Run up in horse racing

[cj]

Here is something I'm working on for our blog. Any feedback is appreciated.

Run Up and the Effect on Final Time

What is “run up?”

Run up is the distance traveled before the timing of the race begins. It gives horses a running start. The timing of the race doesn’t begin until the first horse hits the point where the official distance of the race begins, referred to commonly as tripping the beam. This horse triggers the beginning of electronic timing of the race.

Why does run up lead to misleading times?

Obviously, the most accurate way to time horse races would be to do it from the gate and to run exactly the official distance. But, that isn’t how it is done. Horses are timed from a running start. The effect is that every horse EXCEPT the horse that hits the beam first is timed using the official distance of the race PLUS the distance the first horse is ahead when it begins the timing. This doesn’t seem to be that big of an issue. After all, horses that trail at the beam did so because they weren’t as fast as the leader. But that logic doesn’t hold when examined more closely.

Here is an example involving two theoretical races at Santa Anita Park, both run at one mile on the dirt course. The track uses a very long run up for these races to try to bring the start farther away from the first turn. Sixty yards is not uncommon.

Race 1: Horse A out-sprints the field early and trips the beam five seconds after break from the gate. The next horse, Horse B, doesn’t cross the beam until .30 seconds later. The leader goes much too fast early, quits badly and staggers home last, well back of the others. Horse B takes over the lead and wins the race. The final time is reported as 1:36.00.

Race 2: Horse C breaks alertly, but not as quickly as Horse A above. He trips the beam 5.30 seconds after the start. He takes the field wire to wire, winning the race, and the final time is once again reported as 1:36.00.

This happens every day in horse racing. It isn’t always this extreme, though it isn’t a rare occurrence either. Which horse ran the faster race, B or C, or is it a tie? Both horses took 5.30 seconds to complete the untimed portion of the race (doesn’t that alone sound ridiculous?), and both horses are reported as having ran one mile in 1:36.00. Examine the entire race to properly answer the question.

Race 1: The total time needed to complete this race from gate to wire was 1:36.00 seconds plus the five seconds it took to reach the beam, for a total of 1:41.00.

Race 2: The total time needed to complete this race from gate to wire was 1:36.00 seconds plus the 5.30 seconds of “untimed” racing, for a total of 1:41.30.

Clearly, Horse B ran the faster race. In racing lingo, that is nearly two full lengths. Why are races timed with this arcane, inaccurate method as we approach 2014? At some point, isn’t “we’ve always done it that way” just a tired excuse?

Effect on Figure Making

It is important for figure makers to understand these differences. Many handicappers complain when races run on the same track and at the same distance are given different final time figures. They complain figure makers are just trying to fit races together like a puzzle. It is important to remember the data used for the figure process is flawed. Many times it is wise to not just trust the times when assigning ratings. This is just one example of the issues those trying to rate races come across. It is far from the only one. But, this one is unnecessary and can easily be removed.

[stuball]

I agree CJ not on that race at that track but we have to fit same distance
and different distances from other tracks....a seemingly impossible task--
yet we keep trying to put the puzzle together in some kind of order...
Thank You CJ for your diligence to duty to try to figure out a way to
account for this...I am now importing the runups and have about 3 years
of data....I should take it into account somehow but have never quite got my
brain to compute this.
Thanks for your figs....hope some day to get data files again
to import to my puter....

Stuball :ThmbUp:

[cj]

I agree CJ not on that race at that track but we have to fit same distance
and different distances from other tracks....a seemingly impossible task--
yet we keep trying to put the puzzle together in some kind of order...
Thank You CJ for your diligence to duty to try to figure out a way to
account for this...I am now importing the runups and have about 3 years
of data....I should take it into account somehow but have never quite got my
brain to compute this.
Thanks for your figs....hope some day to get data files again
to import to my puter....

Stuball :ThmbUp:

There is so much more wrong with run up, this only touches on it. We have varying run up for different distances at the same track. We even, occasionally, have it for the SAME distance at the same track. I touched on final time, but the effect on fractions is even greater.

But in a sport that desperately needs new blood, how do you explain something like my example?

[Jeff P]

Q. How do you explain something like that (the untimed portion of the race?)

A. Imagine how much more popular The Olympics (or NCAA Track and Field) would be if they had to explain something like that...


-jp

.

[GMB@BP]

Dont think most execs at tracks equate having to provide accurate information with betting. Degenerates bet whether they have good info, whether the takeout is high or low, etc.

[cj]

Dont think most execs at tracks equate having to provide accurate information with betting. Degenerates bet whether they have good info, whether the takeout is high or low, etc.

They probably don't, but eventually shrinking pools might wake them up. The bettor of 2013 is not the same as the one of 1963.

[PICSIX]

I don't understand why they can't have the timer start at the opening of the gate and let us know this "1 mile race" was 1 mile plus 85 feet? :confused:

[cj]

I don't understand why they can't have the timer start at the opening of the gate and let us know this "1 mile race" was 1 mile plus 85 feet? :confused:


Me either. This would also eliminate many other errors with timing that involve the beam being tripped early or not at all. Just tie the timer to the opening of the gate.

[cj]

Here is another one that is tough to explain. The run up for a mile race is 200 feet, or about 67 yards. Some places card races at 1m 40 yards and/or 1m 70 yards, but this race is called "one mile". It makes no sense to do things this way.

I'm sure that track record thing will be trotted out as an excuse, as if that is a reason to keep around a bad method.

[GMB@BP]

They probably don't, but eventually shrinking pools might wake them up. The bettor of 2013 is not the same as the one of 1963.

I saw that same sentence typed a decade ago.

Horse racing is one of few industries that does not seem to use market principles, like raising prices with declining attendance.

[lamboguy]

a few years ago i won a race at Keeneland, it was a 7 furlong race, i don't remember the time, but it looked pretty normal. a friend called me to congratulate me for breaking the track record with a 3 year old. i said what? i looked up the race in equibase and sure enough it said track record. the distance was slightly more than 7 furlongs and it was the very first time that distance was ever run at Keeneland.

i went to the Preakness this year and saw more than one finish line, i asked someone about that and was told they use different finish lines for different races.

these type of things must be driving the figure makers of the world crazy these days. it doesn't help the people that are betting these things based on figures and times either. anyone that is relatively new to this game rely's on the accuracy of these numbers in same way to make their wagering decision's.

another form of timing that is completely useless are workouts from the pole. some horses start closer to the pole than others and the ones further back will alway's be going faster once the horse gets clocked from the pole.

