The Truth About the Pick Three

At one time, the goal of most racetrack gamblers — if they had a goal at all — was to pick winners. Read any handicapping book published before Al Gore invented the Internet (when a “cell phone” was still a phone used by prison inmates) and you’ll see what I mean.

In the original “Betting Thoroughbreds,” for example, Steve Davidowitz discusses Beyer speed figures — then produced by the man himself and not the Daily Racing Form — by relating how they affected his success rate.

“In the first season, the results said 176 winners in 310 picks. Fifty-three percent,” Davidowitz wrote. “A flat bet profit in all categories.”

Nowhere in his book does Davidowitz mention what that flat-bet profit, or return on investment (ROI), was. That’s because, back then, it didn’t matter. The assumption was that if one could “out-pick” the crowd, one would win.

Two decades later, Andrew Beyer realized how wrong this premise had been.

“At Gulfstream Park in 1990, all of my worst fears about the nature of the pari-mutuel competition materialized,” the dean of speed-figure handicapping wrote. “My speed figures — which had now advanced into the computer age — had never been so refined. I was watching races, assessing individual horses’ efforts, and detecting track biases more astutely than ever before. I was handling my money and my emotions with skill and maturity — a great step forward from the time I punched a hole through the wall of Gulfstream’s press box in a rage over an unjust disqualification. I felt that I was at the very top of my game as a gambler. And still I couldn’t win. My lack of success was due not to bad luck or photo finishes or any of the other traumas that plague all horseplayers. My frustration was best demonstrated by some of the winners I picked — by horses like Memorable Skier.”

Beyer went on to say that Memorable Skier had finished out of the money (worse than third) in all of his nine career races against maiden competition and was now facing winners, which, under normal circumstances, would make the horse an instant toss for most handicappers.

“But his speed figures were competitive and, in his last start, [Memorable Skier] had been forced to race wide on a track with a strong rail-favoring bias,” Beyer explained. “Now he was running again on a day when the rail was an advantage, and he had drawn post position 5, with four slow-breaking horses inside him. I concluded that the maiden would be able to drop to the rail and lead all the way. When I went to the track that day, prepared to make a killing, I thought Memorable Skier embodied all of the handicapping skills I had spent a lifetime learning.

“The race went just as I expected,” Beyer went on. “Memorable Skier popped out of the gate, angled to the rail, led all the way to win by six lengths — and paid $6.20. A pitiful $6.20.

“Even at a track heavily populated by tourists and retirees, the betting public had become smart and well-informed,” Beyer concluded.

While I don’t necessarily believe that the crowd had suddenly become “smart and well-informed” (I think that was always the case to a certain degree), it is noteworthy that Beyer was complaining about Memorable Skier’s price after the race had been run (and he had presumably bet). Outside of his reference to making “a killing,” he gave no indication as what his expectations were before the race.

Fast forward another 20 years and little has changed in the minds of some handicappers. Despite declining field sizes, arguably fewer quality races, drugs (both legal and illegal) and a host of other new variables to muddle the handicapping picture, many folks analyze races today like they did when Sheena Easton won a Grammy for Best New Artist — they focus entirely on picking the winner.

Of course, many realize, as Beyer did, that prices are not what they once were on standout horses, particularly those with a speed or class edge. Enter the horizontal wager: the pick-3, pick-4, pick-5 and pick-6.

Because these types of bets focus almost entirely on picking the race winner, bettors flock to them like foam fingers to Miley Cyrus. But are these kinds of bets such a great deal? Personally, I have my doubts, so I decided to conduct a little test.

I looked at my database stats for Saratoga during the first week of August in an effort to determine whether or not pick-3 wagering, in particular, made more sense (and more dollars, for that matter) than straight win betting.

First, because so many pick-three bettors are looking to beat the favorite in one or more of the legs, I checked to see how the post-time favorite in pick-3 sequences fared over the studied period (Aug. 1-5):

Favorites: 49

Winners: 12

Win Rate: 24.5%

ROI: -36.43%

Next, I determined what the average $2 pick-3 returned:

Avg. Pick-3 Payoff (46 sequences): $353.78

Then, I computed what a simple $2 parlay on the same races would have produced:

Avg. Three-Race Parlay (46 sequences, same races as above): $295.90

“Ah-ha,” I can hear some of you exclaim. “You see, you see (you’re a very excitable lot), the pick-3 offers better value than a parlay” (which is typically the argument for pick-3 wagering).

OK, that’s true; has been for many years. This is due to the fact that the pick-3 pool extracts takeout and breakage, albeit at a higher rate, just once, whereas each race in a three-race parlay is subject to the track and government’s wrath.

What this viewpoint ignores is something that economists refer to as “opportunity costs,” or the foregone benefits of other options. And I would argue that those benefits, in many cases, are better than those offered by the pick-three.

Permit me to explain: Let’s say that you like Horse A, at, say, 4-5 odds, in a particular race. But because of the short price, you decide to bet a pick-3 singling Horse A in the opening leg. Assuming that Horse A wins, what kind of a return do you need to beat the resulting $3.60 payoff?

Many of you will say $3.61… and many of you would be wrong. A $3.61 payoff assumes a single $2 bet and a 100 percent strike rate in each of the two remaining legs of the pick-3… which could only be guaranteed if one were to cover all the possible combinations.

So, for the sake of argument, let’s pretend there are eight horses in each of the last two legs of our mythical pick-3. Assuming that we hit the “ALL” button, our ticket cost now stands at $128 ($2 x 1 x 8 x 8). Hence, we would need a pick-3 payoff of $230.40 to match (not exceed) what a corresponding $128 win bet on Horse A would have returned.

This is not good news considering that the three pick-3’s in my study that led off with an odds-on winner returned an average of $166.47. Granted, a three-race sample is hardly proof of concept, but it’s not encouraging either.

“But Derek,” I hear some of you complaining, “nobody hits the ‘ALL’ button on two legs, especially with a big favorite leading off.”
A valid point, but keep in mind: The minute one starts limiting the pick-3 combinations, one’s winning percentage goes down. So what started as a high percentage play on a 4-5 shot now becomes a much lower-percentage one.

And that uncertainty must be taken into account if one is playing for profit and not just the gratification of having a winning ticket.

Are You Ready For Some Football?

With the kickoff of the NFL season on Thursday, I thought I would re-post a video I did for Youbet.com a few years ago.