A new kind of experimental rating

by Brett Sturman

For decades, harness racing giant Stan Bergstein made predictions as to which 3-year-olds would be the sport’s fastest based on how the crop raced the previous year as freshmen. These predictions came to be known as the Experimental Ratings, and the ratings were an annual industry favorite right up until the time of Mr. Bergstein’s passing in 2011.

Attempting to predict the success of 3-year-olds in advance of the upcoming season almost seems sacrilegious following the work of Mr. Bergstein, but it’s something whose time has come. The problem with statistical measures today in harness racing is that they’re largely outdated.

For example, the USTA maintains a publicly available Average Earnings Index for sires and broodmare sires, but that doesn’t quite tell a complete story. What nothing until now considers is the strength of the sire relative to the family of the dam it’s been bred to, and conversely the strength of the dam’s side relative to the sire. It becomes a complex circular reference. Fortunately, modern analytics allows the ability to figure all of this out.

In 2017, American Ideal and Bettor’s Delight showed a near identical average of earnings for all starters. At least specific to the top speed of 3-year-olds, which is what this analysis entailed, American Ideal rated more than twice as high as Bettor’s Delight and I’d contend is the more likely sire between them to produce superior levels of speed.

Although Bettor’s Delight recently produced the 1:47.2 winner Betting Line, he’s been aided by the fact that six of his fastest 12 starters over the past three years were out of dams sired by Western Hanover – a highly-rated broodmare sire by these calculations. Conversely, American Ideal has recently produced a number of fast 3-year-olds including Dude’s The Man and In The Arsenal, while doing it with a far lesser calculating caliber of broodmare sires.

When predicting which horses will be at the top of the 3-year-old pacing ranks in terms of fastest record, there are first some basic statistics to work with. Everyone knows that top 2-year-old speed doesn’t always equate to top 3-year-old speed, but the rate at which horses fail to develop their freshman season speed may be a surprise. There are 182 2-year-olds that appear on the USTA’s Top Performers list from 2014 through 2016 and of those 182, only 58 appeared on the same list as a 3-year-old. What this tells us is for all of the freshman pacers there were in 2017, less than half of them will be that fast (relative to other horses in the division) the next year.

Although less than a third of starters could replicate their high level of speed between ages two and three, starters from certain sires fared better than others. And it’s no surprise that starters from Somebeachsomewhere exceeded their starters from their sire counterparts. Of the 31 starters from Somebeachsomewhere that made the Top Performers list for fastest 2-year-olds, 14 of them turned the trick as a 3-year-old; far better than the overall average. Similarly, other sires have exceled in getting their starters to progress from 2 to 3, including Well Said (4 out of 9) and Dragon Again (an incredible 5 for 5, led by Fear The Dragon and Wakizashi Hanover).

While only simple math yielded these statistics, they give some validation to what a more complex analysis for 3-year-old shows. Through this modelling, Well Said came out with the highest rating for sires to produce top speed in 3-year-olds; even higher than Somebeachsomewhere. On the surface, it may not make sense – especially when compared to an average earnings index – but the underlying data supports it.

Like the American Ideal/Bettor’s Delight example, the broodmare sires connected to Well Said rate less than that of fellow Hanover sire Somebeachsomewhere. With regards to top potential speed, Well Said has done well for himself. Control The Moment was a fast 2-year-old and obviously was fast at 3, winning the Meadowlands Pace in 1:48.2. He also produced Lyon’s Snyder (1:48.3), Boogie Shuffle who paced in 1:48.2 over the 5/8th’s mile track at Harrah’s Philadelphia, as well as multiple other sub-1:50 3-year-olds; all with a smaller number of horses to work with than Somebeachsomewhere.

The interesting thing is when doing the same type of modelling and it’s restricted to top 2-year-old speed, Well Said rates below Somebeachsomewhere as well as also rating below sires Western Ideal, Dragon Again, Art Major, and others. But Well Said really improves them from ages two to three. In fact, it’s partially the reason the reason why I think that the unbeaten 2-year-old multiple PASS winner Done Well (a Brian Brown trainee) has a chance to be one of the fastest 3-year-olds this upcoming year.

This same type of logic holds true for the calculation of broodmare sires, although the broodmare sire ratings carry only half the weight of that of the sire in this modelling. The reason being that since the broodmare sires are a generation removed from the sire, their genetic value is worth only half as much as a starters sire (or dam).

One of the top-rated broodmare sires for producing the highest levels of speed by my calculations is Western Hanover. While he’s teamed up with super sire Somebeachsomewhere to produce Huntsville, Pure Country, Filibuster Hanover, etc., he’s also been paired with lesser rated sires such as Yankee Cruiser, Stonebridge Regal and Cam’s Card Shark to produce 3-year-olds in recent years with sub-1:50 speed.

A horse that I predict will be one of the fastest this year will be I’m A Big Deal; a colt trained by Chris Ryder and by Somebeachsomewhere. In addition to being by the Beach, he’s out of the speedy filly Big McDeal, who in turn is by Mcardle. In a limited dataset, Mcardle only had one horse represented by him as a Top Performer and it was Boston Red Rocks. Boston Red Rocks was the fastest starter from sire Rocknroll Hanover, which makes Mcardles influence that much worthier.

Based on the analysis as described along with some personal recalibration thrown in to account for intangibles that can’t simply be picked up on pedigree analysis alone, here are my predictions for the top 3-year-old pacers of 2018 (in Stan Bergstein format):

Stay Hungry (Somebeachsomewhere) 1:50.2 – 1:47.1
Hayden Hanover (Somebeachsomewhere) 1:50 – 1:47.4
Hitman Hill (American Ideal) 1:50.4 – 1:48.1
Youaremycandygirl (f) (American Ideal) 1:50 – 1:48.2

Done Well (Well Said) 1:51.2 – 1:48.3
Springsteen (Rock N Roll Heaven) 1:50.4 – 1:48.4

I’m A Big Deal (Somebeachsomewhere) 1:51.3 – 1:48.4

Pedro Hanover (Somebeachsomewhere) 1:51 – 1:49

Drew The Dragon (Dragon Again) 1:51.4 (h) – 1:49.1

Band Stand (A Rocknroll Dance) 1:51.4 – 1:49.2

Nutcracker Sweet (Bettor’s Delight) 1:50.2 – 1:49.3

Grand Teton (A Rocknroll Dance) 1:51.4 – 1:49.3
Jan (Rock N Roll Heaven) 1:51.4 – 1:49.4

— statistical information from the U.S. Trotting Association