Lecture 7: The Brain in Contemplation of Itself: A Comment on Research and Statistics- Part 2: Stats on the Hoof: The Bart

I made my first serious attempt at interpreting statistics in the late 1970s. It started with a visit to my local race track. It was here that I got the misguided notion that, with so much available information about the horses, I would be able to earn an income. It seemed to me that, with a little brain work and a little common sense, I could figure it out. As anyone with true common sense knows already, the entire notion was false. I would soon discover this through experimentation, parting with a few hundred dollars along the way, but no more! The idea was to get some fresh air and sunshine and to see what I could learn, not to lose scads of money. I had no background in formal statistics at the time.
For those who have never been to the track, the tools of the trade are a program and the Racing Form. The program, much as a program at any concert, will tell you who is playing or, in this case, who is running. It includes an estimate of the odds when the race starts. (Odds fluctuate until the time of the race based on the cash amount of bets tendered.) The program also tells you who owns the horse, the stable, and who the jockey is, although jockeys may also change in advance of the race. It tells you each horse’s sire and dam. It will also give information about the race itself, that is, if the horses that are running can be claimed for purchase and how much it would cost to buy such an animal or if it is a stakes race for a cash prize with no opportunity for purchase.
By contrast, the Racing Form will tell you anything else you want to know about the horse… almost. This is where a naïve “investor” has the most opportunity to sucker himself with the available data. And this is where more than one would-be scientist has been born. The Form provides data on each horse’s history, where and when it ran, its position at every point in the race, the conditions of the track when it ran, who rode it etc. This is where one can test out theories, hypnosis by numbers I prefer to call it.
While it may be easy to dismiss this type of experience as useless, there are certain lessons that one may learn, some of which have nothing to do with horses. First, it is best to approach a set of data with a working hypothesis. This is no different than any other type of research—medical, psychological, actuarial, or whatever. If you don’t approach the Form with a theory you can lose track of what you are trying to discover. Approaching numbers with a theory is what scientific researchers do routinely. In fact, that is what they are trained to do. You put forth a research hypothesis, then determine how best to collect information and to interpret it. In this case, the information is collected for you. You just have to decide what kind of rational (or irrational grid) you want to put on it in order to draw your conclusions.
Where it applies to horses, I soon learned that there was not enough information available to me to allow me a reasonable guess as to which horse might win a race. For example, I developed hypotheses about whether a horse started fast, did better down the stretch, or was a good closer, hypotheses that were refuted almost as soon as the gate flew open. When my best reasoning was so easily refuted, I started wondering what the horse and jockey had for breakfast that morning, which side of the bed they got up on, whether they were in a good mood, were having a domestic squabble… or whether the horse may have been offered extra incentive that might affect the outcome.
I turned this thinking into a hypothesis during one trip to the track. On this outing, I inspected the horses before the race, albeit at a distance, with the notion that I might be able to tell how everyone was feeling and whether their mood might turn into a win. During one of the races, I saw a horse with his mouth taped, his tongue lolling out, and with a wild look in his eyes. I remember thinking how ill-bred and uncontrollable the horse looked. It seemed to me unlikely that a horse in this condition could possibly win. Of course, that horse won. And it was back to the drawing board.
At the time, I was working as a cook in a restaurant with counter service. It was around noon when a nicely dressed, older man sat down at the counter, then ordered a hamburger and a coffee. While he was waiting, he reached in his jacket pocket and pulled out a racing program and a pencil. His facial expression was thoughtful as he began scratching down numbers. I was particularly interested in the fact that he did not even have a Racing Form. This meant to me that he was not trying to figure anything out, but that he already knew what he was up to. I got up the courage to ask him what he was doing and he told me, “Parlaying my bets.” It didn’t take me long to turn that bit of information into the discovery that I was way out of my league if I thought I could win at the horses.
Parlaying, as I understood it, was to take one’s winnings from one bet and bet that money again, then continuing to turn any winnings into further bets. This presumes, of course, that you won some money in the first place and that you continued winning, which was something I had never done. This man was making that very assumption and he did it without any statistical data at all. Which led to my second lesson in statistics, which is, if you have direct experience of a phenomena, such that, you know how a thing operates, then you’ve got a leg up on anyone who is simply relying on numbers.
When the older man came back some time later for another hamburger and coffee, with another program open in front of him and with his pencil out, I asked him the two questions that were on my mind. First, whether he had won the last time I saw him, to which he answered that he had and quoted some figure in the thousands. Then I asked him what exactly a jockeys’ agent does. He explained that it was his job to arrange for horses to have jockeys and jockeys to have horses. That meant to me that he not only knew the horses and jockeys, but also the trainers and the owners. In short, he had all the information I could not get even with a program and Racing Form. Inside information, as we all know, is the best information, if reliable. (Shortly, I will describe how this affects issues of mental health.)
Still, with this new found knowledge, I decided to go to the track one more time just for fun. I recall that it was a cool Labor Day at Bay Meadows race track near San Francisco. It was getting on the end of the day and the feature race was about to go off. I think it was the Golden Gate Stakes (but, if it was called something else, please forgive me). I reviewed the Form to see how the field looked. Glancing up, I noticed that one of the horses was going off at less than even odds, meaning that, if the horse won, you might collect 75 cents for every dollar bet. I also noticed he only came in second in his last race. Thinking that I did not stand to win very much if I did bet on the favorite and that he did not win his last race anyway, I noticed another horse that had won his last few races. The odds on this particular horse were about 3 to 1, a better payday. Logic dictated that the second horse would be a better bet.
When the race started, my horse was ahead. Halfway through, he appeared to be about eight lengths ahead. Thinking I was doing pretty well, I asked the guy next to me why the horse running second posted less than even odds even though he lost his last race. The guy answered briefly, “That’s The Bart. He lost to John Henry by a nose.” “Oh,” I said, just as The Bart gave me my first real lesson in horse racing by taking off. He closed the eight lengths before I could breathe and won going away. It was common knowledge that he lost to John Henry, the same John Henry that had been Horse of the Year for two years running. It was common knowledge to everyone, but me. What I had never seen before nor did I appreciate until that moment was how exceptional a horse can be.
The Bart taught me my third lesson, that is, there are no statistics adequate to describe what is exceptional. I never saw John Henry run, but, considering what I had just witnessed, I could only imagine how fast he was. In a matter of seconds, I had learned to admire the speed and prowess of an exceptional horse, but I never went to a horse race or bet on a horse again.