I began handicapping NASCAR Sprint Cup races to improve my results in fantasy NASCAR, but over time I grew interested in betting NASCAR as well. As I learned more about sports betting and line shopped each race, I started to get a feel for the odds the sportsbooks considered break-even, but I didn’t have any hard evidence my hunch was correct.

After last season, I realized I had accumulated several years of published race picks I could use to at least identify my odds of picking a race winner and a top three finisher. Over the past couple weeks, I set out to do just that.

I decided to include only the picks I’ve published in this blog, which stretches back to early in the 2010 season. I also decided to ignore my picks from the restrictor plate races because there’s simply too much randomness in those races—I would never bet them anyway, so why include them in my study? Finally, I decided to also exclude the races for which I produced only early rankings and didn’t follow up to update the picks after practice and qualifying.

That gave me 85 races to work with, which produced these results:

• Number of wins by my top picks: 19
• Number of top three finishes by my top picks: 29

OK, so my top picks scored 19 wins out of 85 races. That’s a 22.3% success rate for win bets, and when you convert that rate to odds, you get 3.48–1. My top picks should therefore eke out a small profit at 4–1, but I think that edge is too small. I can’t see betting to win with less than 5–1.

Several sportsbooks also offer a top three finish bet (a “show” bet in the race horse world). My 29/85 result works out to a 34.1% success rate for show bets, which converts to 1.91–1 odds. That tells me 2–1 odds are break-even and not worth the time or risk. 3–1 might entice me to bet my top pick to show.

Why do I want to pad my odds? Due to the enormous uncertainty in NASCAR racing, I’m not sure 85 races is enough to provide rock-solid statistical relevance. Consider this: I’ve refined my handicapping model a great deal since 2010, but my top picks’ results over the past three years include wild swings. Check this out:

• Number of wins by top pick in 2010: 10
• Number of wins by top pick in 2011: 1
• Number of wins by top pick in 2012: 8

Does this mean I was a better handicapper in 2010? Nope. It simply reveals the crazy hot and cold streaks a NASCAR handicapping can expect to face over time.

Bottom line: When your money is on the line in NASCAR betting, caution is the word of the day!