Secretariat Talk

Chick Anderson called the Belmont finish this way: "Here comes Secretariat to the wire .........., he hits the finish......25 lengths in front....". We are left with a call that in the aftermath of the race was largely forgotten. With the astonishment at the finishing time and the with the zeal at the call of what was to become the official margin of victory, Anderson's initial calculation faded. Anderson himself in summarizing the race concluded that estimating large victory margins is a difficult task, "it could conceivable have been larger [than 25]". Was Anderson, an experienced announcer, close to the truth? How was the official margin determined? Are there ways of probing the distance, of corroborating the official mark? Lets try.

On examining Secretariat's Belmont in slow motion using video equipment with a built-in timer calibrated at 30 frames per second, 6 frames per fifth, the following observations were recorded:

- Secretariat ran the Belmont in 2 minutes and 24 seconds as second place Twice A Prince (TAP) ran it in 2 minutes, 28.2 seconds;

- Secretariat ran the final quarter in 25 seconds as Twice A Prince completed it in 26.2;

- Secretariat clipped the wire at a rate near 52 feet per second (fps) while Twice-A-Prince in successfully winning the duel for second also clipped the finish at a speed near 52 fps;

(Please note that all VHS decimal times are expressed in fifths.)

Scenario 1
We will begin our examination with the third bullet, both win and place clipping the wire at about 52 fps. These figures were obtained using VHS technology equipped with a timer, the footage viewed in slow motion investigating multiple trials using differing feeds with varying visual qualities before settling on the speeds. The probe continues:

Using 52 fps as the measure, dividing this by 5 gives the feet covered in fifths per second, or 10.4 ft per fifth . Dividing this by 8 ft, the standard measure for a length, gives 1.3 lengths per fifth. This 1.3 applies to both Secretariat and Twice A Prince since they were both traveling at about the same speed at the finish. From the first bullet, given that Place finished 4.2 seconds behind Secretariat, expressing 4.2 in fifths gives 21. This figure multiplied by 1.3 gives 27.3, or the difference in lengths between First and Place. Accordingly, Secretariat finished approx. 27 lengths in front of Twice A Prince;

Scenario 2
Approaching the inquiry from a different angle, using data from the second bullet, we derive an average speed for Place on through the final quarter. Twice a Prince running the final quarter in 26.2 seconds gives an average speed in fps of 50.4. Dividing this by 5 gives 10.08 ft per fifth of a second. This figure divided by 8 gives 1.26, the average length the horse is traveling every fifth of a second down the stretch. The time elapsed between Secretariat's finishing time and that of TAP is 4.2 seconds or 21 fifths. The product of 21 and 1.26 is 26.46, the deficit in lengths TAP traveled in the wake of Secretariat's performance. This demonstrates that the slower the horse is moving, the closer it has to be to Secretariat in order to travel the final distance in 21 fifths. In this scenario, Secretariat's margin of victory is approx. 26.5 lengths;

These figures found proximate corroboration with a few independent on-line pieces, one found at http://www.horseracegame.com/community/content/blogs/stretchrun/27-05-2009/did-secretariat-really-win-belmont-stakes-31-lengths-sorry-we-do , entitled "Did Secretariat Really Win the Belmont Stakes by 31 Lengths? Sorry, We Don't Think So." In that analysis, the writer(s) estimated a 25 length victory. A second piece estimated the margin at under 25.

