A Funny 'Cide' Thing - by Walter Seip
A funny thing happened during the running of the 129th Kentucky Derby. The "thing" is not that the supposedly superior Empire Maker was beaten by the double-digit longshot Funny Cide. The funny "thing" is that the results were predictable. Read on.
Recall Affirmed running and winning all three of the 1978 Triple Crown races. You may even recollect that Alydar ran second in each race. Obviously, Affirmed was faster than Alydar. But was he? Perhaps Funny Cide is faster than Empire Maker. But is he? After reading the rest of this article you will conclude that Alydar was the faster horse (and that he deserved higher speed ratings than those published). You will also be able to make an informed decision about when a funny "thing" will happen at your favorite track.
The competition between the two great three year olds of 1978 documents the funny "thing" which when ignored is a serious disadvantage to horseplayers. Speed rating adjustments (reward/penalty) are suggested to identify this handicapping oversight and turn it into an advantage. These adjustments confirm the superiority of your selection and uncover overlay winners and higher odds horses to use in exotic combinations. The obvious results are an improved return on investment.
Some questions: Are outside post positions a disadvantage to a horse? Does this apparent disadvantage more correctly result from how far a horse is from the rail when it gets to the turn? What is the disadvantage of racing wide in a turn? Can this disadvantage be predicted by pace and running style and then quantified in speed rating points? And, how can knowledge of this disadvantage be turned into a betting advantage?
HIGH SCHOOL MATHRemember high school math and solving problems relating to lengths of sides of a triangle? Recall the Pythagorean Theorem that allows that the length of the side opposite a 90 degree ("right") angle is equal to the square root of the sum of the square of the sides making that right angle.
Now visualize a right triangle with the side making the right angle the distance from the center of post position one (PP1) of the starting gate to the first turn (usually somewhat over 1000 feet). The other side is the distance from the center of PP1 to post position 5 (PP5). The distance between each post position of a starting gate is 3.5 feet (I’ve measured it). Therefore, the length of this side of the triangle is 3.5 feet times 4 (the distance from the middle of PP1 to the middle of PP5) or 14 feet. Thus the distance to the turn for the PP5 horse is the square root of the sum of the square of the distance to the first turn from the center of PP1 plus the square of 14 feet.
A small survey of several tracks determined that the distance from the starting gate to the turn ranged from a reported 1320 feet at Hollywood to 2010 feet at Emerald Downs. Solving the equation for PP5 at Hollywood and at Emerald Downs yields the distances of 1320.07 feet and 2010.05 feet respectively. PP12 is 38.5 (11 times 3.5) feet from PP1. The distances to the turn for PP12 are 1320.56 feet (Hol) and 2010.37 (Emd) feet. Although highly unlikely at either track, PP16’s distances are 1321.04 and 2010.68. Clearly; regardless of post position number (PPN), there is hardly more than a foot of difference (at Hol -- 1320.00 - 1321.04 or 1.04 feet and at Emd -- 2010.00 - 2010.68 or 0.68 feet) in the distance any horse travels when racing to the turn. Disadvantage? The numbers don’t support any. Still, horseplayers insist that a horse with a higher (outside) PPN is at a disadvantage.
High school math can not only identify, but seemingly quantify the horse’s very real disadvantage. Recall that the circumference (the distance around) of a circle is equal to two times the radius (the distance from its center to the outside edge) of a circle multiplied by the constant 3.14159 (known as "pi"). Affirmed and Alydar ran joined at the hip for the final 7/8ths of a mile during the 1978 Belmont Stakes. This included the final turn (half the circumference of a circle). Assume that 3.5 feet (recall the starting gate) was the distance between the centers of the two horses. The radius of Alydar’s circle on the outside was this 3.5 feet longer than that of Affirmed on the rail.
Use any radius you choose and the result will always be that the outside horse has to run 3.5 feet times pi or about 11 feet further than the inside horse. Alydar had to run these 11 "extra" feet to remain even with Affirmed when both horses exited the turn and began their historic stretch drive. If the distance between the center of this pair had been 4.5 feet, then the "extra" distance would have been 14 feet. Three horses side by side require the outside horse to run 25 feet further than the inside horse. Assume that the "length" of a thoroughbred is about eight and a half feet. Then the "first over" outside horse is at a disadvantage of 1.3 to almost 1.7 lengths to the inside horse and a disadvantage of 3 lengths or more for the three wide ("3w" in the chart’s comments) horse.
