Trakus data now available for Churchill

[cj]

http://www.churchilldowns.com/racing...capping/trakus

One thing I noticed on Derby day is that the Trakus times are consistently faster, and not by a small amount, sometimes 2 or 3 fifths of a second for a race.

[Robert Fischer]

I noticed that as well.

Is there a run-up or similar that isn't being accounted for?


side note /You would expect some minor differences, and whenever I see them always match perfectly, I assume they have decided to use Trakus as their primary timing device. However, these differences did appear more significant from what I remember.

[elhelmete]

There are instances, and this is one of them, where I get this sinking feeling about how much we rely on bedrock data (time) that often proves to be so inaccurately measured.

[cj]

Originally Posted by Robert Fischer

I noticed that as well.

Is there a run-up or similar that isn't being accounted for?


side note /You would expect some minor differences, and whenever I see them always match perfectly, I assume they have decided to use Trakus as their primary timing device. However, these differences did appear more significant from what I remember.

Trakus times from the gate, but the time of the run up is not reported and is subtracted from what they put in their data. Tracks, from what I have been told, force this upon them.

The more I looked, and it brought back memories of looking at this before, most of the difference comes in the first 1/4 mile. So that tells me the run up part of the equation is off somewhat for Trakus. Electronic timing in place is accurate in most case, barring something tripping the beam early or human error.

[Tom]

Tracks demand that errors be presented as facts.
Why buy Trakus is you refuse to use it?

[Cratos]

Comparing the sensor timing of horseracing to beam timing is like comparing apples and oranges.

The sensor timing is measuring both distance and time; and therefore the resultant will be speed and this is done algorithmically with derivatives because both X ( the distance from the start) and Y (the distance from the rail) must intersect to determine position/location; this is a nonlinear calculation.

The beam timing is often called "fixed post" timing because each timed point is at a fixed post of the race distance and it is not necessary to measure distance because it is pre-measured and therefore using this methodology distance is always exact with recorded time being variable.

Given this difference in measurement methods there is always a possibility of different time for the expected same race point and if they are the same at any race point, it is coincidental.

Why the sensor technology is not displaying race distance runup? I suspect from my knowledge of working with sensor design it would take a more complex and expensive sensor because of the variability in the runup distance for the different race distances.

However I firmly believe with Longine entering the timing of horseracing, fixed post timing will eventually become obsolete.

[Robert Fischer]

Originally Posted by Cratos

Comparing the sensor timing of horseracing to beam timing is like comparing apples and oranges.

The sensor timing is measuring both distance and time; and therefore the resultant will be speed and this is done algorithmically with derivatives because both X ( the distance from the start) and Y (the distance from the rail) must intersect to determine position/location; this is a nonlinear calculation.

The beam timing is often called "fixed post" timing because each timed point is at a fixed post of the race distance and it is not necessary to measure distance because it is pre-measured and therefore using this methodology distance is always exact with recorded time being variable.

Given this difference in measurement methods there is always a possibility of different time for the expected same race point and if they are the same at any race point, it is coincidental.

Why the sensor technology is not displaying race distance runup? I suspect from my knowledge of working with sensor design it would take a more complex and expensive sensor because of the variability in the runup distance for the different race distances.

However I firmly believe with Longine entering the timing of horseracing, fixed post timing will eventually become obsolete.

You seem to be making trakus sound a lot 'smarter'(more complicated) than it actually is.

[Cratos]

Originally Posted by Robert Fischer

You seem to be making trakus sound a lot 'smarter'(more complicated) than it actually is.

It is not a matter of being "smart/complicated" as you characterized; it is what it is and that is a RFID chip technology which can be very smart if you invest the money.

This is not new technology, it has been around a long time. What is more interesting is Longine's claim that they can measure at an interval of 5 cm between horses.

At any rate, sensor measurement is superior to fixed post measurement and the reason being establishing a redundant straight line between two points except in theory is virtually impossible.

[cj]

Originally Posted by Cratos

Comparing the sensor timing of horseracing to beam timing is like comparing apples and oranges.

The sensor timing is measuring both distance and time; and therefore the resultant will be speed and this is done algorithmically with derivatives because both X ( the distance from the start) and Y (the distance from the rail) must intersect to determine position/location; this is a nonlinear calculation.

