More Trakus problems, this time SA

[classhandicapper]

Originally Posted by cj

With Trakus I worry a bit about the ground loss that is reported. It is better than nothing, but I've seen some pretty big errors along the way.

It's not even consistent from track to track.

I glance at it while I'm doing my trip notes, but you more or less have to know how Trakus works at that track and there will still be some errors.

Still, though it can be helpful. It's just a shame there aren't resources being put into getting all this stuff right.

[Robert Fischer]

Originally Posted by cj

With Trakus I worry a bit about the ground loss that is reported. It is better than nothing, but I've seen some pretty big errors along the way. It probably would be a valuable addition but I'd have to learn to trust it. I certainly don't trust the timing with good reason.

With TPD, we don't really know. Nobody seems to have public access right now, though some people obviously do and I'm sure they are using it. I would. That doesn't really seem fair to the betting public since it is being used as official timing at several tracks. I can't find a link to buy it anywhere. I would hope it would be provided and soon.

The data is a little bit buggy.

Best use is to 'flag' certain patterns.
Then if you have a race that indicates the flag, you'd have to check and use common sense.

Sometimes those Trakus 'chicklets' behave erratically.

We're just talking about it, but the data is there, and there are some potentially valuable ways to manipulate it.

TPD is very interesting.

[iamt]

Of all of these different systems that have been installed, as far as I know there isn't a single one in the world that ever has the chiklets displayed prior to race start. They all only start to display them about 50 meters into the race.


There is a simple explanation for this in my mind. If people saw how inaccurate the plotting was when the starters were standing still and supposed to be in a straight line, they would lose any trust they had it he system.


That is the problem with applying this technology to a sport where an error of a meter or two matters, both form a timing perspective and if you then try to apply it to ground loss.

[Dave Schwartz]

Originally Posted by cj

No, I think they use radio transmission.

The new GPS timing systems, run by this group, uses satellites. It seems better than Trakus but it still has a margin of error and isn't perfect by any means. But what system is?

http://www.totalperformancedata.com/

I would have thought it would be far more accurate than a satellite that was so far away.

Thanks for the info, CJ.

[Suff]

After the GS 2nd race , 5.5 dirt, Peter A was rolling through the payoffs and crescendo'd with "FINAL time of the race....pause....we'll have that in a moment?"

[cj]

Originally Posted by Suff

After the GS 2nd race , 5.5 dirt, Peter A was rolling through the payoffs and crescendo'd with "FINAL time of the race....pause....we'll have that in a moment?"

Garden State is open again!?!?!?

Just kidding, I'm a stickler for track codes!

[Suff]

Originally Posted by cj

Garden State is open again!?!?!?

Just kidding, I'm a stickler for track codes!

You’re right. Schooled! Lazy typing. I was thinking of spelling out Aelliole as the guide post but I just figured out how to spell Collmuss recently. Good funny reply though !

[cj]

Originally Posted by Suff

You’re right. Schooled! Lazy typing. I was thinking of spelling out Aelliole as the guide post but I just figured out how to spell Collmuss recently. Good funny reply though !

Did you though?

[Suff]

Originally Posted by cj

Did you though?

It was part of the joke no?

[cj]

Originally Posted by Suff

It was part of the joke no?

I figured, but I'm feeling rambunctious.

[jay68802]

Originally Posted by cj

I figured, but I'm feeling rambunctious.

Go get em tiger.

Don't worry Suff, he busted me a few times also.

[GBL]

Originally Posted by Robert Fischer

IDK about 'timing'

but give me

1. the trakus (or Total Performance) Data...
2. and, a decent programmer

and I could manage the construction of a model that flags 'meaningful' ground-loss, and a model that flags 'significant' race-flow patterns.


I believe that those would be useful tools.

Hi, Robert
Any idea what you're trying to accomplish here?

Been downloading and using Trakus data manually for years for variety of purposes using VBA for Excel. Pretty much everything that I could think of using Trakus data including data mining, reports, customized charts, ground loss, and several other velocity handicapping tools.

Wondering if you have newer ideas which I haven't yet discovered.

Thanks.

[castaway01]

Originally Posted by cj

Garden State is open again!?!?!?

Just kidding, I'm a stickler for track codes!

It's definitely not open, but there are some luxury apartments and about 100 chain restaurants ready for you to stop by.

[cj]

Originally Posted by castaway01

It's definitely not open, but there are some luxury apartments and about 100 chain restaurants ready for you to stop by.

Sad, I spent a lot of my 10 years stationed at McGuire in Cherry Hill for both GS and other things too. Good mall, great night life for which I always preferred to get away from the base.

[Robert Fischer]

Originally Posted by GBL

Hi, Robert
Any idea what you're trying to accomplish here?

Been downloading and using Trakus data manually for years for variety of purposes using VBA for Excel. Pretty much everything that I could think of using Trakus data including data mining, reports, customized charts, ground loss, and several other velocity handicapping tools.

Wondering if you have newer ideas which I haven't yet discovered.

Thanks.

Hi GBL,

Those are two models that could occasionally be significant.

Ground loss is the easy/obvious one... You're basically starting by flagging All ground loss..., and then filtering out only the instances where the ground loss happened in places or scenarios, where the ground loss actually matters in races.

The Race Shape, pattern recognition stuff is a little more complicated, and requires a competent programmer, probably with some related experience(in data patterns), who can follow my instructions and translate that into an application/model.