[JustRalph]

Damn, starting to feel like a real job yet? :lol:

[cj]

Damn, starting to feel like a real job yet? :lol:

Nah, still love it!

[Just a Fan]

I think I like having the runup. Since we analyze the sport using fractional times, it would be awkward to analyze the early pace of a race if the race was timed from a standing start. But you make good points.

Not entirely relevant to this post but worth mentioning - QH races are timed from a standing start. Since races are timed to the thousandth of a second, you have to hope that the guy that drives the tractor that pulls the gate to the starting line is very precise and consistent at putting the gate in just the right spot. With TBs, the runup takes that variable out of the equation. I have a tough enough time parallel parking, I would never be able to pull that gate onto the starting point.

[ronsmac]

This is why conventional pace and speed figures have so many flaws. Runup and wind direction was probably the first things I picked up on as a youth playing and watching races. Unfortunately, smarter and smarter players are noticing every facet of the game that may have given me a small edge 20 yrs ago. I believe 2014 will be my last year betting on horses. lol

[Tom]

How does Takus do it?
Do they have a run up or start as the gates open?

[cj]

How does Takus do it?
Do they have a run up or start as the gates open?

Run up is still included, though the methods are different. Without run up, fractions would look MUCH slower than we are used to seeing.

[PhantomOnTour]

CJ - we've discussed it before, and the difficulty involved, but hand timing these races on my own is looking more and more like a possibility....at least for NY races.

Some of these shenanigans are ridiculous. A track should NEVER run the same distance on the same day with different run ups...GP is the only one I know of so far.

[Clocker]

If variations and inconsistencies in run up are significant, it would seem to make more sense to handicap based on internal fractions. Using internal fractions would factor the run up out of the data. It would also give you an edge in that you are using better raw data than the public, or the public handicappers, are using.

[RXB]

Run up is still included, though the methods are different. Without run up, fractions would look MUCH slower than we are used to seeing.

Pimlico 6f races and Gulfstream 1m dirt races being the prime examples with their five-foot runups.

[cj]

If variations and inconsistencies in run up are significant, it would seem to make more sense to handicap based on internal fractions. Using internal fractions would factor the run up out of the data. It would also give you an edge in that you are using better raw data than the public, or the public handicappers, are using.

Excluding part of the race doesn't really work.

[barn32]

At some point, isn’t “we’ve always done it that way” just a tired excuse?I know this answer is used to explain a lot of things in life, but can anyone tell me what the original reason was for deciding to use a run-up in the first place?

[Clocker]

I know this answer is used to explain a lot of things in life, but can anyone tell me what the original reason was for deciding to use a run-up in the first place?

One reason would be for short route races on a mile track, the gate would have to be far enough up the stretch from the finish line, and the tote board, so it could be pulled off the track into the infield or onto an inner track.

Another reason would be to get more accurate times back in the day when races were hand timed. If the official timer is high up in the grandstands, it is hard to see exactly when the gate opens or when the lead horse passes the starting pole in a sprint starting on the back stretch or a chute.

Back in the day, there was a flag man standing at the starting pole. When the horses were all in the gate, he would raise the flag over his head. (And the track announcer would say, "The flag is up.") The flag man would drop the flag when the lead horse passed the pole. Dropping the flag as the horses went past would be more precise than trying to drop it at the exact moment the gate opened.

[classhandicapper]

This is another of an almost endless number of reasons I stopped being so literal about figures.

The first time I realized the run-up was a bigger issue than generally perceived was when I saw a horse pretty much beat the gate and bolt to a 2-3 length lead right at the start. I realized that everyone else's figure was going to get penalized because he was so quick out of the gate and broke the beam so much faster than average.

There's also an issue when the start is especially rough and almost all the horses get banged around (maybe it happened in the JCGC this year??) . When they break the beam they often haven't really hit their typical stride yet or maybe one horse did and the rest didn't.

The more accurate the better, but I'm pretty much at a stage where I just want to know if the race was fast, average, or slow for the class.

IMO quality figures are an almost indispensable part of handicapping (especially among lightly raced maidens and limited winners where there is a wide range of ability and the horses haven't sorted themselves out yet), but when people start forming strong opinions over a few points either way, I think they are probably in dangerous territory.

[banacek]

Some of these shenanigans are ridiculous. A track should NEVER run the same distance on the same day with different run ups...GP is the only one I know of so far.

They should never have different runups at all in my opinion.

Hastings was doing this a couple of years ago at 6f...sometimes 100 ft runup, sometimes 24 ft runup. And there was no way to tell in the form.

They seem to doing 100 ft. run-ups for 2yo races now and calling them about 6f and the 24 feet ones for older horses. But I still watch the replays to make sure. It is still a problem to make the adjustments, but I think I've got something that works.

[PICSIX]

They should never have different runups at all in my opinion.

Hastings was doing this a couple of years ago at 6f...sometimes 100 ft runup, sometimes 24 ft runup. And there was no way to tell in the form.

They seem to doing 100 ft. run-ups for 2yo races now and calling them about 6f and the 24 feet ones for older horses. But I still watch the replays to make sure. It is still a problem to make the adjustments, but I think I've got something that works.

The run-up distance(s) for past races should be listed on the results charts.

[illinoisbred]

The run-up distance(s) for past races should be listed on the results charts.
Equibase does give the run-up for each race in their charts.. though not always correct. A tough one is the 6+1/2 at Arlington. The run-up is given at 64' yet the pole is not 64' nor the true start of the race. I asked AP for an explanation and was told if the pole was placed where it needed to be, it would be out on the track(the 6+1/2 starts just up the chute).

