Advent of Alpha Day 17: Form Calibration to Markets

Newspapers that want to give the impression of data density like to print a form line, with the most recent result at the right, going back to the left. On race cards “-” indicates a break to the last season and “/” indicates a break to previous seasons before that.

On a race card you might see:

U4/301-211

This indicates the horse won the last two times out, came second before that and last season had a 3rd place before not coming in the top 9 and then on their last race of the season finishes the season with a win. The previous seasons had an unseated rider before a 4th place.

Football papers might have something like this:

LLWWDW

Same deal, but with Wins, Draws and Losses.

There is statistical significance to form lines: horses that consistently win, tend to win more often in future than those that don’t. Obvious. Maths is easy here.

However, we don’t care about winners, we care about value, and these form lines don’t indicate value, which just needs you (I think), to start thinking about that past performance against what the market expected to happen.

Let’s suppose we had two form lines: results (as we have now), vs market expectation. Let’s take our two examples, the horse first:

Result:   U4/301-211
Expected: 01/134-422

See how a different story is told? What about the football?

Result:   LLWWDW
Expected: DDLLWW

What’s the story now? Compared to the form line alone, you can see a story arc puts into context the market in front of you today.

In horse racing there is a cliched back and forth about this. Someone will say “it won the last three times out”, or “it won its last by 7 lengths”, to which a reasonable reply might come “but what did it beat?”. Some people try and over-complicate this and identify a strength of form line: well, X beat Y and Y has beat Z and X is now racing Z so should beat it. OK, but is that reasonable?

Is it not more interesting to think about how the last three times out the result was better than the market had predicted? Or considerably worse? I think it points to the fact that a horse/team/player is consistently poorly assessed by the betting public, and that should be priced in.

Most of your counter-parties are just looking at result form lines. If you go just one layer below into looking at result vs expected, you’re beating a ton of people who are taking a view.

But in 25 years of following racing, I’ve never heard of anybody who looks at what the market looked like on previous events, in any real depth.

The closest we have to this today is the beaten favourite marker on some cards, indicating that beaten favourites will do “better” next time because they were beaten last time - I’ve not done the data analysis to understand if that’s true, but it sounds like BS. Even if it is a BF, I’d want to quantify it - was a 3.0 fav in a 20 runner handicap on heavy ground, or a 1.1 on good to soft against 4 other horses? Those are two very different things.