Some may say that the mark that matters is the margin established when the winner's nose clips the wire, not the time elapsed between this and when Place finishes. In response, the following image is telling: (double click image to view)



Scenario 3
This image taken from Secretariat.com displays Secretariat in the stretch inside the final 16th, an observation verified using various VHS resources. Using the same resources, it was found that the number of rail studs between the 16 poles is 41. Dividing the distance of a 16th, 330 ft, by this figure gives 8.05, the measure of a length. These studs can serve as a measuring instrument in estimating the margin of victory. It can be seen here that TAP and My Gallant are in heated competition just past the final 16 pole while Ron Turcotte looks towards the teletimer, perhaps in surprise at the fractions his colt is setting. Counting the rail studs between First and Place sums to an apparent 23. At this point, Secretariat's distance to the 16 pole in studs is 26.  Subtracting 26 from 41, the stud count between the poles, gives 15. The product of 15 and 8.05 gives 120.75, the distance in feet from Secretariat to the finish.  Using the average speeds for the final quarter derived from the second bullet, Secretariat, traveling at an average of 53fps will traverse this distance in about 2.3 seconds. If his counterparts are traveling at an average 50.4 fps., then Secretariat's velocity is about 2.6 fps faster. This figure, 2.6, multiplied by 2.3 seconds gives 5.9 ft or .74 of a length. At this rate, Secretariat will at most increase his distance over his nearest rival by about 3/4s of a length to the finish. Adding 3/4s of a length to 23 rounds to 24, so it is evident that Secretariat did not win by 31. A range starting at 24 with a ceiling at not more than 27 can be established. I believe Chic Anderson is vindicated, ......posthumously vindicated*.

This scenario, however, reveals discrepancies with the other two. A proper counting of the studs and thus the lengths from Secretariat to TAP sums to about 23. We count the misplaced studs as if they were placed in their proper position. If the time elapsed between Secretariat's finishing time and TAP's clip to the wire is in fact 4.2 seconds (and it is), then TAP's (and My Gallants) closing velocity would be a much slower 46 fps, and not the higher speeds derived above. To close the final distance of 24 lengths (approx. 193 ft ) in 4.2 seconds suggests momentum at Place and Show would have rapidly decayed in the final yards of the race. I don't believe this is accurate. On reviewing Belmont footage, TAP's final furlong time at 13.2 seconds, or 50 fps, was verified. This figure is less than the quarter average of 50.4 quoted above, but not significantly. Either the image above is not the final 16th (it is in fact the final 16th) or rail studs between Winner and Place are missing. Using ESPN Classics video footage, a good quality feed of all the horses passing the 3/16 pole on into the final eighth was found; at that point, again carefully counting studs placed Place approximately 23 to 24 lengths behind the leader. If that distance was 24, then the rate differential between Secretariat's and TAP's quarterly averages should have added close to 4 lengths through the final sixteenths, thus Secretariat leading by approximately 28 at the finish. If 23 is accurate, then Secretariat's final margin would have been closer to 27. This latter mark more closely coincides with the derivations in the first 2 scenarios. Presently, the discrepancies between the third scenario and the other two cannot be accounted for. A full range graphic of the final 16th would help resolve these questions.

So how was the official margin of victory established?  That has never been found. Some years back I contacted Belmont Park with this question, and never received a reply. The judges that day came up with the figure 31, but who were they? Are some of them still living today?

This leaves a few more questions: was Count Fleet's margin of victory really 25? What type of technology was used in 1943 to make that determination? Also, a quick highlight from Dorothy Our's 2006 work on Man O' War: in a footnote she addressed his 100 length victory in the 1920 Lawrence Realization and surmised that his actual margin of victory was probably just under 90 lengths (a furlong plus about 6 lengths which is equivalent to 88.5 lengths). Large margins are very difficult to establish.

This piece does not detract from Secretariat's finishing time which was the true marvel of the day; in fact it may prove to be a marvel for centuries to come (see my blog on Singularities).

*  Chic Anderson went to our Maker in March of 1979. He called America's last 3 Triple Crown winners.
**note: Secretariat, on average, opened at least 1 length for every 16th ran that day.

