Murphy Thinking Outside the Box

[Wandering Eyes]

“It's invisible,” says Rob Murphy. “And sometimes it moves. Sometimes it's an oval; sometimes an oblong with no corners.”

On one level, then, there's nothing there at all. Yet those, for many years, were the parameters of his working life. As a relief pitcher for eight MLB teams between 1985 and 1995, all his focus was on that void behind the batter, and the elusive margins between strike and ball.

Little wonder, then, if another great passion of Murphy's life is all about giving definition to something most of us consider hopelessly amorphous: namely, what makes some Thoroughbreds superior to others.

“What if we typed into some AI platform, 'Who will win the Kentucky Derby?'” he suggests. “There's a correct answer. I can give it you right now. 'A 3-year-old Thoroughbred with four legs and a tail.' Right? Okay, so how many of those do we have? Twenty-thousand. So it's a question-and-answer, but it's not the right question. And what I've learned, as AI has started to permeate, is that you have to ask the right questions. But when I look at what I was doing in the 1980s? Well, those were the right questions.”

He was ahead of the game in more ways than one; became an object curiosity, indeed, as the guy who hauled a computer into the bullpen itself, honing his analytics on batters. And that was an era when you almost needed the physique of a professional athlete for this kit to be in any sense “portable.” As a southpaw, in fact, he was always careful to use his right arm to carry his laptop, which came in at 30 pounds.

He waves his phone and chuckles. “It was 9.4 hertz,” he says. “So a millionth of the processing power we now carry around in our pockets. A millionth!”

It was three years before anyone else in baseball started using a computer. In the meantime, Murphy was also pioneering its use in another walk of sporting life.

Murphy's maternal grandfather, Frank Ashley, called races for over 30 years at Arlington Park, Washington Park and Hawthorne; while his dad's stepfather Freddie Smith rode for Colonel Bradley, winning the 1940 Preakness and Belmont on Bimelech. Growing up in Miami, however, his most immediate Turf stimulus was a string of racehorses his father kept at Calder.

It was at the lunch tables of the Turf Club there that Murphy began a 50-year quest to unravel an enigma. Simply as a handicapper, at first–but gradually all those grids, all those numbers and symbols, began to strike a deepening chord. For his, patently, is a mind with its own way of looking at things.    When the word SLEW (for his equine paragon) reversed out of a printer, for instance, few of us would have clocked, as he did straightaway, the latent alternative: M375. Hence his racing program, M375 Thoroughbreds.

Sure enough, even as an adolescent, Murphy discovered a peculiar aptitude for deciphering the data by which we interpret equine ability. He was just 17, in fact, when hitting a freak streak: the trifecta, six days running, with a three-horse box.

Getty Images

Closing day at Calder, they walked onto the Turf Club patio for the last–then the only trifecta race of the day.

“So who do you have?” asked his dad.

“Three, five and 10.”

“What!?”

The moderate favorite plus two longshots.

After the race his mom sighed, “Well, nobody's got that.”

He pulled out the ticket. “We do.”

It paid $9,800.

“I'm not even old enough to cash the tickets,” Murphy says. “My parents had sent me to a great Catholic high school, so when my dad returned from the windows I handed back five grand. 'That's for my school,' I said. 'And my car.'”

Everything, then, still remained sheer instinct. Murphy wouldn't be introduced to computer science–the first class in the state–until his senior year of high school.

“But you know what?” he remarks. “It was via numbers. It was the process, learning a system.”

That said, the class opened immediate horizons. Asked to volunteer a practical application, for grading, Murphy persuaded his teacher to let him handicap the Derby.

“The program outgrew the memory of the school's system,” he recalls. “The whole thing was still loaded on paper tape, an inch wide. But we ran the program and it said: Affirmed, Alydar, Believe It.”

Grade A, then–albeit Murphy had to settle for the race call over the PA between innings at the baseball district championship. For his other, more physical prowess was meanwhile creating opportunities of its own. By 1981, his college career at the University of Florida made him a first-round draft pick by the Cincinnati Reds. Mind you, he didn't break out of minor leagues until 1985. If baseball didn't happen, he wanted a contingency plan.

So patiently, painstakingly, he extrapolated his handicapping values into pedigrees. He pored over stallion registers and produce records. His first compendium totaled 12,000 pages; the second, 18,000. He read and annotated every entry.

For most of us, data in such volume would tell you so much that it tells you nothing.

“I know that good sires come in all different shapes and sizes,” Murphy accepts. “Good dams, the same: might be Grade I winners, might be unraced. But I'm a trends person. I see something, I make a model. Seattle Slew fits; then Northern Dancer fits too. So then it's, 'All right, how can I show a validated search-and-find process?'

“So next I started going through yearling catalogues. Over 200,000 pages, inputting the data. Then I put these parameters together and, in the sale that produced Seattle Slew, he was one of just two that popped positive. Who cares what the other horse was? You picked that one, right? Then Spectacular Bid, a couple of years later, his dam fits the model–even though she got her black type in a $12,500 stakes at Bay Meadows. Amazingly enough, Flying Paster's dam won that race. So it doesn't have to be the Alabama winner every time. Leslie's Lady got her black type at Hoosier. Arrogate's dam won hers at Sam Houston.”