[jasperson]

This is why conventional pace and speed figures have so many flaws. Runup and wind direction was probably the first things I picked up on as a youth playing and watching races. Unfortunately, smarter and smarter players are noticing every facet of the game that may have given me a small edge 20 yrs ago. I believe 2014 will be my last year betting on horses. lol
Are you planning to die in2014?;)

[cj]

Equibase does give the run-up for each race in their charts.. though not always correct. A tough one is the 6+1/2 at Arlington. The run-up is given at 64' yet the pole is not 64' nor the true start of the race. I asked AP for an explanation and was told if the pole was placed where it needed to be, it would be out on the track(the 6+1/2 starts just up the chute).

There are all kind of silly things like this that go on around the country. Totally unnecessary but nobody does anything to change them.

[Cratos]

The “run-up” affect on the final time on a horse in thoroughbred racing is an invariable discussion among horseplayers; especially among the speed figure and pace figure aficionados.

However what should be first understood is the timing of the thoroughbred race.

Currently there are two methods, the teletimer methodology which is the legacy system that times the race and not the horse at predetermined distances; historically called points of call (POC).

In essence the legacy system times the race and not the horse.

The second method is Trakus, the “new Kid” on the block and is a speed-distance curve measurement system for thoroughbred horseracing that measures the speed of each horse in the race with respect to its distance travelled in the race and from that measurement, the distance between horses at the time of POC (point of call) can be measured.

Again, Trakus times the horse and distance travelled by the horse during the race.

In contrast, the legacy “beaten length” system is a point measurement system which measures the time of the leading horse at the pre-determined POC of the race and does not measure distance. Distance between horses is calculated in this system from the non-standard metric of the length.

In essence the legacy “beaten length” system does not measure distance and only measure the time of the leading horse at the aforementioned predetermined POC. Distance between horses using this method is fanciful.

In neither case is the run-up distance included in the final time and it shouldn’t be, because that would violate the stated distance metrics of the race (e.g. if a race is stated to be run at 5 furlongs and it is timed a 5 furlongs plus the run-up distance, then the stated distance is incorrect).

From what researchers who have studied the impact of “run-up” on final time of a horserace understands, it is that “run-up” exists in thoroughbred racing to achieve acceleration; a thoroughbred have to be convinced to run.

During their research they found that a thoroughbred although considered to be the fastest land animal it does not accelerate fast from a stand still when compared to a cheetah, a quarter-horse, and a human being.

When they compared a thoroughbred against a cheetah, a quarter-horse, and a human being and all were faster over the length of a basketball court (94 feet) than the thoroughbred.

However beyond that point the thoroughbred was faster. In their findings it was found that the maximum stride length for a horse is about 23 feet, with a stride rate of 2.25 strides/second and a top speed of 43.5 miles/hr and the average stride length of the thoroughbred racehorse are approximately 20 feet long and it can take up to 150 strides per minute.

Therefore should the “run-up” be included in the final time? As stated earlier the answer is no. If you have the parameters of the run-up you can calculate the acceleration by first calculating energy dissipation by the horse for the “run-up” and converting it to get the force; and use an estimated mass to calculate the acceleration.

To me that is too tedious to prove a nebulous impact on the horse’s final time because the geometric shape of the racetrack, the side force in the turn, and the starting positions of each horse have far more impact on final time.

[cj]

The “run-up” affect on the final time on a horse in thoroughbred racing is an invariable discussion among horseplayers; especially among the speed figure and pace figure aficionados.

However what should be first understood is the timing of the thoroughbred race.

Currently there are two methods, the teletimer methodology which is the legacy system that times the race and not the horse at predetermined distances; historically called points of call (POC).

In essence the legacy system times the race and not the horse.

The second method is Trakus, the “new Kid” on the block and is a speed-distance curve measurement system for thoroughbred horseracing that measures the speed of each horse in the race with respect to its distance travelled in the race and from that measurement, the distance between horses at the time of POC (point of call) can be measured.

Again, Trakus times the horse and distance travelled by the horse during the race.

In contrast, the legacy “beaten length” system is a point measurement system which measures the time of the leading horse at the pre-determined POC of the race and does not measure distance. Distance between horses is calculated in this system from the non-standard metric of the length.

In essence the legacy “beaten length” system does not measure distance and only measure the time of the leading horse at the aforementioned predetermined POC. Distance between horses using this method is fanciful.

In neither case is the run-up distance included in the final time and it shouldn’t be, because that would violate the stated distance metrics of the race (e.g. if a race is stated to be run at 5 furlongs and it is timed a 5 furlongs plus the run-up distance, then the stated distance is incorrect).

From what researchers who have studied the impact of “run-up” on final time of a horserace understands, it is that “run-up” exists in thoroughbred racing to achieve acceleration; a thoroughbred have to be convinced to run.

During their research they found that a thoroughbred although considered to be the fastest land animal it does not accelerate fast from a stand still when compared to a cheetah, a quarter-horse, and a human being.

When they compared a thoroughbred against a cheetah, a quarter-horse, and a human being and all were faster over the length of a basketball court (94 feet) than the thoroughbred.

However beyond that point the thoroughbred was faster. In their findings it was found that the maximum stride length for a horse is about 23 feet, with a stride rate of 2.25 strides/second and a top speed of 43.5 miles/hr and the average stride length of the thoroughbred racehorse are approximately 20 feet long and it can take up to 150 strides per minute.

Therefore should the “run-up” be included in the final time? As stated earlier the answer is no. If you have the parameters of the run-up you can calculate the acceleration by first calculating energy dissipation by the horse for the “run-up” and converting it to get the force; and use an estimated mass to calculate the acceleration.

To me that is too tedious to prove a nebulous impact on the horse’s final time because the geometric shape of the racetrack, the side force in the turn, and the starting positions of each horse have far more impact on final time.

Well, the problem is that having an "untimed" portion of the race, and that is exactly what it is, results in the timing of races being inconsistent. My example races in the first post happen every day around the country. There is no need for it. It is a flawed system.

[Saratoga_Mike]

1) From what researchers who have studied the impact of “run-up” on final time of a horserace understands, it is that “run-up” exists in thoroughbred racing to achieve acceleration; a thoroughbred have to be convinced to run.

2) Therefore should the “run-up” be included in the final time? As stated earlier the answer is no. If you have the parameters of the run-up you can calculate the acceleration by first calculating energy dissipation by the horse for the “run-up” and converting it to get the force; and use an estimated mass to calculate the acceleration.



