably high on the chart. Obviously they were not able to convert that success into productive Cup careers.

According to this chart, Danica has her work cut out to be a true star, but the examples set by Stewart and Kahne (former open-wheelers like Danica), show that there are potential paths to success.

The Concerning Sign for Danica
Past research (to be posted later) suggests the best predictor of future success is the ability win races early in one's career (not top 5s or top 10s, but wins only). It's easier for a driver to build consistency later if they have proven their ability to win early on.

Let's consider that point when looking at the best finish for these same drivers in their first 59 Nationwide races:

1 Brad Keselowski
1 Carl Edwards
1 Clint Bowyer
1 Dale Earnhardt, Jr.
1 Denny Hamlin
1 Greg Biffle
1 Jeff Gordon
1 Jimmie Johnson
1 Kasey Kahne
1 Kevin Harvick
1 Kyle Busch
1 Martin Truex, Jr.
1 Matt Kenseth
1 Ryan Newman
1 Sam Hornish, Jr.
1 Tony Stewart
2 David Stremme
3 David Ragan
4 Danica Patrick
5 Josh Wise
13 Regan Smith

Notice the big difference here in drivers who won a race compared to the drivers who didn't. The drivers in bold (Stremme, Ragan, Wise and Smith) have not made an impact in Cup. The concern for Patrick is that without a single win in her first 59 Nationwide races, she could find herself with unimpressive Cup results, similar to these four drivers.

When you look at the two main factors above: median finish and early-career wins, both are stacked against Danica, statistically suggesting she will not be a future Chase-caliber driver. Hopefully she can prove this analysis wrong. If she finds a way to on-track success, she will have created a new path that past statistics would not have expected.