Secretariat , ManOWar, and Citation bring the best and worst out of us. There simply is no way to know with factual certainty which was finally the better. All the poles of so called experts are built on speculation, opinion, and fixed memories of particular performances that left impressions on observers. State culture plays a role as well, for instance, the adulation of ManOWar in Kentucky in distinction to that of Secretariat in New York. Contrast this to the following of Citation in California. Kentucky Horse Park (KHP)is a monument to ManOWar just as Belmont Park is the same to Secretariat. In a similar way the Blood-Horse publication out of Kentucky competes with the New York's Daily Racing Form (DRF) for communities. All are influenced by the cultures from which they arise. The old timers referenced in one list could easily be countered by a list of old timers in another. Fixed allegiances are difficult to break. For every Sylvester Veitch is a Holly Hughs, both Hall-of-Fame trainers who 'saw them all', yet both with differing opinions as to what they saw. For every Ed Bowen, writer and editor for Blood-Horse, is a Joe Hirsch, the late writer and chief-editor for the DRF, both of high distinction yet with differing opinions when it came to 'the best'. For every Phil Georgeff or Tommy Trotter, etc, can be found an Allen Jerkins, Hall-of-Fame (HOF) trainer, or an Eddie Arcaro, HOF jockey, who both had experiences and judgments countervailing the former. Add Charles Hatton (who 'saw them all'), Andy Beyer, handicapper and co-developer of the respected Beyer Figures, Steve Davidowitz, Teddy Cox, Mike Casale, Tim Capps, and others to that list. And finally, the Blood-Horse panel of seven which found both Secretariat and ManOWar tied for first, both receiving 3 votes, but differing by one at place giving ManOWar the edge. Hardly a consensus, especially when considering that one member placed Secretariat outside the top 10 (14^th), a determination whose judgment on a Thoroughbred that as a three year old set or equaled 6 track records, two of which were world marks, and was a mere 1/5th off of two more, both on hand rides and one in the mud; completed the Triple Crown in record time; in two seasons set or broke 8 top graded stakes records (7 in one season), five of which have never been broken; and set at least one record at all 6 distances ran between and including 8 to 12 furlongs (8, 8.5, 9, 9.5, 10, 12); should have been questioned and possibly discarded as an outlier. One can only guess at the employed criteria. Nevertheless, this panel brought into contention the earlier AP panels that decidedly awarded ManOWar the trophy with little discussion.

A question concerns the listing itself. How does one interpret it? Is the second place horse a nose off the first, a head, a half-length, a length or more? A pound lighter, 2 pounds or more? Is the horse in the last spot 100 lengths off the first, or 100 pounds lighter? Or does the list rank an overall career impression, both on and off the track? An overall contribution to the sport? If the former, then Timeform in 1999 asked the same questions and derived rankings based on computer models and imposts . If the latter, then it is difficult to say whether on the track horse 'x' is the better of the rest. Was Exterminator really ten places better than Sir Barton when in fact Barton defeated the former giving weight? Or were other factors given weight in giving Exterminator the edge?

This takes us to those commentators who assert that ManOWar faced competition on par with the runners of 1973. This is not correct. Taking a quote from 'A Century of Champions', a work by Randall and Morris who in 1999 transformed data compiled by Timeform modeling into book form, this text said in regard to ManOWar:"A more rational appraisal, in light of the undoubted increase in competitiveness of Stateside racing since his day, must place 'the horse' (ManOWar) on a slightly lower pedestal, but there remains no doubt of his greatness". This statement corroborates my work (www.truevine.net/~sec@truevine.net/Legends/) completed some years ago that found similar conclusions. In that work I found that the average finishing times of Kentucky Derby winners decreased by as much as 4 seconds between 1920 and 1973. Of those 4 seconds, 2 could be attributed to the horse alone, the remainder to all other factors, improved track surfaces, lighter equipment, training, etc.... The improved specimens can be traced not only to better nutrition but to influential blood lines imported into the states from Europe, lines which included but are not limited to the potent sires Bull Dog and Nasrullah (traceable to Nearco and Phalaris), and the highly influential broodmare sire Princiquillo. According to Randall and Morris American horses achieved parity with their European counterparts in the 1960s and forged ahead in the 70s in their golden age of racing.

