New Guy Has Questions

[gentz]

Hi to everyone at Pace and Cap. I recently started with RDSS and although I am familiar with velocity ratings there are a few I would like a definition for if someone could help. They are E/EP#, L/LP#,DCL and FRACTALS. Any help would be appreciated. I also have a few questions for Ted. Is there any chance of manually being able to adjust a horses beaten lengths and what is, or will be in the future the purpose of the Custom Screen? Thanks in advance and I am looking forward to learning as much as I can, and sharing as much as I can with this interesting group of people.

[Ted Craven]

Hi Greg, and welcome!

On the Energy screen:

E/ep stands for Early Energy Potential - a blending of all fractional velocities balanced toward the first 2 fractions
L/ep stands for Late Energy Potential - a blending of all fractional velocities balanced toward the 3rd fraction
Tpp stands for Total Pace Potential - the combination of E/ep and L/ep

This is the velocity based counterpart to the EPR + LPR = TPR ratings and is arguably a more accurate portrayal of the balance of early/late energy distribution: factors definitely worth modeling by track distance and surface. The ranks of E/ep and L/ep, when summed, form the Balance figure shown on the BL/BL screen under the heading BAL. Thus, the lower the Balance, the better the rank components of BOTH the E and the L components, in other words. This is not to say that lowest is best, and this Balance is not a final decision point. Long study has shown that a Balance of 8 or above are poor Win contenders, though the relative distribution of a horse's energy, given today's matchup of energy disbursement styles, may suit it for in-the-money (e.g. high E/ep and low L/ep may fade but not die; or low E/ep and high L/ep may close against tiring Earlies but not enough to win, etc). The Tpp aggregate is a strong overall ranking, not unlike Total Energy (the simple sum of the 3 fractional velocities), and appears in the Supplementary Factors section of the BL/BL screen.

DCL (on the Velocity screen, and ranked on the BL/BL screen) stands for Deceleration - a simple measure of decline of velocity from the 2nd call point through the 3rd fraction. It is shown as a percentage figure (i.e. percent that the 3rd fraction velocity is of the 2nd call velocity), thus the higher the value, the less the horse decelerated. I say simple measure, as this calculation does not take into account how relatively fast or slow a line's 2nd call was, thus a horse travelling relatively slowly to the 2nd call would tend to decelerate less in the 3rd fraction than one where the 2nd call velocity was higher. Translation: a good deceleration number may merely indicate a slow horse slowing down less than the others - doesn't mean a good ranking alone, in a matchup with the other contenders today. Try to use the combination of TS (True Speed) plus Deceleration as a more realistic marker of whether the horse was going fast enough in the first place for its good deceleration to mean anything.

OR - use the V/DC ranking, the much more complex and (perhaps ultimate) measure of Deceleration relative to Velocity, which takes into account the above arguments and blends in the prime Bottom Line 'power number'.


Just a quick note on Fractals for now - they are a ranking of a collection of energy disbursement blends, Early, Late and Normal (Unbiased), depending on which energy disbursement style you judge the collection of contenders to run as. For example, if 3 Earlies (ESP = E, or maybe EP) and they are of fairly equal early energy ability, consider the higer ranked Late Fractal readout (i.e. because the Earlies will wear each other out). Perhaps more about Fractals at another time - Sartin expounded on them in some Follow Ups at some time if I can dig that up. You can model them and use them if they are useful, but you could also live without them.

The Custom screen is where I am developing new or modified report layouts, or have developed private readouts for users. So far, only I can put stuff in there. Down the road, who knows.

You cannot adjust the beaten lengths manually, at the moment. You could do this in Speculator (one of RDSS' ancestors) so I will provide that ability here as well - eventually. In the meanwhile, just mentally fade the velocity or other readouts a bit in the direction (up or down) to what beaten length adjustment you were thinking of making.

Hope that helps a bit, but keep it coming!

yours,

Ted

[gentz]

Thanks for the thorough answer Ted. It is very much appreciated. I would like to read more about fractals when you could find the time to try and dig up the Follow Ups. Greg

[lsosa54]

Originally Posted by gentz

Thanks for the thorough answer Ted. It is very much appreciated. I would like to read more about fractals when you could find the time to try and dig up the Follow Ups. Greg

Greg: I'm not sure Fractals is a concept I completely buy into but here are extracts of the two Follow Up articles on the subject. FU #43 is from around Jan 1994 and FU #44 is from around March 1994, so keep in mind the concept is from 14 years ago. Lou

[lsosa54]

Some further reinforcement of what Ted said regarding E/EP and L/LP in Doc's words from one of the manuals.

[Ted Craven]

Thanks Lou. That's the earliest decription of the Fractal concept Doc was working on, and I thought there might be something more in Issues 70+, because while he might have retained the name Fractals (because it sounded cool), the math behind the readouts in Val/Spec/RDSS is probably not the same. I think Doc thinks that in general Fractals was an inconclusive concept, yet left them in anyway because of demand. RDSS provides them for backward compatibility with Val and Spec.

I never grasped the additional utility of those readouts, compared to E/ep, L/ep, Tpp. Perhaps a model of those Fractal factors by track, surface, distance will point you to an extra win or exotic hit per cycle (or perhaps not). I would recommend treating them as an Oats, Peas, Beans collection of factors, i.e. whatever the heck they stand for, model them and use them if they seem predictive - just don't get exercised about the label. But remember to pay attention to when the Match Up in today's race trumps the model...

Ted

[lsosa54]

Originally Posted by Ted Craven

and I thought there might be something more in Issues 70+

Ted: Only other blurb I found on Fractals although I just did a quick eyeball scan search from FU 68 to 88. Lou

[Turbulator]

Originally Posted by lsosa54

Ted: Only other blurb I found on Fractals although I just did a quick eyeball scan search from FU 68 to 88. Lou

Makes perfect sense to me--not.

How can a simple number determine inter-fractional movement? Seems a bit arcane to me.

[Tim Y]

Benoit Mandelbrot (look him up at Wikipedia), a mathematician, was one of the first to practice fractal math in regard to ongoing pictures. Fractals are the optical representation of the math.

Read Chaos by Gleick to learn all about them,http://www.amazon.com/Chaos-Making-S.../dp/0140092501

The LARGER concept I think Sartin was getting at when he advised reading this book, was the ORDER that is hidden within what appears chaotic, an order that reproduces over and over in the fractal image. Sartin also suggested another good book called Complexity by Waldrop where the Santa Fe Institute mathematicians took the concept to the next level and showed self-ordering systems occurring along the edges of chaos. It was the parallel belief that the program brought out an ordered understanding of what appeared to be a chaotic event.
http://books.google.ca/books?id=Nb-h...um=1&ct=result