1) What researchers have written on this matter? Do you have links to some of their work? Sounds worth reading.

2) Could you provide an example of this calc using a race from GP with the 5-foot run-up?

[Stillriledup]

It seems that the runup isnt the problem necessarily, the problem is that the DRF/Equibase, etc don't list the actual distance of the race.

Why not just measure exactly 5 Furlongs or 6 Furlongs and put the gate right at the spot? Put the gate right by the timer so when the gate opens and the first nose is one foot out of the gate, the timer gets tripped. Why wouldnt the gate be right on the beam?

[Saratoga_Mike]

SRU - you just made the case for run-up being a problem (issue to be dealt with).

[Stillriledup]

SRU - you just made the case for run-up being a problem (issue to be dealt with).

Either they can put the starting gate right at the beam...OR, the official charts can list the run up time in 100ths in the chart. Seems like putting a man on the moon might be easier.

[cj]

Either they can put the starting gate right at the beam...OR, the official charts can list the run up time in 100ths in the chart. Seems like putting a man on the moon might be easier.

You know, I'd be perfectly fine with leaving things as is if the run up distance and actual time of the run up was reported as well.

[Stillriledup]

You know, I'd be perfectly fine with leaving things as is if the run up distance and actual time of the run up was reported as well.

Maybe its easier to record the run up time than it is to move the gate right to the beam.

I wonder why they don't put the gate right at the beam, maybe its just an unwritten rule in racing that there has to be runup for some reason.

[PICSIX]

What about track records? If horse A sets the record with no run-up & then horse B breaks that record with 60 feet of run-up that is a problem! :ThmbDown:

[cj]

What about track records? If horse A sets the record with no run-up & then horse B breaks that record with 60 feet of run-up that is a problem! :ThmbDown:

I mentioned this a few times already. But, it wouldn't be a problem if they just gave us the actual run up distance (not a guess) and the time it took to run it. Then, you just report distance and time as is done now. Problem solved.

That said, I don't know many people that really care about track records. They are a function of the track speed as much as they are the horses most of the time.

[classhandicapper]

I'm going to play devil's advocate and suggest that understanding that there are issues with run up distance and time gives you an advantage over people that don't consider issues like this. So maybe we don't really want more accurate information to be available to everyone. ;)

[highnote]

I was going to say the same thing, Class. Andy Beyer once said that you should be happy that races are not timed accurately because it gives you an edge if you can figure out how to do it better.

Ragozin used (uses?) video playback with a time code generator hooked up to it. The time code is "burned" into a portion of the video and he could (can?) get an accurate time in the race to about a 30th of a second.

The U.S. television broadcast standard is 29.97 frames per second -- or about 30 frames per second. So each frame of video is 1/30th of a second. So a 6 furlong race might take 1 minute 10 seconds and 3 frames -- or 1:10.1 -- from gate to wire. Then you can factor in the runup by looking where the gate was placed in relation to the first beam timer. If you know how many feet are between the fence poles you can get a pretty good estimate of the distance from the gate to the first beam. This would allow you to get closer to knowing the actual distance of the race.

Ragozin wrote in his book that because Saratoga is a casual meet, the starting gate is not always placed in the exact same position for every distance. In other words, one day the runup to 6 furlong races might be different than the runup the following day.

The best way to know for sure is to watch video.

I can imagine that the one reason the big betting syndicates have a big advantage is because they hire people to watch video and calculate accurate speed figures using accurate times and distances.

I'm going to play devil's advocate and suggest that understanding that there are issues with run up distance and time gives you an advantage over people that don't consider issues like this. So maybe we don't really want more accurate information to be available to everyone. ;)

[cj]

There are problems with timing off of video. First, you don't really know the distance of the run up. You can get fairly accurate. Second, you can't always tell where actual timing of the race begins. It can be tough to see, and when you can see it, it is often at an angle.

Even if you get both of the above correct, you will never be able to time fractions off of current video. So you'll have a good final time, but no fractions.

[cj]

I'm going to play devil's advocate and suggest that understanding that there are issues with run up distance and time gives you an advantage over people that don't consider issues like this. So maybe we don't really want more accurate information to be available to everyone. ;)

I understand that point of view, but I'd rather have accurate data and see if I can use it better than most.

[highnote]

Even if you get both of the above correct, you will never be able to time fractions off of current video. So you'll have a good final time, but no fractions.

You shouldn't have to time the fractions off of video because that's done by Equibase and 99.9% of the time they should be accurate -- except at Gulfstream and a few other cases.

So all you really need to do is time the runup and estimate the distance and add those two values to the final time and the condition book distance.

Right? Or am I missing something. I've never used video to do this. I'm just taking Ragozin's word. Intuitively, it makes sense to do it this way.

[highnote]

I understand that point of view, but I'd rather have accurate data and see if I can use it better than most.


Good point -- the hard part, as you know, is getting accurate data.

[cj]

You shouldn't have to time the fractions off of video because that's done by Equibase and 99.9% of the time they should be accurate -- except at Gulfstream and a few other cases.

So all you really need to do is time the runup and estimate the distance and add those two values to the final time and the condition book distance.

Right? Or am I missing something. I've never used video to do this. I'm just taking Ragozin's word. Intuitively, it makes sense to do it this way.

I wouldn't say 99.9%, but it is a pretty high number. You are right, just timing the run up and distance and adding it to the official times would seem to be enough. I'm just saying it isn't that easy given the angles. Tough to judge the actual "beam" and tough to judge the actual run up distance.

[highnote]

I wouldn't say 99.9%, but it is a pretty high number. You are right, just timing the run up and distance and adding it to the official times would seem to be enough. I'm just saying it isn't that easy given the angles. Tough to judge the actual "beam" and tough to judge the actual run up distance.

to judge the distance, assuming you can find the beam, you could probably make a pretty good estimate by eyeballing the rail posts. Say they are about 10' apart. You know a horse is about 8' long from nose to tail. How many horse lengths are there in the distance between rail posts and if you know how many posts there it should give a pretty good idea of distance.

however this a lot of shit, mess and bother. Maybe par times are nearly just as good given the randomness of a horse race since par times would capture the effect of the runup even if you dont actually measure it -- think Monmouth runups vs Belmont at 6 furlongs.