The ease with which ManOWar dominated his fields cloaks the real miracle, the fact that this specimen that was generations ahead of his time came from a crop of only 1600, plus or minus a few. What are the odds that such a specimen would emerge from such a small crop? By way of his sire Fair Play, ManOWar introduced into the American stock a speed line that would strongly influence several decades. By way of his Dam Mahubah, ManOWar possessed a stamina he would pass to his daughters that they in turn would pass to their progeny well into the future. So the mixture of speed with stamina became central to the industry. As a sire, the horse all but singlehandedly improved an American stock that was well behind European standards; and it is true that at his time, ManOWar's competition was good yet modest, not only to him, but to future generations. ManOWar will have had far stiffer competition in the 1970s than in the 1920s had roles been reversed, and this despite better nutrition, tracks,and equipment. Some declared Sir Barton the better to Riva Ridge (RR), an observation taken from the Blood-Horse panel ranking which placed the former 9 places better. My work contradicted this. Using Regression Modeling to compare Riva Ridge's Kentucky Derby, a descendent of ManOWar, to Sir Barton's Derby performance, after making all necessary adjustments for extra-equine factors, Riva Ridge outperformed Sir Barton by approx 8 fifths, or about 8 to 10 lengths. A large difference between the two was RR's inability to handle off tracks. The horse sub-performed in the mud which preempted his 1972 Triple Crown run, performing badly in the Preakness on a wet track. The same condition may have short circuited his 1972 Woodward, finishing a distant fourth to Key To The Mint. However on a good track, Riva Ridge was formidable, equaling or setting 4 track records and breaking 2 world marks. His world record for a mile and three sixteenths still stands some 40 years later. In the Timeform listing of America's top 100, RR was imposted at 133 pounds; unfortunately neither Sir Barton nor Exterminator showed.

After adjustments for all factors other than the horse, my findings placed Secretariat ahead of ManOWar from close to 1 second at 10 furlongs to 1.6 seconds at 1.5 miles. In the comparison races at 1.5 miles, both were on hand rides for the distance. Timeform imposted Secretariat, Citation and Spectacular Bid ahead of ManOWar, 144 , 142,141, and 139 pounds respectively. Timeform also imposted Forego at 138 lbs, the Hall-of-Fame gelding both Secretariat and Sham defeated by 10.5 and 8.5 lengths respectively in the Kentucky Derby. One wonders how well Sham would have performed had he continued his racing career.

ManOWar left a strong impression on horsemen, both on and off the track. Probably no single horse performed as beautifully both on the track and in the shed, helping to rejuvenate a stock far less than its international counterparts. He was often referred to as the 'Super-horse', 'Wonder Horse', or the 'Miracle Horse' by journalists who recounted his performances in their Sunday chronicles. In communication with Steve Davidowitz, handicapper and racing historian, this author stated that those who recalled the horse referred to him as 'stout', and 'unforgiving'. The horse might well have provided the greatest influence on a single industry searching for a footing in a country going through an industrial revolution, wars, to a sporting world in rehabilitation. At the close of 1920, a remarkable sporting year that covered a summer Olympics, a Bill Tilden Wimbledon title in tennis, and Jack Dempsey victories in boxing, both ManOWar and Babe Ruth topped the list of the most influential sporting figures to populate the headlines. Some writers made a case for ManOWar as the most crucial. Crowning sobriquets such as the 'The Greatest of All Time', 'The Horse of the Age', and others clothed his ascent to royalty. But by the 40s, the balance began to tip. New blood-lines infusing crops that grew in size by factors of 6 to 15 began taking effect, producing greater performance lines which arguably peaked in the 60s and 70s. Stronger gene pools generated the leading performances of Secretariat, two additional triple crown winners and a host of HOF runners spurring observers to reference these decades as America's golden age of racing. Thus far, the question is unanswered as to whether a second golden age remains, ready to spawn a new generation of record breakers, bearers of a heightened speed and stamina that will lay the foundation for the new century. Breeders and owners work hopefully and diligently to bring this forward.

