Horses Don't Bounce

Of all the racing “truisms” that are, well, not true, few grate on me more than the concept of the “bounce.” Conceived by Len Ragozin, founder of the “Sheets,” the bounce theory asserts that a peak effort, as measured by a horse’s speed figures, is likely to lead to a regression of form, especially if the horse in question returns to the races too quickly.

No less an authority than Hall of Famer Bobby Frankel was a big proponent of the idea. In his formative years as a trainer, Frankel often speculated as to why his horses seemed to perform so randomly.

“I wondered why a horse would run so good one day and then come back and run so bad," he told Andrew Beyer of the Washington Post back in 2003.

Aided by Ragozin’s Sheets and his own belief that rest was the key to avoiding these dreaded form bounces, Frankel won over 3,600 races in his illustrious career, including the Belmont Stakes (Empire Maker, 2003) and the Breeders’ Cup Classic (Ghostzapper, 2004).

As a result, Frankel was among those who helped usher in the era of the “freshened” horse — and he wasn’t afraid to buck tradition in that regard.

The late conditioner (he passed away in 2009) was downright angry with Ramon Dominguez after the former champion jockey suggested that High Limit, who had just four career starts prior to competing in the 2005 Kentucky Derby for Frankel, “probably doesn’t have enough experience to compete at the highest level yet.”

“There might be a new rider on this horse,” Frankel informed the New York Post after seeing Dominguez’s comments in the Courier-Journal. “I haven’t made up my mind yet. I’ll give him a call. I mean, if he doesn’t have any confidence in High Limit… I don’t know. Most likely he’ll ride the horse, but I have to see whether he was quoted right.”

Frankel went on to say that he’d “had a lot of horses with no experience who went on to win big races.”

Unfortunately, High Limit wasn’t one of them, as he finished last in the Run for the Roses. However, the aforementioned Ghostzapper remains the only Horse of the Year in the history of the Eclipse Awards to have started fewer than five times during the calendar year in which he was honored (Main Sequence hopes to add his name to that short list).

ECLIPSE AWARD WINNERS: HORSE OF THE YEAR

2013 Wise Dan (7 starts)
2012 Wise Dan (6)
2011 Havre de Grace (7)
2010 Zenyatta (6)
2009 Rachel Alexandra (8)
2008 Curlin (7)
2007 Curlin (9)
2006 Invasor (5)
2005 Saint Liam (6)
2004 Ghostzapper (4)
2003 Mineshaft (9)
2002 Azeri (9)
2001 Point Given (7)
2000 Tiznow (9)
1999 Charismatic (10)
1998 Skip Away (9)
1997 Favorite Trick (8)
1995 Cigar (10)
1994 Holy Bull (10)
1993 Kotashaan (10)
1992 A.P. Indy (7)
1991 Black Tie Affair (10)
1990 Criminal Type (11)
1989 Sunday Silence (9)
1988 Alysheba (9)
1987 Ferdinand (10)
1986 Lady's Secret (15)
1985 Spend A Buck (7)
1984 John Henry (9)
1983 All Along (7)
1982 Conquistador Cielo (9)
1981 John Henry (10)
1980 Spectacular Bid (9)
1979 Affirmed (9)
1978 Affirmed (11)
1977 Seattle Slew (7)
1976 Forego (8)
1975 Forego (9)
1974 Forego (13)
1973 Secretariat (12)
1972 Secretariat (9)
1971 Ack Ack (8)

Still, as I’ve written on previous occasions, I think the “bounce” is more often a return to the mean or simply the result of different race dynamics than it is a physical phenomenon that can be cured by a little R&R.

To prove this, I devised a series of tests, utilizing the largest of my racing databases, comprised of over 14,300 races.

To start with, I looked for horses with at least 10 starts over the past two years that recorded a lifetime-best Brisnet speed figure in their most recent race, which must’ve been contested over today’s general track surface (dirt/synthetic or turf):

Number: 6,674
Winners: 1,015
Win Rate: 15.2%
Impact Value (IV): 1.16
Odds-Based Impact Value (OBIV): 0.81
In-the-Money Last (Bounce) Race: 69%
In-the-Money Today: 45%
Avg. Change in Finishing Position: -1.24

As the IV and OBIV indicate, there’s not much to infer from the overall numbers, so let’s break them down — first by days since last race. If Frankel and the “Sheets” guys are right, we should see a significant difference among horses returning to the races quickly and those given some time off:

RETURNING TO THE RACES IN 14 DAYS OR LESS
Number: 1,696
Winners: 297
Win Rate: 17.5%
IV: 1.29
OBIV: 0.92
In-the-Money Last (Bounce) Race: 67%
In-the-Money Today: 48%
Avg. Change in Finishing Position: -0.97

RETURNING TO THE RACES IN 15-29 DAYS
Number: 3,021
Winners: 441
Win Rate: 14.6%
IV: 1.12
OBIV: 0.78
In-the-Money Last (Bounce) Race: 73%
In-the-Money Today: 44%
Avg. Change in Finishing Position: -1.49

RETURNING TO THE RACES IN 30-59 DAYS
Number: 1,516
Winners: 222
Win Rate: 14.6%
IV: 1.14
OBIV: 0.78
In-the-Money Last (Bounce) Race: 65%
In-the-Money Today: 44%
Avg. Change in Finishing Position: -1.17

RETURNING TO THE RACES IN 60 DAYS OR LONGER
Number: 441
Winners: 55
Win Rate: 12.5%
IV: 0.99
OBIV: 0.72
In-the-Money Last (Bounce) Race: 53%
In-the-Money Today: 39%
Avg. Change in Finishing Position: -0.74

Well, well, well… these digits beautifully illustrate why most studies on the bounce effect are flawed. Just as Frankel asserted, horses coming off of a 60-day layoff or greater do, in fact, hold their form better than those returning to the track in a more timely fashion (0.74 average increase in finishing position). However, that form is typically not as good to start with as it is for horses coming back in less than 60 days.

Just 53 percent of the horses that returned to the racetrack in 60 days or greater finished in the money (third or better) in their last race — compared to 67 percent wheeling back in two weeks or less, 73 percent returning in 15-29 days and 65 percent coming back in 30-59 days.

Hence, from a betting perspective, “freshened” potential bounce horses represent terrible value, losing over 33 cents on every dollar wagered (0.72 OBIV). By contrast, potential bounce horses returning to the races in two weeks or less cost the bettor just 17 cents on every wagered dollar (0.98 OBIV) — about the same amount of red ink that favorites produce.

But I didn’t stop there. I wanted to test the premise that a bounce is primarily caused by a particularly good effort, e.g. the 2009 Woodward that many, including the great Steve Haskin, believe “gutted” Rachel Alexandra.

Thus, I looked for horses that not only recorded a personal-best Brisnet speed figure in their last race, but that also met or exceeded today’s par in that race.

Duck for cover, you’re about to witness another myth explode.

To summarize:

A) Horse recorded a lifetime-best Brisnet speed figure that meets or exceeds today’s par in its last race, which must’ve been contested over today’s general track surface (dirt/synthetic or turf).

B) Horse must’ve started at least 10 times over the past two years.

RETURNING TO THE RACES IN 14 DAYS OR LESS
Number: 1,051
Winners: 210
Win Rate: 21.0%
IV: 1.44
OBIV: 0.93
In-the-Money Last (Bounce) Race: 62%
In-the-Money Today: 52%
Avg. Change in Finishing Position: -0.50

RETURNING TO THE RACES IN 15-29 DAYS
Number: 1,894
Winners: 333
Win Rate: 17.6%
IV: 1.33
OBIV: 0.82
In-the-Money Last (Bounce) Race: 68%
In-the-Money Today: 49%
Avg. Change in Finishing Position: -1.00

RETURNING TO THE RACES IN 30-59 DAYS
Number: 1,113
Winners: 181
Win Rate: 16.3%
IV: 1.25
OBIV: 0.80
In-the-Money Last (Bounce) Race: 60%
In-the-Money Today: 46%
Avg. Change in Finishing Position: -0.75

RETURNING TO THE RACES IN 60 DAYS OR LONGER
Number: 353
Winners: 47
Win Rate: 13.3%
IV: 1.05
OBIV: 0.72
In-the-Money Last (Bounce) Race: 46%
In-the-Money Today: 41%
Avg. Change in Finishing Position: -0.23

Notice that every single category outperformed its counterpart, completely destroying the notion that a superior performance — as gauged solely by speed figures, which is what the bounce theory is based on — leads to a subpar performance next time out.

In future tests, I’ll examine the impact of distance, age and running style, but, for now, I think it’s safe to say that the bounce theory has about as much merit as Willow Smith’s conception of time.

“I mean, time for me, I can make it go slow or fast, however I please, and that’s how I know it doesn’t exist.”
~ Willow Smith