[JustRalph]

30 frames a second is a nice number to banter about, but toss in compression and the fact that some cable companies and SAT services compress video multiple times before it arrives at your playback device, no way I would rely on it

[highnote]

I would recommend buying the actual videos on DVD from the racetracks to make sure you are getting broadcast quality. That's what the Wizard does. So probably anyone can buy them.

Just talked to a broadcast video engineer friend. HD broadcasts are still 30 frames per second.



30 frames a second is a nice number to banter about, but toss in compression and the fact that some cable companies and SAT services compress video multiple times before it arrives at your playback device, no way I would rely on it

[Cratos]

To me the solution to the question on whether run-up leads to misleading times is solvable by solving for the slope of the curve.

IMO there isn’t any need to use videos or any other graphic data to confirm this answer because all of the needed quantitative parameters are given for a theoretical solution. I say theoretical because that should be the starting point and the answer from that exercise can easily be adjusted to practical usage as more data is gathered.

Also it is worthwhile to remember that this solution is about the ideal thoroughbred conformation which is a subject many hardened racing people has spent a lifetime studying, yet still never completely conquered. But to even the most experienced, it all too often remains an area of mystery.

It is a mystery because the search for discovery is centered on the two key requirements to produce power and those being speed and strength.

Yes, every racehorse needs speed to be competitive and to be a winner, but it is strength (the focus of this discussion) that it needs to start.

Strength will give the horse a good initial jump, the power to maximize top end speed and a large stride from its strong hind quarters.

Will there ever be an absolute answer to the run-up dilemma? I don’t think so because of the difference in track configuration, environment conditions, each horse’s innate ability, etc.

However I believe that we can calculate a range that will have a high degree of accuracy if not precision.

Therefore what we are looking for by example is to calculate the slope of the curve at the end of the race if a horse starts exactly at the zero starting position; this curve is inverse “bathtub” shaped with its origin at zero.

Next we extend the starting position incrementally (adding the run-up distance) in –x (distance) to see what happens to y (speed) as the run-up distance becomes longer. At some point there will be an optimal distance.

While I think it might be a neat calculation, I am skeptical about its usefulness because of the aforementioned variables that will affect it.

Also what I haven’t found in this discussion is the starting line placement of horses. There appears to be much concern about the “run-up” distance, but not any discussion about why the starting positions aren’t staggered.

With the starting positions in line at typically 90 degrees to the rail, the outside horses will always be penalized by extra distance to the first turn.

The question now becomes whether this situation cause adversity to the use of the horses’ energy due to its outside post position.

[highnote]

Good point, Cratos.

Also, even though there may be small timing errors, is the time and effort needed to determine more accurate race times worth the cost?

There are other factors that determine the outcome of races and all these other factors taken together are probably more valuable than determining slightly more accurate times.

I doubt that making measuring the time of races more accurately will produce noticeable differences in speed figures since speed figure ratings are usually rounded to whole numbers. A speed figure of 89.3 may be more accurate than a speed figure of 89.5, but how many more winners will it produce and is the splitting of hairs really worth the hassle?



To me the solution to the question on whether run-up leads to misleading times is solvable by solving for the slope of the curve.

IMO there isn’t any need to use videos or any other graphic data to confirm this answer because all of the needed quantitative parameters are given for a theoretical solution. I say theoretical because that should be the starting point and the answer from that exercise can easily be adjusted to practical usage as more data is gathered.

Also it is worthwhile to remember that this solution is about the ideal thoroughbred conformation which is a subject many hardened racing people has spent a lifetime studying, yet still never completely conquered. But to even the most experienced, it all too often remains an area of mystery.

It is a mystery because the search for discovery is centered on the two key requirements to produce power and those being speed and strength.

Yes, every racehorse needs speed to be competitive and to be a winner, but it is strength (the focus of this discussion) that it needs to start.

Strength will give the horse a good initial jump, the power to maximize top end speed and a large stride from its strong hind quarters.

Will there ever be an absolute answer to the run-up dilemma? I don’t think so because of the difference in track configuration, environment conditions, each horse’s innate ability, etc.

However I believe that we can calculate a range that will have a high degree of accuracy if not precision.

Therefore what we are looking for by example is to calculate the slope of the curve at the end of the race if a horse starts exactly at the zero starting position; this curve is inverse “bathtub” shaped with its origin at zero.

Next we extend the starting position incrementally (adding the run-up distance) in –x (distance) to see what happens to y (speed) as the run-up distance becomes longer. At some point there will be an optimal distance.

While I think it might be a neat calculation, I am skeptical about its usefulness because of the aforementioned variables that will affect it.

Also what I haven’t found in this discussion is the starting line placement of horses. There appears to be much concern about the “run-up” distance, but not any discussion about why the starting positions aren’t staggered.

With the starting positions in line at typically 90 degrees to the rail, the outside horses will always be penalized by extra distance to the first turn.

The question now becomes whether this situation cause adversity to the use of the horses’ energy due to its outside post position.

[cj]

I really don't get all the negativity about this. It is 2013. There is no reason we shouldn't have the exact distance run and the exact time it took to run it. End of discussion as far as I'm concerned.

[thespaah]

Here is something I'm working on for our blog. Any feedback is appreciated.

Run Up and the Effect on Final Time

What is “run up?”

Run up is the distance traveled before the timing of the race begins. It gives horses a running start. The timing of the race doesn’t begin until the first horse hits the point where the official distance of the race begins, referred to commonly as tripping the beam. This horse triggers the beginning of electronic timing of the race.

Why does run up lead to misleading times?

Obviously, the most accurate way to time horse races would be to do it from the gate and to run exactly the official distance. But, that isn’t how it is done. Horses are timed from a running start. The effect is that every horse EXCEPT the horse that hits the beam first is timed using the official distance of the race PLUS the distance the first horse is ahead when it begins the timing. This doesn’t seem to be that big of an issue. After all, horses that trail at the beam did so because they weren’t as fast as the leader. But that logic doesn’t hold when examined more closely.