The program Murphy constructed was isolating 1% of yearlings; of those, 20% proceeded to win stakes. Many multiples, in other words, of the industry average.

Then a teammate, seeing Murphy buried in pedigree books on the bus, introduced him to a buddy who owned horses. The guy was intrigued, and they agreed to target a juvenile sale at Hialeah in 1984.

Of 233 horses, Murphy's program approved only an Ack Ack colt. The physique checked out, and they bought him for $30,000. Artillerist won three of five juvenile starts; he ran second in stakes on the other two, and again when reappearing in a Grade II.

Then a single horse sieved out of a Washington-breds sale became a stakes winner. “So the real-life application's working,” Murphy says. “And I'm tightening the screws all the time. But then, yes, baseball happened. Eleven years of it happened.”

He was particularly talented at it; his ERA in 1986 was 0.72. In the history of Major League Baseball, no National League pitcher who has thrown more than 50 innings in a year has a lower one.

Nonetheless he maintained and tweaked his database throughout. There was no internet, of course, and he was constantly on the road. But he could get the Daily Racing Form in cities, there was BRISnet, he even used an early telex machine, with fast-fading printouts on thermal paper.

In August 1994, eight days after the Yankees plucked him from the Cardinals, and from last to first in the standings, the Major League Players' Association went on strike. The silver lining was that he could now go to the September Sale. With a couple of pals pitching in, Murphy shook two out of the catalogue: the colt won a graded stakes, the filly was stakes-placed before becoming a stakes producer.

Entering the stretch with baseball, then, Murphy knew he had something to fall back on. Next, a couple of mares: one produced a millionaire in Japan; the other, claimed for $5,000, came up with Platinum Tiara (Cozzene). Carrying the silks of M375 Thoroughbreds, she won a stakes by 11 lengths and was beaten half-a-length in the GI Breeders' Cup Juvenile Fillies.

“And of course we've got the mother, the brother, the sister,” Murphy says. “So out of that five-grand mare, we sold millions of dollars of horses. Just from throwing the right dart.”

Another bull's-eye was Grade II winner Golden Spikes (Seeking the Gold), at $250,000 his only pick at the 2006 September Sale. There was admittedly a period out of the game, to prioritize family, but Murphy has been regrouping since. In 2021, for instance, having missed the Saratoga Sale through illness, he saw that West Point had bought his top pick. He approached Terry Finley and was allowed into the City of Light colt we now know as graded stakes winner Battle of Normandy.

Obviously there's no single, simple formula; nor, equally, can Murphy be expected to divulge much even of its complexities. Clearly, however, he has devised a colander that retains a disproportionate percentage of superior horses.

“There's probably 75 or 80 data points,” he says. “The algorithm is 100-plus calculations. But it does say yes or no, there's no gray. Even with technology now, you can't just ask ChatGPT: 'Which horse should I pick out of this sale?' It's about how data is processed, how it's curated, whether it fits the pattern you're looking for.”

His focus is increasingly on dams.

“I've broken them down: stakes mares, winning mares, others,” he explains of his yearling sales research, dating back to 1975. “The numbers are the same across each group. And you wouldn't expect that. But it's finding a balance: mares with zero race record must be from better families, that kind of thing. There was an article recently saying how science has discovered that broodmares are best with their second through sixth foals. I knew that in 1984. Okay, Secretariat's dam was 19, so we're going to miss some. But that's something I always accepted, always been part of the model.”

Murphy tested three Saratoga Select Sales, between 2019 and 2022 (omitting the Covid year). From 649 horses, an initial filter produced a longlist of 89. But that 14% produced half the elite horses graduating from these auctions. Expensive, some of them–but not as expensive as the many duds released through the colander.

If mares widen the filter, compared with Murphy's original sire focus, they also widen opportunity: it's less about finding Nureyev at $1.3 million than his sister Fairy Bridge at $40,000. Remember that mares will gradually move from one column to another, with age, and Murphy incorporates a corresponding sliding scale.

But what he has always done so laboriously, longhand, plainly has scope for exponential acceleration through AI. Murphy has accordingly begun talking to potential partners about turbo-charging the process.

“As always, it's about relationships,” he says. “I've been putting in 10, 12 hours a day moving this forward: it's not labor, it's passion. Right now, I have parts of 25 horses. We still want to keep the quality high and the quantity low. But we're growing it back and looking forward to big things. I can't wait to see what the future brings.”

Talking about that invisible box, Murphy had used an instructive expression. “I've got the ball, he's got the bat,” he says. “I'm in charge.”

That's obviously not the case in racing, where so much is vested in those raising, training, riding your horse. “On the other hand, in baseball, second place is last!” Murphy says. “Second at the Breeders' Cup, they're going to write you a check. And you have black-type for the mother, brother, sister.”

However rare the athletic attributes that brought Murphy fame, they are harnessed to no less exceptional mentality.

“I was raised by the world's two greatest optimists,” he says gratefully. “If the word 'can't' ever came out of my mouth, as a kid, my dad would say, 'Do not say that.' So whether it's work harder, or read more, whatever it may be: I'm a 'can do' person.”

The post Murphy Thinking Outside the Box appeared first on TDN | Thoroughbred Daily News | Horse Racing News, Results and Video | Thoroughbred Breeding and Auctions.

View the full article