## Sunday, August 30, 2009

### The Best Simple Metric for Predicting the full NASCAR points standings

Forget, for a second, everything you know about the current NASCAR points system: 185 points for a win, 34 points for last place, 5 points for leading a lap and 5 points for leading the most laps.

Imagine if the NASCAR points system consisted only of this:

1 point for finishing the race
1 point for finishing on the lead lap
1 point for finishing in the top 10
1 point for winning the race

In this system, you would get 0 points for not finishing and 4 points for winning. How relevant a points system would this create? Who would the champion be in this system?

What if I told you that in seven of the last eight seasons, the champion in the traditional NASCAR points system was the same as the driver in my fictional system?

For our purposes, let’s call this the Watermill Score. Ignoring the reset of points caused by the Chase, in every single Chase year the driver who scored the most total points all season was the same driver who had the highest Watermill Score. The year 2002 is the only exception because of a rare closeness in the competition, where seven drivers were within 226 points of the champion.

In 2009, for example, the top three under this system are Tony Stewart, Jimmie Johnson and Jeff Gordon. They also are the top three in the traditional standings.
In fact, going back every single year this decade, there is a .98 correlation between the Watermill Score and the traditional NASCAR points system, not just at the top of the standings, but for every spot down the line.

In every single year, the Watermill Score is more correlated than any other individual metric of Top 10s, Top 5s, Most Wins, Lead Lap Finishes, Racing at the Finish, Average Finish, Average Start, etc. There is no single other metric that is a better estimator of the points standings.

Why is this interesting? And why is this important?

Despite the complication of the current system, and all of the talk that it needs to be changed, this revelation is interesting because it tells you that the essential factors that go into winning a championship are finishing the race, finishing on the lead lap, getting Top 10s and getting wins. Two of the four points you receive just for being consistent and the other two points you receive for finishing up front. The points system is half-based on consistency and half-based on running up front.

It's important because it suggests a new way to approach race strategy. If you are a crew chief who is thinking about gambling for fuel at the end of the race, think about this math:

If you do not gamble, you are guaranteed a Top 10 finish, which is a guaranteed 3 points in this system. But if you do gamble, and you try to take the win, the risk is that you will fall off the lead lap. Your reward in this case is 4 points; your risk is going home with 1 point. Assuming a 50/50 chance of running out of gas, the average of that is 2.5 points. You would need an 80 percent chance of saving fuel all the way to get back to a 3-point-average scenario. And would you really want to risk that? Is that a good trade?

So when you simplify a situation like this, you can make a quick guess as to what to do. Come in now for the gas, take your guaranteed Top 10 and go home; the guaranteed 3 points is better than anything else. And I have just proved to you that if you can win based on this points system, you will also win in the real NASCAR points system.

Using the Watermill Score as a guide, crew chiefs will take the guaranteed Top 10. They will take the 3 points per race because NO DRIVER THIS DECADE has finished a season above a 3-points-per-race average. Tony Stewart has 66 points in 2009, which is a 2.75-points-per-race average. So whenever you can get a guaranteed Top 10, you should take it every time. The gambles are not worth it in the long run.

By simplifying the points system to this easy checklist, crew chiefs can better come up with strategies during the race without getting caught up in how other drivers are doing, where they might finish the race, etc. Just focus on these four categories (finishing the race, finishing on the lead lap, getting a Top 10, winning) and everything else will work itself out.

Also note the importance of finishing on the lead lap. It is one-fourth of the entire points system here. It is just as important as a Top 10 or winning. Again, if you are having a bad day, if you can still finish on the lead lap, you will earn yourself 2 points. Think about what that means for crews making adjustments to bad cars. Do you swing for the fences on a big adjustment, another gamble, to try to get the car way up in the front, but carry the risk of going the wrong way and sending your car to the back? Or do you focus on smaller adjustments you know will work, even if it means riding around the back of the lead lap?

As we see in this system, the answer is to go for the guaranteed small adjustment. First priority is not losing that lap. If your car is not a Top 10 car this weekend, then focus on finishing on the lead lap, and again the points will take care of themselves.