Been looking for

[steveb]

Originally Posted by denniswilliams

I just use .16. It could be more exact but it slightly alters the published beaten lengths and creates consistent ones for the last 800m split sectionals that they added last year. Gives me a fuller view of the race.



Close finishes are the norm in HKG - especially at HV. It seems that any number of horses in a given field have run a race that could win that day. Do you think speed or pace figures actually capture what I see as very small differences between these horses? And, if they don't, is it because the data is not exact enough? What would be needed; more precise timing; exact position data? .....

depending on jurisdication, it varies.
thus beaten 6 lengths here, will not equal 6 lengths there.
most might use .16, somewhere else .17, over there .15
if you use the individual final times, then it matters not.
not all places give the times from which the margins are derived though.


HV is weird IMO, i will get around to studying why one day, but the median/avg implied prob of the winners appears to be much less there.

speed and pace is much more obvious at Shatin.
in fact just about every factor is much more random at HV than Shatin.

[Nitro]

Originally Posted by castaway01

I think the point was Nitro again trying to turn a thread into "what's the point of posting because nothing works except my secret analysis of Hong Kong toteboards" rather than criticizing how Hong Kong times its races. But that's just me because it's happened in 348 threads.

I see your reading comprehension hasn’t improved very much. Too bad because you just might learn something if you even bothered to get involved with the racing in HK. But yes, the tote analysis I use is a proprietary system of discovering the playable entries in any given race. I’m not interested in picking winners. I’m interested in making winning plays.

As SteveB pointed out, nothing about horseracing and subjectively interpreting the related speed and pace information is perfect by any means. That’s why I prefer to generally hedge my betting by using multiple entries when Dutching, or playing Verticals. It’s too bad that so many get hung up when they see more than 1 or 2 posted selections per race. There’s more than one way to skin a cat and enjoy the potential profiteering that this game offers.

[PaceAdvantage]

Originally Posted by Nitro

There’s more than one way to skin a cat and enjoy the potential profiteering that this game offers.

Especially if the skinning occurs after the results have been posted.

EZ game

[Nitro]

Originally Posted by PaceAdvantage

Especially if the skinning occurs after the results have been posted.

EZ game

Sharp comment if you’re into imaginary world of fiction and drivel.

[cj]

I don't get these threads. I guy asked a legit question about something, and mostly it gets poo pooed for no reason because others think they know better.

Maybe the information would lead to nothing, but this is exactly how you learn things in the game. You go down a lot of blind alleys and many are dead ends. But, occasionally they are not.

[MJC922]

Originally Posted by cj

I don't get these threads. I guy asked a legit question about something, and mostly it gets poo pooed for no reason because others think they know better.

Maybe the information would lead to nothing, but this is exactly how you learn things in the game. You go down a lot of blind alleys and many are dead ends. But, occasionally they are not.

I wasn't clear about what the actual question was and after re-reading it I'm still not, but then when you answered it I thought that covered it.

My only history with track dimensions relates to wind and average angles of all pace segments relative to the stretch. In that case I had stretch lengths and circumfrences (is that a word?) of all tracks in a DB. When they were hard to come by I would result to google earth and the measuring tool. So a custom function had to be written to derive the average angles of all segments in the race in order to apply the wind variant to all segments then stitch those together and start looking at pace for step two.

If he's looking to estimate the paths out on the turn a decent way is what Ragozin mentioned in his book i.e. when they didn't have ground being done by anyone for a certain track they could estimate by looking at the lengths, the running positions and the post positions. I was doing this many years before I'd ever read the book so I smiled when I heard he would do that too. Basically if you know a horse is 4th but less than a length back on the turn and his post was furthest out from the other three horses you can say four path on the turn is at least in the ballpark. You can do this for every horse in the race and probably won't be too far off.

[Belkster]

Originally Posted by Nitro

Before you get too carried away or excited about a race running factor with limited impact and value you might want to take a peek at a somewhat related thread that I started over 5 years ago titled:

“A Critique of Speed: Fallacies in Creating Accurate Speed Figures”.
http://www.paceadvantage.com/forum/s...d.php?t=148582/

I stand by my original beliefs and as far as I know electronic tracking of these race horses really hasn't improved very much since it was mentioned back then. I do agree that workable technology that produces reliably consistent and accurate results could answer many speed and pace related questions.

(My how time flies!).
.
.

Even if such technology will ever be there, it still won't be able to predict changes in speed and pace. Or could it? I'm trying to find a practical use for this, not a statistical one.

[Belkster]

Originally Posted by MJC922

I wasn't clear about what the actual question was and after re-reading it I'm still not, but then when you answered it I thought that covered it.

My only history with track dimensions relates to wind and average angles of all pace segments relative to the stretch. In that case I had stretch lengths and circumfrences (is that a word?) of all tracks in a DB. When they were hard to come by I would result to google earth and the measuring tool. So a custom function had to be written to derive the average angles of all segments in the race in order to apply the wind variant to all segments then stitch those together and start looking at pace for step two.

Google Earth? It's a very large scale, don't you think? There could be a lot of inaccuracies.

[MJC922]

Originally Posted by Belkster

Google Earth? It's a very large scale, don't you think? There could be a lot of inaccuracies.

If I recall in most cases the information was available, whether in the american racing manual or online. So for example such and such track is a one mile oval with a stretch length of 990 ft. Finish line to first turn might not have been available, you could use the measuring tool to estimate. I only used it or needed to for some obscure tracks in north america where I couldn't even find the most basic information so it would be better than nothing. What I'm getting at with the OP is that if you have circumfrence and the full length of the stretch you can figure out how long the turn is. With length of stretch to the finish line you can also figure out how much of the final quarter for example was on the turn initially, which then changes the average approach angle of the segment. That angle plays into how a wind variant from the stretch (call it 0 degrees) is applied to other pace segments.

[ubercapper]

Originally Posted by steveb

depending on jurisdication, it varies.
thus beaten 6 lengths here, will not equal 6 lengths there.
most might use .16, somewhere else .17, over there .15
if you use the individual final times, then it matters not.
not all places give the times from which the margins are derived though.


HV is weird IMO, i will get around to studying why one day, but the median/avg implied prob of the winners appears to be much less there.

speed and pace is much more obvious at Shatin.
in fact just about every factor is much more random at HV than Shatin.

As a reference, and I can't speak for all speed figure makers in the U.S., but many are using six lengths per second, so .167 (rounded) seconds per length.

[steveb]

Originally Posted by ubercapper

As a reference, and I can't speak for all speed figure makers in the U.S., but many are using six lengths per second, so .167 (rounded) seconds per length.

i don't think it's valid to use any figure, because it's so variable.
but i was referring to the official margins, given by the various tracks, that as far as i know, are all derived from the times.

[Nitro]

Originally Posted by Belkster

Even if such technology will ever be there, it still won't be able to predict changes in speed and pace. Or could it? I'm trying to find a practical use for this, not a statistical one.

I believe it could, because you would no longer be relying on the typical static picture of beaten lengths at each point of call, but rather on the dynamic running of each individual horse during the entire race. There would be no statistics involved just pure unadulterated real time velocity profiles of every entry. Whether that technology ever comes to fruition is another story.