Major league baseball has been revolutionized by statistical
To varying degrees, teams have
recognized that they simply can’t be competitive without a thorough
understanding of the factors that go into winning baseball games.
Because many of those factors are subtle (and
surprising) enough that common sense and raw observation doesn’t detect or
measure them properly, virtually all teams now employee a number of experts in
statistical analysis whose sole job is to find areas where the team can gain a
This mirrors the direction which
successful businesses in many industries are headed, as “Moneyball” has taken
hold in all sorts of industries.
Ironically, the value provided by statistical analysis is
greatest in the industries where it is the least widespread. In baseball, the early adopters (such as the
well-publicized Oakland A’s) received the greatest benefits when they were the
only teams taking advantage of market inefficiencies. Now that everybody is doing it, they have to
work a lot harder for a much smaller edge.
Based on the lack of public discussion of statistical
analysis of pedigrees, I’d have to assume that it’s virtually non-existent in
thoroughbred horse racing. In baseball,
there was a twenty-five year stretch where amateurs with no connection to
baseball published a wide variety of insightful studies and analyses. Despite that, those within the sport remained
resistant to new ideas, and preferred to rely on traditional wisdom and ‘common
sense’ over statistical evidence and measurements. While I would expect faster adoption of any useful
findings in horse racing (due to the much greater number of people in decision
making positions in the sport), there still should be some degree of ‘internet
chatter’ generated. To date, there are
relatively few people talking about any kind of statistical analysis in the
sport, and most of what has been done consists of backward-looking data mining,
rather than true studies of predictive analytics.
Those who take the plunge of embracing analysis are going to
be rewarded with a wide variety of valuable learnings. How valuable?
That depends on the size of their investment in the sport. In 2012, there were 65 buyers of over $1
million of yearlings at public auctions.
For those 65, buyers, even a very modest 5% improvement in their return
on investment would have been worth a minimum of $50,000 for the year. Then consider all the purchases of weanlings,
two years, and broodmares, plus all of the private purchases of
thoroughbreds. Add in the farms that are
breeding their own races horses. In all,
must be many hundreds of owners operating on a scale where even a slight
improvement in their ability to predict racing success would provide them with
tens of thousands of dollars per year of value…in some cases hundreds of
thousands. And there are thousands of
people involved in the sport in various ways for whom it would be worth $100 or
more to have a better understanding of the factors that go into predicting the
success of a racehorse based on their pedigree.
It’s uncertain when racing’s statistical revolution will come, but it’s
going to be big business when it does.