As ManOWar's 100^th birthday approaches, the stately monument marking his remains at KHP quietly overlooks the American horizon as it has for the past 70 years, confirming the reverence still held for this horse, his place in history fixed. It stands with an air of invincibility, the same certainty of will he stood with at the start of the post-war boom. Yet as the objective modeling of Timeform, the expanded mathematical examinations of Charles Justice ('The Greatest Horse Of All'), the handicapping skills of Andy Beyer, and further statistical modelings corroborate, whereas the reverence belongs to ManOWar, the track belongs to Secretariat; whereas KHP belongs to ManOWar, Churchill Downs and Belmont Park belong to Secretariat; whereas two to three decades of sire influence belonged to ManOWar and the future to his son, the broodmare sire War Admiral, the X-Factor belongs to Secretariat, his daughters and the future they bring. These along with Citation, Count Fleet, Seattle Slew, and a few others benchmark American racing even today.

Secretariat's Stride Length and Gait

When I did research on Secretariats stride, I initially relied on biographical data such as that found in the works of Raymond Woolfe and Bill Nack, both citing the 25 foot mark taken just after one of the horses' workouts. Apparently, it was the longest one they found within the 5 furlong distance he ran that day. On perusing the work Thoroughbred Champions: Top 100 Racehorses of the 20th Century, in the piece addressing Secretariat, the horses' stride length was recorded at 26 feet, a figure probably taken from the Marlboro Invitational where for the first and only time Riva Ridge and Secretariat competed. In that meet, with the two running side by side down the stretch, both jockeys in post race discussion estimated Secretariat's stride 3 feet longer than Riva's which was known to be 23 feet. At Secretariat.com in the Ask Penny Archives, Secretariat's owner commented that though they measured his stride at 25 feet after one of his workouts, she thought it reached 28 feet down the backstretch of his Belmont Stakes. Searching the internet, I found information relating to a horses' stride angle, ways of measuring stride length, and a few other topics. I also found a variety of statements claiming Secretariats stride measured such and such, ranging from 22 to 25.25 feet. At Kentucky Horse Park , a scale exists that allows vistors to compare the stride lengths of some of the finest Thoroughbreds foaled in this country. There, evidently, Secretariat's stride is listed at 24 feet (where this measure was taken from, one can only guess); ManOWar's at 28.

To begin with, why is stride length even a worthy topic of discussion? Good stride length is important because it is a product of good conformation and mechanics, and the best classic formations of both should result in a sound running, athletic specimen getting an edge in competition. Some debate various aspects of this. Ron Turcotte, Secretariat's jockey, stated that stride length is of less importance than the time it takes a horse to complete a full stride, or how fast a horse can accelerate in the critical moments of a race. In the end, the ground a stride covers is of great importance, but not the only component that counts. In the 2004 Preakness that found the big striding freshman Rock Hard Ten and the much smaller but seasoned Smarty Jones in head to head competition, Smarty Jones ran a Beyer Figure of 118 leaving Rock Hard Ten some 8 to 10 lengths behind in the stretch. More recently, in human terms, track champion Usain Bolt's long strides and rapid acceleration led him to record breaking performances at both sprint (100 meter) and 200 meter distances in the 2008 Olympics. It is said he will compete at the 400 meter distance in the 2012 heats. So both length and stride time are of importance. Usually a sprinter covers less ground per stride but with faster stride times while distance runners cover longer spaces with slower acceleration rates. Where does Secretariat fall? Writers have considered him a hybrid, somewhere in between. In 1973, a heavily muscled corpus resembling a sprinter standing 16.2 with length close to a distance runner, Secretariat demonstrated strong acceleration with great stamina, with the ability to easily traverse classic distances. The horse displayed long stride length.

