The first Saturday in May is less than two weeks away so it’s time for the casual horse racing fan to consider which three-year-old will win the Kentucky Derby
(NASDAQ:CHDN) in Louisville. It is certainly possible to enjoy the Derby for the sport and the beauty alone, without bourbon (See also: The Whiskey Breath of Wall Street: What the Invention of Bourbon Has to Do With Credit Derivatives
) or betting, but who would want to?
As I discussed last year
, most Derby horses are known quantities. The expert handicappers and insiders make it very difficult for a casual amateur to find a bet with a positive expected value. This is not the case for lesser horseraces, for which careful data analysis turns up many profitable strategies, even without the benefit of expert or inside knowledge. But that’s work, and Derby Day is for fun. The best hope for the non-professional, non-expert, or non-insider is to find a darkhorse, a horse going off at long odds (that is, a horse deemed to have a small probability of winning that will pay its backers a large multiple of their stakes in the unlikely event that it crosses the finish line first), not because everyone knows of reasons it will not win, but because no one knows of reasons that it might win.
This year, the Derby changed the selection rules to make it harder to find darkhorses. The old rules ranked horses based on the amount of money won in “graded” races (that is, races of sufficiently high quality). But grade 3 (lowest quality) counted the same as grade 1 (highest quality), and money won in seven months before the Derby counted the same as money won in the weeks leading up to the race. Because Derby runners are so young, that seven months can make a huge difference. Therefore, every year there were horses running in the Derby that had not faced Derby-quality competition, nor had received significant handicapper attention, nor had run major races at distances relevant to the Derby, nor had significant recent performances to judge.
Under the new rules, horses get points from finishing fourth or better in 36 designated races. That eliminated about 150 lesser races from consideration, forcing Derby contenders to prove their mettle in the biggest races against other Derby hopefuls. Many shorter-distance races were removed from consideration as well. 86% of the points are awarded in the ten weeks leading up to the big race and 56% are awarded in the last five weeks. Handicappers can concentrate attention on these races so they get high-quality, up-to-date observations of Derby candidates in direct contention with each other.
But all is not lost for the darkhorse fancier; there are still opportunities to back contenders of unknown quality, horses for which you have some hope of using your skills to find a positive edge. I started out by computing the record of all Derby contenders against other Derby contenders in Derby-eligible races (only one of the 36 qualifying races has yet to be run, The Derby Trial on April 27).
Current betting favorite Verranzano has beaten four probable Derby runners (Falling Sky, Java’s War, Normandy Invasion, and Vyjack) and lost to none. Revolutionary (won over Code West, Golden Soul, Mylute, and Palace Malice) and Orb (outran Frac Daddy and Itsmyluckyday) are also undefeated. Unsurprisingly, these are three of the five favorite horses at the moment.
Goldencents (with 5% owned by Rick Pitino, coach of the NCAA champion and hometown Louisville Cardinal men’s basketball team) is also among the top five, winning over Itsmyluckyday, Mylute, Super Ninety Nine, and Tiz a Minister, but he burned out in the San Filipe Stakes after a blistering pace opening duel. He ended up finishing fourth, behind Derby candidate Tiz a Minister. He is also of interest because if he wins, it will be under the first black jockey to win the Derby since 1902. In the 19th century and very early 20th century, many of the top jockeys were black, but as in baseball, racism would force them out of the top levels of competition.
The only other one-loss horse is Overanalyze (won races over Falling Sky, Frac Daddy, Normandy Invasion, and Oxbow, but lost to Vyjack). He is considered a longshot at the moment at 18-1, but things change. Normandy Invasion is the fifth favorite at 7-1 odds currently, despite a one win and four loss record versus other Derby contenders in qualifying races. The two in-between horses, neither favorites nor longshots, are Vyjack at 10-1 odds with a one win and two loss record, and Itsmyluckyday at 12-1 with a two win and three loss record.
All of these horses have been thoroughly examined with plenty of relevant race data to analyze. It’s hard to believe that any of them will leave the starting gate at odds mispriced enough to overcome the track or bookmaker’s percentage. If any of them do, it will not be a casual fan who figures that out.