The beam timing is often called "fixed post" timing because each timed point is at a fixed post of the race distance and it is not necessary to measure distance because it is pre-measured and therefore using this methodology distance is always exact with recorded time being variable.

Given this difference in measurement methods there is always a possibility of different time for the expected same race point and if they are the same at any race point, it is coincidental.

Why the sensor technology is not displaying race distance runup? I suspect from my knowledge of working with sensor design it would take a more complex and expensive sensor because of the variability in the runup distance for the different race distances.

However I firmly believe with Longine entering the timing of horseracing, fixed post timing will eventually become obsolete.

It is pretty hard to beat the accuracy of beam timing barring something triggering a time to be registered early or human error. When it functions properly, the difference between those times and Trakus times is going to be, in my opinion, an error of some sort in the Trakus system. Something isn't being measured as precisely as it is with the beam system. Distance or time, it is one or the other.

You are not correct on why Trakus doesn't display run up distance and times. They are not allowed to do so. The information is available in their system. I have seen it with my own eyes. They do not rely on "reported" run up distances, they use the sensors to measure it. Trakus times the entire race, uses the distance recorded for the total race, subtracts the "official" distance, and the remainder must be the run up. They subtract the time of the run up from the total time to get an official time.

I personally do not think Trakus measures distance as precisely as is widely believed. Time is time, not difficult to measure. But trying to calculate distance based on horses carrying sensors, horses that don't run straight lines, is not nearly as easy. I've also detailed the rounding error that causes the distance horses travel to be overstated by 7 feet per two furlongs many times. It is undeniable.

[cj]

Originally Posted by Cratos

At any rate, sensor measurement is superior to fixed post measurement and the reason being establishing a redundant straight line between two points except in theory is virtually impossible.

The point between the beams never changes. Not any different than something like track. The 100 meters in the Olympics (and any other big meet) is always run at 100 meters. The distance doesn't vary. Doesn't seem virtually impossible to me.

[Cratos]

Originally Posted by cj

The point between the beams never changes. Not any different than something like track. The 100 meters in the Olympics (and any other big meet) is always run at 100 meters. The distance doesn't vary. Doesn't seem virtually impossible to me.

I am not going to argue or debate the points that I made in my earlier posts. I have studied and designed in sensor technology environments enough to make such statements.

Does this make me an expert? No it doesn't, but I damn well know what I am talking about both from science and experience; and can prove it.

It seems to me when Trakus is mentioned the same rehortic is given based on conjecture.

From what I understand Trakus was started in part by some people from MIT which should make it easy to get to someone in the know at Trakus through the alumni association. Also the president of Trakus is an electrical engineer and I believe his Masters is from MIT.

I only mentioned this because we are now talking engineering and I am very comfortable with my aforementioned statements about RFID chip timing versus beam timing.

However if you don't like Trakus; don't use it.

[Tom]

if you invest the money.

5 pretty big words there.

[cj]

Originally Posted by Cratos

However if you don't like Trakus; don't use it.

I never said I don't like Trakus. I'm merely stating that it isn't perfect. It is not as precise as many people pretend it is. Funny how you clam up when you can't dispute what is posted.

[classhandicapper]

There's another ground loss issue over and above the ~7 feet per 2 furlong question. The ground covered at some distances is almost certainly not accurate.

I was tracking some of the data for close to 4 months and found that the ground loss was very overstated for 1 specific distance at 1 specific track. Since I wasn't following it every day for every track, I can only assume there could be other distances that are "off" for some reason.

I still find the product very useful when used in combination with the DRF Charts and replays. When used in combination it's easy to figure out where all the horses were and take trip notes the first time you watch the race. It cuts down on the workload even if you don't use it's actual ground loss or times.

[cj]

Originally Posted by classhandicapper

There's another ground loss issue over and above the ~7 feet per 2 furlong question. The ground covered at some distances is almost certainly not accurate.

I was tracking some of the data for close to 4 months and found that the ground loss was very overstated for 1 specific distance at 1 specific track. Since I wasn't following it every day for every track, I can only assume there could be other distances that are "off" for some reason.

I still find the product very useful when used in combination with the DRF Charts and replays. When used in combination it's easy to figure out where all the horses were and take trip notes the first time you watch the race. It cuts down on the workload even if you don't use it's actual ground loss or times.

This is 100% correct. (Other than the DRF part, I prefer TimeformUS )