Here is an example involving two theoretical races at Santa Anita Park, both run at one mile on the dirt course. The track uses a very long run up for these races to try to bring the start farther away from the first turn. Sixty yards is not uncommon.

Race 1: Horse A out-sprints the field early and trips the beam five seconds after break from the gate. The next horse, Horse B, doesn’t cross the beam until .30 seconds later. The leader goes much too fast early, quits badly and staggers home last, well back of the others. Horse B takes over the lead and wins the race. The final time is reported as 1:36.00.

Race 2: Horse C breaks alertly, but not as quickly as Horse A above. He trips the beam 5.30 seconds after the start. He takes the field wire to wire, winning the race, and the final time is once again reported as 1:36.00.

This happens every day in horse racing. It isn’t always this extreme, though it isn’t a rare occurrence either. Which horse ran the faster race, B or C, or is it a tie? Both horses took 5.30 seconds to complete the untimed portion of the race (doesn’t that alone sound ridiculous?), and both horses are reported as having ran one mile in 1:36.00. Examine the entire race to properly answer the question.

Race 1: The total time needed to complete this race from gate to wire was 1:36.00 seconds plus the five seconds it took to reach the beam, for a total of 1:41.00.

Race 2: The total time needed to complete this race from gate to wire was 1:36.00 seconds plus the 5.30 seconds of “untimed” racing, for a total of 1:41.30.

Clearly, Horse B ran the faster race. In racing lingo, that is nearly two full lengths. Why are races timed with this arcane, inaccurate method as we approach 2014? At some point, isn’t “we’ve always done it that way” just a tired excuse?

Effect on Figure Making

It is important for figure makers to understand these differences. Many handicappers complain when races run on the same track and at the same distance are given different final time figures. They complain figure makers are just trying to fit races together like a puzzle. It is important to remember the data used for the figure process is flawed. Many times it is wise to not just trust the times when assigning ratings. This is just one example of the issues those trying to rate races come across. It is far from the only one. But, this one is unnecessary and can easily be removed.
Ok..I have a question. At what point does the Trackus timing and distance system begin recording each horse's run?
Out of the gate? Or when the horse actually passes the official start marker for that race distance?

[cj]

Ok..I have a question. At what point does the Trackus timing and distance system begin recording each horse's run?
Out of the gate? Or when the horse actually passes the official start marker for that race distance?

From what I understand, they record from the gate, but adjust the clocking to match the official distance of the race.

[GMB@BP]

I really don't get all the negativity about this. It is 2013. There is no reason we shouldn't have the exact distance run and the exact time it took to run it. End of discussion as far as I'm concerned.

Its an old school mentality IMO and if anyone wonders why the game does not grow this thread is a good example why.

[Clocker]

I really don't get all the negativity about this. It is 2013. There is no reason we shouldn't have the exact distance run and the exact time it took to run it.

It's not negativity, it is acceptance of reality. We don't have it, so the short term issue is how to make the best of the situation.

The reason we don't have it is lack of incentive for the tracks. Most people aren't aware of the problem, and those who are aware and care about it, still play. We can continue to try to push for better timing, but in the meantime we have to deal with it as it is.

[cj]

It's not negativity, it is acceptance of reality. We don't have it, so the short term issue is how to make the best of the situation.

The reason we don't have it is lack of incentive for the tracks. Most people aren't aware of the problem, and those who are aware and care about it, still play. We can continue to try to push for better timing, but in the meantime we have to deal with it as it is.

It's been around long enough that anyone serious about it (including me) knows how to make the best of it by now. Your point is exactly what I'm doing, make people aware.

[highnote]

I really don't get all the negativity about this. It is 2013. There is no reason we shouldn't have the exact distance run and the exact time it took to run it. End of discussion as far as I'm concerned.

The Quarter Horse racing industry uses gate timers.

The technology to hook a timer up to the starting gate must be fairly simple and can't cost much money. It would be a simple matter for a track official to make a note of the position of the starting gate. There are plenty of places on the rail that could be marked in paint with the distance from the first beam.

[GMB@BP]

It's not negativity, it is acceptance of reality. We don't have it, so the short term issue is how to make the best of the situation.

The reason we don't have it is lack of incentive for the tracks. Most people aren't aware of the problem, and those who are aware and care about it, still play. We can continue to try to push for better timing, but in the meantime we have to deal with it as it is.

Or maybe the type of players who would be aware of it see things like that are no longer interested in the sport, but yea the dying demographic of the sport is unlikely to even know it exists.

[Cratos]

It's not negativity, it is acceptance of reality. We don't have it, so the short term issue is how to make the best of the situation.

The reason we don't have it is lack of incentive for the tracks. Most people aren't aware of the problem, and those who are aware and care about it, still play. We can continue to try to push for better timing, but in the meantime we have to deal with it as it is.

You are correct, it is not about negativity, it is about customer demand and apparently the customers (bettors) are not demanding such info.

Solve the equation and two things will become evident. First you will find out the longer the race, the less impact of the run-up distance and the closer the start is to the turn the run-up has more impact.

But in either case, it is not as significant as some other factors.

[cj]

You are correct, it is not about negativity, it is about customer demand and apparently the customers (bettors) are not demanding such info.

Solve the equation and two things will become evident. First you will find out the longer the race, the less impact of the run-up distance and the closer the start is to the turn the run-up has more impact.

But in either case, it is not as significant as some other factors.

That's funny...as if bettors demands are ever taken into account.

[Cratos]

That's funny...as if bettors demands are ever taken into account.

I don't want to go awry of the thread's topic, but my response was in reply to the inference in the poster's message which I concur.

However what should be done and who should do it?

[cj]

I don't want to go awry of the thread's topic, but my response was in reply to the inference in the poster's message which I concur.

However what should be done and who should do it?

Horse racing should enter the 21st century before the 22nd gets here. It isn't that difficult. Time the full race, not just most of it.

If Usain Bolt's 100m races were timed with run up, he would never hold the world record.

[Cratos]

Horse racing should enter the 21st century before the 22nd gets here. It isn't that difficult. Time the full race, not just the the most of it.

If Usain Bolt's 100m races were timed with run up, he would never hold the world record.

