Beating the Chalk

Each time I go to Vegas I stop by the Gambler's Book Shop (which I think might be no more) to see what they have in stock that's racing related. Some of what they carry is out of date, which, as far as I am concerned is all the better. Just like we don't want to bet what others are betting, we don't want to be using angles that others are using, too.

I once picked up a book, which shall remain nameless, and I ran all their hot "angles" through a database. All the angles - not some, but all - were ROI negative, despite having the expected good impact values.  Some were of an angle variety which we all love and yearn for: Here's how you beat the chalk.

It's great to beat the chalk. Not only are you getting a great payoff, but it feels good. We outsmarted the crowd! Unfortunately, it is one of the most difficult, if not the most difficult things to do in horse racing, in an ROI positive fashion.

When you try and subset data to beat the chalk, it shows just how hard it is.

Beware the chalk off extended layoffs! Well, here is this year by favorite, by days since last start, non first time starters (click to enlarge):

 Horses off greater than 300 days, who are the favorite, win at 37.72% with an ROI of 0.91. This is well ahead of horses who have recency in their profiles. If you try to beat the chalk that way, you're road kill.

Maybe we don't want to bet chalk if they are from lower ranked connections? Beware!

I guess there's something to it, but if you try and pitch out the 4th or 5th ranks, you aren't making much headway - their ROI is at or near the top connections.

Jcapper has a neat algorithm for horses who look ugly - no works for awhile, terrible form, ugly, negative-angle horses. They win at 4.6% of the time. What do they do as a favorite? A 0.81 ROI and a 29% win percentage. When ugly horses filled with negative angle after negative angle payout what any other horse does when they're chalk, it's a mighty tough gig. 

Most angles you often hear to beat the chalk tend to be just like most angles: A fast way to the poorhouse.

Here are a few things I look for when I want to fade a chalk, and they all have a common theme:

• Perfect trip beaten favorite.
• Post parade inspection. Does the horse look off, is he or she less keen and happy? 
• Cold trainer. Jamie Ness has run off the board with a few lately. Why bet him at 4-5 if he's cold?
• A hot "Replay" horse. He was checked and would've won for fun. But everyone saw it. These are the types who are even money and have a 35% chance to win.

I only find profitable fades over time when I go "off book", with tiny subsets. That works for me.

What works for you?