ALYDAR WAS FASTERWhat does this have to do with Affirmed and Alydar? The tapes of their 1978 Triple Crown races will confirm that Alydar was "first over" for at least one turn during each race. The Kentucky Derby charts state that Alydar "commenced to advance from the outside after six furlongs, continued wide in the stretch, swerved to bump ...." The Preakness charts note that Alydar "advanced willingly outside of horses in backstretch, engaged Affirmed well out from the rail...." Alydar ran at least 3 times the 11 extra feet further (perhaps twice this number) than the rail hugging Affirmed. The charts document that Affirmed took a total of 6:22 2/5 to run the three races and that Alydar’s total losing distance was less than two lengths. Do the math (speed equals distance divided by time). With the times being all but identical and Alydar covering much more than 4 lengths further, Alydar was the faster horse. As with all horses that do not go "wire to wire" on the rail, Alydar’s true speed ratings were reported lower than earned as the total distance he ran was ignored. Using low speed ratings places a horseplayer at a very serious disadvantage.
EMPIRE MAKER FASTER?The summary of the running of the 129th Kentucky Derby notes that both Funny Cide and Empire Maker were both a little wide in the early going with Funny Cide then edging nearer the inside while "Empire Maker ... was six or seven wide during the early stages, ... continued five wide on the backstretch and into the far turn, advanced steadily five wide leaving the three-eights pole... and wasn’t good enough." Consider the five wide (backstretch -- far turn -- to the three-eights pole) as 55 "extra" feet. These six "extra" lengths confirmed that Empire Maker was more than good enough to win, but he didn’t -- and for many it was a disappointment.
DISAPPOINTMENT PREDICTEDEmpire Maker’s lack of early pace and an outside post position in a large field suggested that his speed ratings should be reduced because of the strong likelihood that he would be forced to go around slower horses. This predictable reduction resulted in a slight speed rating advantage for Funny Cide and with his higher odds he became the selection. When the race was run, Empire Maker’s disadvantage was even worse than predicted and he "wasn’t good enough." If the 2003 Belmont Stakes has a small field, Empire Maker will be "good enough." Can this disadvantage be turned into an advantage (can a funny "thing" be identified)? It can, especially for the handicapper willing to adjust reported speed ratings to get an edge on those that do not.. For better or worse, reported speed ratings are based solely on final times. This time is compared to the best time run at the track at that distance within the last three years. The difference in fifths of a second is then subtracted from 100 (example: 3 seconds is 15 fifths producing a speed rating of 85). This figure usually is accompanied by a subjectively determined "variant" that accounts for the condition of the running surface. Other arguably more accurate methods compare the final time to an objectively determined "par" for that particular type of race.
SIMPLY ADJUSTSurely these speed ratings can be adjusted to a number that would more accurately reflect the real speed of a horse (such as Alydar). The easiest approach is to simply add (reward) or subtract (penalize) speed rating points to those you presently use. Basic mathematics has quantified some useable figures. Above it was suggested to add 3 points for a 3w horse (a 4w horse would receive a 4 point reward). A caution -- "fanned out" or "swung" 4w does not justify the full 4 points as the horse was not 4w for the entire turn. Since the horse was probably 2 or 3 wide to allow it to fan out or swing 4w or even 5w, a reward of 2 or 3 speed rating points is fully justified.
Unfortunately, not all comment lines include information translatable to speed rating point adjustments. Shedding light on or perhaps further complicating the reward/penalty issue are the legitimate considerations of total number of runners, numbers of expected front runners, a horse’s PPN (recall the comments relative to going wide into the turn), and the horse’s running style.