While it wouldn't make any difference to me, there are two ways this situation could be resolved.
(a) Put a second timer beam at the start and sync it with the gate opening. This would measure the time of the first horse from the gate to the start, but it wouldn't measure distance.

(b) Use a Trakus type methodology which puts a sensor on each horse and this would allow for both time and distance to be measured.

I would think that the first option might be easier to get done because it is less expensive.

[ronsmac]

Horse racing should enter the 21st century before the 22nd gets here. It isn't that difficult. Time the full race, not just most of it.

If Usain Bolt's 100m races were timed with run up, he would never hold the world record.
With his running style wouldn't he be winning and seperating from his rivals more than he does without a run up? If the 100 meter finals was 105 with runup. The early leaders would be decelerating at the end more than Bolt and he'd win with a larger margin considering he finishes stronger or decelerates to be more precise less than any runner in the world. The extra 5 meters would be to his advantage. Maybe I'm misunderstanding you're statement.

[ronsmac]

Horse racing should enter the 21st century before the 22nd gets here. It isn't that difficult. Time the full race, not just most of it.

If Usain Bolt's 100m races were timed with run up, he would never hold the world record.Oh, I see what you're talking about now, I stand corrected. I'm no english major.

[cj]

Oh, I see what you're talking about now, I stand corrected. I'm no english major.

I probably wasn't that clear. He is a slow starter, so somebody else would always trip the beam, adding to his time. That is all I was saying. He could run the exact same race, one alone, and one with a field of runners, and he would get two very different times.

This is what happens in horse racing every day.

[classhandicapper]

I really don't get all the negativity about this. It is 2013. There is no reason we shouldn't have the exact distance run and the exact time it took to run it. End of discussion as far as I'm concerned.

I think the the run-up is way more important for pace analysis than final time analysis.

There are at least a dozen small issues like run up that impact making final speed figures at the margin. Unless they all line up in one direction, we are typically debating about relatively small degrees of fast or slow when trying to account for them.

But run up can have a huge impact on the first couple of fractions and alter your view of whether the pace was fast or slow.

[banacek]

I think the the run-up is way more important for pace analysis than final time analysis.

Depends on the situation. From my analysis 6 furlongs at Hastings with a 100 foot run up is roughly .81 seconds faster final time than with a 24 foot run up. (CJ probably has more precise values)

Obviously the pace is seriously affected as well, but 4/5 of a second at a sprint distance is close to 5 lengths. That's an immense difference.

[classhandicapper]

Depends on the situation. From my analysis 6 furlongs at Hastings with a 100 foot run up is roughly .81 seconds faster final time than with a 24 foot run up. (CJ probably has more precise values)

Obviously the pace is seriously affected as well, but 4/5 of a second at a sprint distance is close to 5 lengths. That's an immense difference.

That's quite a lot.

I'm going to guess that there is at least some level of balance at the end of the race where the longer run up means they are tiring for slightly longer at the end. So maybe that run up is worth even more than 4/5 to the first quarter, which is gigantic.

[johnhannibalsmith]

While it wouldn't make any difference to me, there are two ways this situation could be resolved.
(a) Put a second timer beam at the start and sync it with the gate opening. This would measure the time of the first horse from the gate to the start, but it wouldn't measure distance.

(b) Use a Trakus type methodology which puts a sensor on each horse and this would allow for both time and distance to be measured.

I would think that the first option might be easier to get done because it is less expensive.

The technology to time races from the moment the gate opens has been existence for quite some time. Quarter horses don't trip a beam. The starter's plunger not only opens the gate, but triggers the timer on the photo equipment - at least that's how QH racing has been timed in my experience. The problem in translating that to TB, I would guess, is the issue of internals and display, but hardly an insurmountable challenge.

[cj]

The technology to time races from the moment the gate opens has been existence for quite some time. Quarter horses don't trip a beam. The starter's plunger not only opens the gate, but triggers the timer on the photo equipment - at least that's how QH racing has been timed in my experience. The problem in translating that to TB, I would guess, is the issue of internals and display, but hardly an insurmountable challenge.

It would be pretty easy to do I'm sure.

[Cratos]

The technology to time races from the moment the gate opens has been existence for quite some time. Quarter horses don't trip a beam. The starter's plunger not only opens the gate, but triggers the timer on the photo equipment - at least that's how QH racing has been timed in my experience. The problem in translating that to TB, I would guess, is the issue of internals and display, but hardly an insurmountable challenge.

I understand what you are saying, but from an engineering point of view the design technology is not that much difference; it is just who “pulls the trigger.”

Although I think the bigger issue in getting the timing change is political. Would the horsemen want the change because they might want to stay with the “stated race” distance measuring concept?

However what I find very odd is that before Trakus was introduced the quantitative analyses in horseracing were made from the “stated race” distances which Trakus has proved to be incorrect in terms of distance travelled.

Therefore what is the thesis of this argument? Is it accuracy or precision or is it just an inquisition to satisfy a false perception about antidotal information. You can use mathematical statistics to determine the impact on any distance increment within in the race or the final time of the race and the good thing about this calculation is that it is transferrable from racetrack to racetrack and distance to distance by making the correct adjustments to the calculation.

[Cratos]

I think the the run-up is way more important for pace analysis than final time analysis.

There are at least a dozen small issues like run up that impact making final speed figures at the margin. Unless they all line up in one direction, we are typically debating about relatively small degrees of fast or slow when trying to account for them.

But run up can have a huge impact on the first couple of fractions and alter your view of whether the pace was fast or slow.

Aren't you just speaking of the shape of the curve with respect to the dissipation of the horse's energy.

Also didn't Sartin energy distribution concept prove this long ago?

[cj]

However what I find very odd is that before Trakus was introduced the quantitative analyses in horseracing were made from the “stated race” distances which Trakus has proved to be incorrect in terms of distance travelled.

I'm sure Trakus has found some distances were not correct as listed, particularly on turf, but how have they proved distances in general were incorrect? Or are you talking about ground traveled by each horse?