In getting estimates of Secretariat's stride length I began with the simplest method available: using film, I estimated the number of strides it took for the horse to cover a known distance, usually a 16^th (330 ft), and then calculated the average. This was not always an easy task. More often, horses don't enter and exit 16^th poles evenly. Many times I had to estimate reasonable stride portions at the margins and work from there. I discovered that depending on the stage of the race, whether it was early, middle or late, length varied. Also, depending on the location of the horse, whether the horse is running outside, in the middle or inside the field of competition, or through the turns, this affected the number of strides taken from beginning to end. Rarely does a horse take a straight, unimpeded path. Pace also affects the stride count.

With Secretariat, counts of 13 to 14 strides per 16^th down the stretch of races were found, for an average range of 23.5 to 25.4 ft per stride depending on the 16th. I found this wanting so other methods were employed, yielding tenuous results. Eventually I found an important piece of research done in 1978 by MIT professor George W. Pratt who studied horse gait in an effort to explain how thoroughbreds suffer bone fractures during competition. The piece was published in 1978's American Journal of Veterinary Research. Using film of Secretariat's Marlboro Cup and Belmont Stakes that could be broken down into frames per second, Dr. Pratt developed methods of estimating stride length and stride efficiency and concluded that horses are at risk when they extend beyond their efficient reach. In estimating stride length, he carefully calculated the ratio of frames per stride to frames per second and multiplied this figure, the stride time, by the horses velocity in meters per second (convertible to feet per second) through those same frames. The result is the distance the horse traveled through each stride in question. Applying those same methods to footage of Secretariat's Derby and Belmont, I was able to get good estimates of the horses' stride lengths through various sections and stages of the races, some of which were corroborated in Dr. Pratt's work. Whereas Dr. Pratt used small sections of stretch runs and concluded Secretariat's stride length at approximately 24.3 ft, I enlarged the sample to differing sections of the Derby and Belmont and found that stride length varied depending on the stage of the heat. First, confirming that the times in seconds and fifths per quarter clocked by the employed technology matched the official recorded times for both the Derby and Belmont, I proceeded to take the required readings through the stretch runs of both races. I found Secretariat crossing the Derby finish with a stride length of approximately 25.2 ft. In that race, it will be recalled that the horse closed the final quarter in an individual 23 seconds flat, the official time at 23.2 seconds and that because Sham entered the quarter one fifth of a second sooner. Secretariat's closing velocity of 54 ft/sec coupled with a stride time of approximately .46 seconds rendered the measure. In the Belmont, stride lengths at or near 26 ft were recorded in the first half of the race, where, on entering the backstretch, Secretariat and Sham were racing at speeds at or near 60 ft per second; Sham in that same space had strides reaching 25 feet. In the second half, down the backstretch, Secretariat's velocities decelerated to roughly 57 ft per second reducing stride lengths to 25 ft. and above. In the final half, through the final quarter, tracking stride times and specific velocities became more difficult as the camera moved further from the subjects; nevertheless an estimated closing time of 52 ft per second at the wire was determined rendering strides just exceeding 24 feet for the champion. This last reading was corroborated in Dr. Pratt's work. Through the last mile, stride time remained constant making velocity the chief determinant of stride length. In the first half, faster stride times at higher velocities gave both Sham and Secretariat their largest strides. As I did not work the Marlboro Cup and Dr. Pratt did not work the Derby, further findings were not corroborated. These observations lead one to conclude that stride length occurs in ranges and is contextual, dependent on what the subject is doing at the time of the measurements. Faster stride times indicate acceleration, slower times indicating longer leveling motions. High velocities are achievable under both circumstances. As average stride lengths in the Thoroughbred community reach 21 to 22 feet, one can see that Secretariat and others who offer lengthy stride measures are most likely good distance runners. With Secretariat's proven stamina, this suggests the horse should have been an exceptional distance runner. With this, Secretariat's stride range is listed at 24 to at least 26 feet (see Additional Findings under the Discussion link), with a long gallop measuring 25 plus, depending on velocity and stage of the race.