RUNNING STYLESThere are handicapping services that classify a horse’s running style as front runner, those close up, those that usually close to contend, or a combination of two styles. Some services provide "pace" figures. (More on these later in this article.) A somewhat skilled handicapper could make these classifications and approximate pace figures by examining past performance data such as times and beaten lengths at selected points ("fractional calls") in a race. Obviously, horses behind others going into a turn must go around them -- usually. Sometimes they find a "hole" in the middle of the pack and sometimes that hole is on the rail. If the horse is to win or at least to contend, percentages can be estimated regarding the manner of how this could happen. Perhaps on the dirt the horse has to go around (wide) about 65% of the time, between horses about 15% of the time, and up the rail 20% of the time. For turf races in might be 70% around (wide), 10% between, and 20% up the rail. Make your own percentages and adjust.
PACEPace figures provided, usually at a cost, by handicapping services resemble speed rating numbers. Some provide pace ratings at several points of call throughout a race. If not included, beaten lengths must be deducted, usually on a 1:1 basis (that is, 2 lengths out of the lead reduces a pace rating by 2 points). By comparing the first call pace number of today’s runners, front runners (based on times and not position) can easily be identified. By comparing all first call pace figures to the highest pace figure, horses that will probably be within say 3 points (lengths) in a sprint and 5 in a route would be labeled front runners. Closers are those horses with pace numbers more than 5 points lower than the highest number in a sprint and 7 lower in a route.
WHAT IFSLet's suppose today’s horse was the only front runner in the race. Like Affirmed, it has the shortest distance to run as it will surely be in front and on the rail for the entire race. It must be rewarded for this circumstance by adding as much as 2 points to its speed rating in a sprint and about 1 point in a route. If several front runners (less that 25% of the number of runners) are expected, this reward (given to all) should be cut in half. When many front runners are expected (25% - 50% of the runners), a penalty of almost a point should be assessed to each. Somewhere it has been reported that if a two horse speed duel develops, 44% of the time one of the dueling horses will win the race. So over half the time, neither one of them does.
Skipping past the "close up" runners to the "closers", penalties are suggested (focusing first on sprint races) based on the number of expected closers. For a small field, the penalty is small enough to be ignored. For fields of 5 to 8 horses, a penalty of 0.5 is suggested. For nine to 12 runners, a 0.8 penalty; and for fields greater than 12, a 1.3 point penalty.Closers in a route race incur more of a penalty for going wide. In a short field, 1.6 points is appropriate. For fields of five to eight, about 1.9 points; and any field exceeding 9, a 2.1 point penalty.
More severe penalties were calculated and would not be inappropriate. These suggested penalties incorporated the probabilities presented earlier and considered both turns of a route.
SUMMARYThe serious disadvantage to a horse drawing a high post position has been shown not to be the distance to the turn, but rather the high probability of being wide into and through the turn. This disadvantage was quantified and a reward/penalty approach to changing speed rating numbers to the advantage of the horseplayer was advanced. Both the simplified "reward for wide" and the "what ifs" approaches produce seemingly small changes to a speed rating. But a couple of points equates to a couple of lengths and to wins at nice prices as other bettors neglect the quantifiable dynamics of early speed and going wide -- a funny "thing."
AIN’T COMPUTERS WONDERFULA computer is a great tool for comparing horses in any given race. Very sophisticated algorithms can be developed to assign reward/penalty points to a horse’s speed rating based on post position, number of runners and their pace. A simple task for a computer is to "normalize" reported speed ratings for weight carried in previous as well as in today’s race. Using consensual weighting and involved lookup tables, two turn reward/penalty adjustments can be instantly incorporated into a speed rating and combined with other selected factors. There are handicapping services that employ these techniques for a fee. This writer developed and started using such an algorithm about two years ago and has stayed ahead of the horses...just barely!
-- Walter Seip is a retired US Army Colonel, is an engineer with advanced degrees, and lives in Las Vegas where he has won several local handicapping tournaments and hit the board in many others. BRIS (Brisnet.com) has been his source of handicapping data since 1998.
@copyright, SEIP Ventures May 2003. All Rights Reserved.
[ Sign-up ] [ Products & Prices ] [ Bookstore ] [ Track Info ]
[ Library ] [ Tech Support ] [ F.A.Q. ] [ Member Services ] [ Contact Us ]