Therefore what is the thesis of this argument? Is it accuracy or precision or is it just an inquisition to satisfy a false perception about antidotal information. You can use mathematical statistics to determine the impact on any distance increment within in the race or the final time of the race and the good thing about this calculation is that it is transferrable from racetrack to racetrack and distance to distance by making the correct adjustments to the calculation.



If horse A wins in a field with a fast breaker that trips the beam quickly and finishes last by 50 lengths, he gets a totally different final time (and fractions) than if that horse were not in the race. There is no mathematical formula that is going to fix that. The start of the race timing is determined by one horse and one horse only. The start of timing should start with all horses at the same point. Without that, the false clockings will persist.

[cj]

Aren't you just speaking of the shape of the curve with respect to the dissipation of the horse's energy.

Also didn't Sartin energy distribution concept prove this long ago?

Yeah, this one is pretty easy, but you can't use a general fix. Horses are tiring at different rates depending on the pace, and in some cases are actually running faster with added distance.

[Cratos]

Yeah, this one is pretty easy, but you can't use a general fix. Horses are tiring at different rates depending on the pace, and in some cases are actually running faster with added distance.

That is why there is a curve.

[Cratos]

I'm sure Trakus has found some distances were not correct as listed, particularly on turf, but how have they proved distances in general were incorrect? Or are you talking about ground traveled by each horse?





If horse A wins in a field with a fast breaker that trips the beam quickly and finishes last by 50 lengths, he gets a totally different final time (and fractions) than if that horse were not in the race. There is no mathematical formula that is going to fix that. The start of the race timing is determined by one horse and one horse only. The start of timing should start with all horses at the same point. Without that, the false clockings will persist.

You are wrong because what you have established is the classical “what if” problem in statistics, but this is not about statistics and this is not about “fixing” anything; it is about the significance of the data. If you want to “fix” something why not encourage the racetracks to publish the body weight of each horse before the race. This would give a good indication to the bettors of the horse’s health from race to race and its energy potential.

I agree with your premise that “the start of timing should start with all horses at the same point.”

However that holds true if you are timing horses, but the racetrack is timing races; not horses and that is why they have fixed distances. A six furlong race to the racetrack is 3,960 feet and not the recorded distance that Trakus discovers through its tracking methodology.

To answer your question: “are you talking about ground traveled by each horse?” Yes I am because that is what needed if you are going to make a correct assessment of the horses’ speed in the race. Speed by definition is the ratio between distance and time. Until Trakus there wasn’t any horseracing data provider that I know of, giving that information.

[cj]

You are wrong because what you have established is the classical “what if” problem in statistics, but this is not about statistics and this is not about “fixing” anything; it is about the significance of the data. If you want to “fix” something why not encourage the racetracks to publish the body weight of each horse before the race. This would give a good indication to the bettors of the horse’s health from race to race and its energy potential.

I agree with your premise that “the start of timing should start with all horses at the same point.”

However that holds true if you are timing horses, but the racetrack is timing races; not horses and that is why they have fixed distances. A six furlong race to the racetrack is 3,960 feet and not the recorded distance that Trakus discovers through its tracking methodology.

To answer your question: “are you talking about ground traveled by each horse?” Yes I am because that is what needed if you are going to make a correct assessment of the horses’ speed in the race. Speed by definition is the ratio between distance and time. Until Trakus there wasn’t any horseracing data provider that I know of, giving that information.

If somebody wants to crusade for weighing horses, I'm all for it. I've been lucky enough to land a job where I have a voice in racing, even if it is a small one. I'll use it to address topics I care about, thanks.

As for "you are wrong"...umm, no, I'm not. The system is flawed. I've pointed out exactly why with details. If a system can give the same horse different times when it runs the same EXACT race, that is a flaw. It has nothing to do with significance. Time isn't a statistic, it is a measurement that should be precise. End of discussion.

Now, onto Trakus...

A six furlong race is 3,960 feet if a horse travels the shortest path to the wire. Trakus, while nice, has a rather large built in margin of error. Look at any straightaway and you will see the distance for the first two furlongs listed for each horse. This shouldn't measure more than a foot or two more than the actual distance for the 1,320 feet. I'm accounting for a little swerving, but even that is a stretch. Yet, you won't find a horse listed as having traveled less than 1,327 feet, and as much as 1,331, maybe more.

Try plugging some numbers into the Pythagorean Theorem and see how much a horse would have to drift, swerve, etc. to run 11 extra feet on a straight and you'll know there is an issue. The answer would be 171 feet, or 57 yards.

If you want to use the time and distance for each horse as reported by Trakus, great, but you should understand the system before you do. I'll be happy to help further if you need.

[Maximillion]

Depends on the situation. From my analysis 6 furlongs at Hastings with a 100 foot run up is roughly .81 seconds faster final time than with a 24 foot run up. (CJ probably has more precise values)

Obviously the pace is seriously affected as well, but 4/5 of a second at a sprint distance is close to 5 lengths. That's an immense difference.

Have very limited experience at Hastings, but dabbled a little in it over the summer and found it to be (for me) very unpredictable.

Thanks for posting the above...maybe its one of the (many) reasons I was confused about a lot of the races there.

[JustRalph]

CJ just validated my high school geometry teacher. Damn you CJ!

[EMD4ME]

I've never seen someone with a tape measure or ruler out there. So....how do we know a race had a 76 foot run up as opposed to 80 foot run up?

[stuball]

Does anyone know how far in advance the runups are posted as sometimes they change race to race --same distance...Would it be feasible that a trainer
cross enters his or her horse and runs in the race that the runup would tip the favor to his horse..just talking hypothetical here..am I crediting the trainer with having Einstein intelligence.(quantum theories)

Just kidding but some of these posts strike a nerve in my brain prompting these thoughts......

Stuball

[cj]

Does anyone know how far in advance the runups are posted as sometimes they change race to race --same distance...Would it be feasible that a trainer
cross enters his or her horse and runs in the race that the runup would tip the favor to his horse..just talking hypothetical here..am I crediting the trainer with having Einstein intelligence.(quantum theories)

Just kidding but some of these posts strike a nerve in my brain prompting these thoughts......

Stuball

Equibase is provided the run up positions before racing starts by the tracks allegedly, but that doesn't mean the actual run up will match. I don't think fans have any way of knowing until the gate is put in place.