We never tire of saying it: The Belmont Stakes is a race unlike any other run in the United States.

It is run at the distance of a mile and a half on the dirt course, yet it is run around only two turns. Much is made of the distance of the Kentucky Derby being the farthest the horses have run up until that point, and in nearly all cases this is true. The horses in the Derby enter after running in races at a mile and a sixteenth or a mile and an eighth. The horses are running at least one furlong farther than ever before. But in the Belmont Stakes, the horses that ran in the Derby will now be required to travel an extra two furlongs, and some of the expected starters will be traveling three furlongs farther than ever before.

Speed and pace figures are viewed with skepticism by many when unusual distances come into play, and not without good reason. Just as a speed figure at five or six furlongs isn’t an ideal predictor of a horse’s performance at a mile, prior pace and speed figures aren’t always as predictive at the exceptional distance of the Belmont Stakes. Tactics become magnified in importance–jockeys will make errors on an oval that is different than any other in the country. (See Spectacular Bid’s Belmont for a controversial case study.) However, a study of the past decade of the race does reveal some interesting trends.

The average winning TimeformUS Speed Figure for the Belmont Stakes the past decade is 111. It is not a race that is likely to be stolen on the front end; despite a memorable longshot like Da’ Tara, the Belmont distance, we believe, is proving less kind to speed than shorter dirt races.

Nine of the 10 horses that have won the Belmont Stakes in the last ten years had run a lifetime-best TimeformUS Speed Figure in one of its last two starts. The other, Union Rags, had done so three starts back.

However, that doesn’t help us in 2014. Each one of the probable entrants has run a top in one of the last two races.

Looking for more clues from prior winners: The average top TimeformUS Speed Figure for the Belmont winner prior to the race was 110. Birdstone in 2004 had the lowest lifetime-best number coming into the race, a 99 TimeformUS Speed Figure. Ruler On Ice had the second-lowest at 105. Using 99 as a cutoff for contenders this year eliminates Commissioner, Matuszak, and Matterhorn. If the cutoff is bumped to 105, Medal Count and Tonalist can be added to the list.

Another clue to the Belmont winner is in how energy is rationed in prior races, and it is here that we have particular interest. All but Birdstone (who is the exception to so many rules) had run a final-time figure equal to or faster than their first-call pace figure in one of their last two races.

Let’s stop here. It’s a straightforward idea but it bears explaining:

Commonly in dirt racing, we see “race shapes” in which the first-call pace figure is faster than the corresponding final-time figure. Often enough, horses attending reasonably fast early paces manage to hold on to win, simply by putting away the other speed horses.

And in the last decade of Belmonts, the early pace call has tended to be faster than our final time figure–even in the one exception, 2007, the leaders faded and Rags to Riches and Curlin came from off the pace to fight it out to the wire. Usually, Belmont winners have shown the ability to win races by exceeding their early pace numbers with their final-time numbers. They can relax early, and they can exceed their early running in a more comprehensive effort.

And so, it is likely the pace will be too fast for the horses up front, and looking at this year’s Belmont contenders doesn’t give us reason to believe that the “race shape” will be any different than it’s recently been. The average first-call pace figure in the last decade of Belmonts is about 13 points faster than the final-time number.

But unlike in so many dirt races we see, disposing of the other early speed horses is typically not enough to win.

To be more specific, leaving aside outlier Birdstone, seven of the remaining nine had run this way in their last race–faster in final time than in early pace.

The other exceptions: Palace Malice’s notorious failed Derby blinkers-on experiment and Drosselmeyer’s Dwyer. Of those being considered for the race this year, Tonalist, Samraat, Wicked Strong, Medal Count, and General a Rod ran faster pace figures than final-time figures last out. Not good. Of those, Samraat and General a Rod did so in each of their last two races.

Using their speed early, and going slower late.

Rarely a good idea in the Belmont Stakes.

Of the Derby runners contending this Saturday, many had troubled trips, but neither Medal Count nor Samraat has shown a history of exceeding his early pace figures with his final-time figures.

Wicked Strong, on the other hand, has.

The Belmont, as mentioned, is a different race than a typical dirt contest in this country. Pace and speed figures are tricky at this distance, but do provide some valuable insights. The race won’t be stolen; the winner will deserve it. Several of the entrants show negative signs as detailed above.

Only three horses were not eliminated by some of the trends mentioned above: California Chrome, Ride On Curlin, and Commanding Curve. And, given his history, we think Wicked Strong fits with this bunch as well. Based on the past decade of pace and speed figures, chances are strong the winner will come from this group in 2014.