One point of interest: in viewing Secretariat's gait in slow motion, when the horse extended its hinds to the rear prior to recoil, it tended to bring its hooves together at a centered focal point. This action provided an extra fraction of kick. In viewing Sham's action, this did not happen. Those rears tended to extend straight before recoiling without the extra frame of thrust. Sham's stride time was thus a little shorter thereby impacting stride length. The leggy 16.2 challenger spent a little more energy in high velocity runs at least in part because of the lack of the lengthier extension. Viewing older film of Citation's Kentucky Derby stretch run revealed similar action. Citation's rears tended to extend straight and slightly outward before recoil, lacking the extra fractional kick. This trait in Secretariat's profile might have been the outcome of the 'sloping rear' that often received conformational critique. In an interview, owner Penny Chenery referred to this sloping rear as the 'jumping bump', a trait inherited from its sire line that identifies jumping horses. She felt it gave the horse extra propulsion. If this horse was a hybrid, it may have been a jumper/runner cross.

A quick footnote: ManOWar's 28 foot stride was never corroborated. Though Walter Farley's semi-fictional account of the champion and Kentucky Horse Park record it, 3 other biographers never confirmed it. Both Page Coopers' (1950) and Dorthy Ours' (2006 ) accounts list it at 25 or more. Biographer Ed Bowen stated in his 2000 BloodHorse publication on the horse that the earliest BloodHorse reference to the champion's stride designated it at 26 ft. A reference is found in the 1974 Daily Racing Manual that listed it at 28 ft without citing sources. This was followed with a change in the 1975 Manual listing it at 25 to 28 feet. The following excerpt taken from a post by John Harris to a Network54 Forum in 2005 summarizes the issue in more detail:

---------------------------------------------------------------------------------

ManOWar Photo & Stride Length
by John Harris

I would like to offer my comments in response to 2 issues posted elsewhere on these web site pages:

#1. Length of MOWs stride:  my brother-in-law (Steve C.) and I conducted considerable research on this subject, in the early 1980s with the assistance of the National Museum of Racing in Saratoga and with Doris Jean Waren at the Keenland Library in Lexington, KY. We uncovered that the maximum length of MOWs stride, as Officially Measured, was 26 feet. This was recorded in both the Thoroughbred Record (Nov. 6, 1920) and Bloodhorse Magazine (April 30, 1966), as follows:

'ManOWars stride, which has been reported variously from to 25 to 28 feet, was measured in the stretch run of the Lawrence Realization at 26 feet.'

C.W. Anderson, in his book, Big Red (c.1943), wrote,  .. 'his stride, which covered 29 feet, is the longest on record.' (see, page 26).

Also, an article by Theodore Von Ziekursch, which appeared in the Saturday Evening Post on July 26, 1930, stated,  'at Harve de Grace they measured his stride ... and in the back stretch, with the
jockey turning on the juice by letting him out, this was increased to 29 feet. '

However, both the book and the article are written with considerable hyperbole and embellishment, so their 29 foot stride claims must be 'questioned' (my term).

-----------------------------------------------------------------------------

Finally, in a footnote in the appendix of Dorthy Ours' work, a comment is found describing the 28 ft mark as 'legend'.