Next consider the records of the horses beaten by the non-longshots. In these records, I show win, loss, and tie. Win means the horse finished among the top three, and ahead of another Derby runner. Loss means the horse finished behind another Derby runner that was among the top three finishers in the race. Tie means both horses were fourth or worse, in which case I don’t consider the relative order very meaningful. Java’s War was 3 wins, 2 losses, and 1 tie; Palace Malice was 5-4-2; Frac Daddy was 4-4-0; Mylute was 4-4-3; Oxbow was 2-5-4; Super Ninety Nine was 2-3-0; Falling Sky was 0-4-1; Code West was 5-2-2; Tiz a Minister was 1-2-0; and Golden Soul was 0-5-5. Handicappers have observed all of these horses extensively against the Derby favorites and each other. The records are not necessarily predictive of Derby performance; Normandy Invasion at 1-4-0 is considered about fifteen times as likely to win as Code West at 5-2-2. Generally speaking, the favorites have done better in the qualifying races than the longshots, but experts look at far more than the raw win/loss totals. They also consider the importance of the race, quality of the running, quality of the other horses, nature of the race, track type and conditions, breeding, training, jockey, and other factors.
We are left with six possible Derby entrants that are not favorites, but have never lost to one. They are all considered longshots at odds from 22-1 to 70-1, but there is considerably more uncertainty about their quality than with the horses mentioned above. Among this group, there is some reasonable hope that you can uncover an underpriced horse.
The favorite among the darkhorses is Will Take Charge at 22-1. This horse qualified for the Derby by winning the Rebel Stakes over lightly-regarded Derby runners Super Ninety Nine and Oxbow, and the Smarty Jones Stakes with no other Derby contenders in the race. He was an also-ran in the Kentucky Jockey Club Stakes and Southwest Stakes, finishing behind Frac Daddy and Super Ninety Nine, and in the same crowd as Java’s War. This adds up to a 2-2-1 record, but none of these are exactly the horses to beat if you want to win the Kentucky Derby. Will the smell of the roses bring out the champion in this not-yet-challenged three-year-old? It’s happened before.
Next is Lines of Battle at 30-1 -- in my opinion, the most intriguing darkhorse in 2013. This Irish-trained horse was unimpressive in one previous trip to the US and qualified for the Derby by running in only one qualifying race, the United Arab Emirates Derby in Dubai. He won that race, but none of the other runners will enter the Kentucky Derby. The pace was slow, and the artificial track is nothing like the dirt at Churchill Downs (although the 1,900 meter distance is closer to the Derby length than any of the other qualifying races). There’s not much for the experts to chew over with Lines of Battle, so they are reduced to speculating based on breeding and workouts, and neither are reliable predictors of success.
Govenor [sic] Charlie has also not raced against other Derby contenders in qualifying races, and is currently listed at 35-1. He did not race at all as a two-year-old (not since Apollo in 1882 has such a horse won the Derby), and qualified for the Derby by winning his third start ever. That was the Sunland Derby in New Mexico, and he won in style by five lengths, setting the track record in the process. No other Derby contenders were in the race. Is he a one-race-wonder with one minor league win? Or is he the fastest three-year-old in the country? Probably the former, but no one will know until May 4.
At 50-1, you have your pick of two untested horses: Black Onyx and Winning Cause. These horses qualified by winning the Horseshoe Casino Cincinnati Racing Spiral Stakes and Lexington Stakes, respectively. Neither horse entered another Derby qualifying race, and no other Derby contenders were in either race. Finally, you can bet on Charming Kitten at 70-1. This horse finished third in the Bluegrass stakes, behind Java’s War and Palace Malice, for an 0-2-0 record versus Derby contenders, but 0-0-0 versus likely Derby winners.
These darkhorses are not the longest of the longshots; that distinction belongs to horses that have given ample evidence they cannot beat top Derby runners. Unless you are an expert or an insider, betting on them is like buying a lottery ticket, and you pay about the same expected 50% loss. Betting on favorites is more like a 5% expected loss. If you can find a second-favorite at short odds (which can’t happen this year unless there is a seismic shift in sentiment without significant data) or a short odds horse getting significantly shorter immediately before the race, you might even get a zero expectation bet, without actually doing any work.
I have no reason to believe that betting on a random darkhorse gives better than random expectation. As far as I know, no systematic study has been done on the topic. But I do believe with darkhorses you have little disadvantage compared to experts. There is not enough information to determine the probability beyond some general statistical analysis available to anyone. Betting on a darkhorse can be a true exercise of judgment, a game you could win, something an intelligent person can take pleasure in doing (unlike, in my opinion, playing casino games or lotteries at known unfavorable odds). It’s not really betting on a horse; it’s betting on yourself, something I think should be a regular habit for everyone, at least in some arena.
So place your bets, take your seat (hopefully at the track, but failing that, at a television screen), pour your drink of choice (just no sugar and whiskey abominations), and let the best horse win, to the profit of the best bettor.
Also see: Dark Horses, Long Shots, and Kentucky Derby Betting