Equation Racing writes:
"The Winning Equation at Charlotte"
Posted by Ol Don on May 29, 2011
Viewed 463 times
Well... on my last post, I talked a little bit about how I use numbers to predict winners at each track. There's a bit of fancy math and statistical modeling, so, I won't bore you with that spiel again. BUT, I will talk a little bit about the numbers that are important at Charlotte vs. the other tracks.
First, starting position is big. It is highly correlated with average position and, most importantly, finish. On some tracks this is not the case, in particular those tracks with lots of wrecks and those that run long. Interesting, the 600 is certainly one that runs long. But the numbers don't lie and indicate that -- for instance -- a driver starting in the top 5 has a 50% chance for a top 10 finish and a 30% chance for a top 3. Not bad if you're getting long shot odds in Vegas for dark horses like Keselowski and 'Dinger.
Second, momentum is a big factor in predicting success at the Coca-Cola 600. This may seem obvious, but I think the more interesting question is: how do you define "momentum"? Is it how the driver did last week... or the last 3 weeks... or in practices... or qualifying? It is not something that exists as a set statistic, so I had to run the numbers to see which of these -- or combo of these -- actually makes the biggest difference. I won't disclose the formula, but the influence of recent races is huge coming in to Charlotte. In this regard, attractive drivers to win are Kyle, Carl and Denny.
Lastly, and probably the most intuitively, site rating is important in Charlotte. This metric really encompasses average running position, laps led, etc and sums up a driver's historical success on the track. In this regard, you have to consider Kyle and Jimmie as favorites.
So... with all these considerations, a formula must be developed to predict winners. Some drivers perform well in none, one, two or even all three of these areas. The model I built calculates -- or tries to calculate -- the exact equation for success.
Here's who the model likes this week (in general), along with some personal commentary on the rankings.
Top tier is Kyle Bush and Jimmie Johnson. There is a pretty sharp drop off after that, but my personal opinion is that Edwards and Hamlin belong in that top tier, with Denny perhaps being the most intriguing dark horse to win.
The next tier has Gordon, Logano, Burton, Kenseth and Mark Martin. In this group, I think Gordon is overvalued. I think the two stable bets for top 10 finishes here are Burton and Martin, with Kenseth having the most upside and chance to win.
After that, you have Biffle, Kahne, Ruetimann, Bowyer and Kurt Bush. This is a high risk, high reward group. One of these guys could certainly make a push based on their skills and proven ability to win (in particular Kahne and Kurt). But guessing which one is anyone's call. At gun point I'd take Kurt Bush but would rather take my chances with guys like Kenseth, Hamlin or even Burton.
There are 3 drivers that just seem to fall way down the predicted finishing list. They are Jr, Harvick and Stewart. It's pretty surprising to see these guys so low -- in particular Harvick -- but if the numbers are right, these are guys that should struggle tonight. And by struggle I mean finding a top 10 or even top 15 finish.
With that, I'll introduce a new "segment" of sorts to my columns. I will give my picks for three different fantasy/betting formats.
1. Safe bets and good value long(er) shots to win
Safe = Kyle Bush, Jimmie Johnson, Carl Edwards
Good value = Kenseth, Hamlin, Burton, Martin
2. Yahoo style fantasy picks:
Group A: Edwards, Kenseth
Group B: Burton, Martin, Truex Jr, Kahne
Group C: Ragan, Labonte
3. Elimination style format where you pick one driver each week (max of using each driver 2 times all season):
Good luck to all and let's go night racin!!
Opinions expressed in blogs are those of the individual bloggers and do not necessarily represent the views of racing-reference.info.