With respect to Secretariat, the question of singularities, or non-repeatable events or achievements, can be approached from this angle. I once did a rough sketch of the odds of a horse winning the triple crown while running all 3 races in record time.  Recalling those races, the Derby was run in 1:59 2/5s; the Daily Racing Form clocking of the Preakness showed a time of 1:53 2/5s; and of course the Belmont's time of 2:24 flat. Though the Preakness time is not official, most in the know accept it because of the substantial evidence supporting it.  In calculating this figure, if I remember correctly, the odds were the product of 2 probability factors covering a window of about 50 years, from about 1926 to 1975 or so. The 2 factors were the odds of a horse winning the crown, and the odds of a horse setting track records in all three events given 'fast' tracks. After much work, the figure derived was one time in some 550 triple crown attempts, and less than two in about 1100. Given that an attempt occurs one time annually, this figure translates to 1 time in about 5.5 centuries and twice in some 11 centuries; or less than twice in 1000 years. Secretariat's Triple Crown was his greatest achievement in his truncated career not so much because he was successful in accomplishing it, but because of the way in which he accomplished it. It is this odds figure that gives some sense to how he did it and it is this figure that sets him apart. Some commentators have tried to diminish the horse's accomplishments over the years, but in truth this colt was a millennium horse. If I were to recalculate inclusive of the racing years to the present day, this outcome would become more remote. One can count maybe one other that had such an achievement in terms of odds, and that was Citation's record 15 straight in a single 'season. However, Citation's figure counts only as far as American horses are concerned for even this singular achievement pales in comparison to Kincem's 17 and 0 in her second season of running. Kincem ran in the 1870s in Europe and achieved the greatest singularity in finishing a racing career with a record of 54 and 0. This filly ran against the best colts, horses and mares of her time and more than once carried loads of up to 150 pounds. Furthermore, on multiple occasions, she competed at distances of at least 2 miles.  These great achievements plus maybe one to two others stand out in racing history.
As a postscript, Rapid Redux recently increased his win streak to 22 this January of 2012 which included a record 19 wins and no losses for the 2011 season. This not only tied Citation's record 19 wins for the 1948 season but also broke his record 15 straight for the year. Rapid Redux's 19 single season win streak also outdid Kinscem's 17 and 0 outcome in her 3 year old campaign. The difference however is not in the streak itself, but in the nature of the streak. Whereas Kinscem and Citation competed at top grade levels and distances, Rapid Redux has competed at allowance levels. The streak itself is insufficent in determining the quality of the specimen. Rapid Redux however has carried high weight in his competitions and ran distances of up to 9 furlongs. Perhaps on a pound for pound basis can the Rapid Redux acheivement be considered among the best, a true singularity.

This paper presented a mathematical approach to two of the Triple Crown races Secretariat ran in 1973, the Kentucky Derby and Belmont Stakes. After using the Statistical method of Regression Analysis, it became clear that Secretariat's Belmont  bested  ManOWar's 1.5 mile Jockey Club performance by 11 to 12 lengths. But for two issues, the methods took into account all the material conditions that separate 1920 from 1973, track speed, etc.... Of the two issues the first was the direction the race was ran, the Jockey Club was performed left handed; it is not known how direction affects outcomes. The second was the weight carried, ManOWar posted 118 pounds to Secretariat's 126. Otherwise, Secretariat's Belmont was and most likely continues to be the best 1.5 mile performance on a right handed dirt track in recorded racing. Faster times are listed on turf surfaces, but that is in large due to the nature of the surface: turf is harder and faster, on average close to 2 seconds better at 1.5 miles. Similar comparisons and methods were used with Secretariat's 10 furlong Kentucky Derby. Issues became apparent at the outset, ManOWar did not run the Kentucky Derby in 1920.  The comparison race was the 10 furlong Travers Stakes on the 9 furlong track at Saratoga. In comparing these 2, Secretariat's performance bettered ManOWar's by 3 to 4 lengths.  In terms of speed, it is not known how different the Saratoga course was from the shorter 8 furlong track at Churchill Downs at that time. No doubt, differences affect time. ManOWar, however, carried 129 pounds in that race to Secretariat's 126.  Field size affects time as well, Secretariat competed against 12 others, ManOWar two. 
Based on these observations, all things equal, it is easily determined that Secretariat would have been the most formidable opponent to ManOWar, one that would have drawn the very best from him, and one that had the capacity to defeat him.