Quote:
Originally Posted by Buzz
Hi Everyone,
I am looking for a definition of the Perceptor. Does anyone know the definition of the Perceptor and how it is computed?
Thanks,
Buzz
|
Buzz,
Perceptor is a concept carried forward from Speculator (and from earlier software like Energy!) which shows factors in terms of their deviation from best, with best given a 0% (i.e. 0% difference - it
is the best). The caption on the Primary Factors screen gives a clue:
In this example, each of the 7 Primary factors (for the PPs of horse itself, in this case) are shown with the 0.0 (in RED) representing the best for that factor. Big Brown's 2008 Derby PP shows that his last race FX was his best FX (0.0%). His second line FX was 0.1% worse than his best.
Part of the idea for showing factors like this is to reveal just how
close or how
separate differently ranked numbers are. In the FX example, the 2nd line FX is ranked #2 but it is only 0.1% different from the best. That small a difference (1/10th of one percent) is
insignificant, and is
indistinguishable from chart-caller error. It could really be co-ranked #1. However, the 3rd line EPR is 21.5% worse than the last line EPR (the best, at 0.0%). Though this EPR is ranked #3 it is a LOT different than the #3 ranking of the FX for line 3, which is only 1.5% different from best. (In this example, we're comparing a Turf line 3 to dirt lines, so the big differences in EPR and LPR compared to the subsequent dirt lines is understandable.) Perceptor style numbers do reveal that not all
ranks, which are shown on the BL/Bl screen, and go into making up the Primary Line Score and thus the BL/BL ranking - are created equal.
In many cases, the actual number values of the factors (which are shown on other screens) are irrelevant (they're a kind of
magic number - we're interested in the
ranks and the
relative differences, or
gaps more than we're interested that LPR is 92.6, for example). So Peceptor does away with the underlying values themselves, and puts all factors on a comparable scale - namely percentage difference from best - then compares the ranks (top 3 coloured) and the gaps in an impactful way.
Perhaps you wondering about
Perceptor Total? Well, carrying the concept forward, it's a more fine-grained, alternate form of Line Score. The Primary Line Score is a sum of ranks of the 7 Primary Factors. But we've just finished noticing above that not all ranks are created equal. So Perceptor Total simply sums all the
percentage differences - instead of summing the ranks - then ranks that sum (the sum is not shown) in terms of difference from best. Again, in the above example, we don't really care that Big Brown's last line sums to 17.4 and represents the 0.0% (the best), and that the 2nd line sums to 23.0 which is 5.6 points worse (the figure shown) than 17.4.
Too much information! 
We simply want to know a rank, and a gap or margin. We want to know how similar or dissimilar lines are to each other, as that is how they will also differentiate themselves when compared against each other (either same horse, or different horses).
Showing the entire Primary Factor grid this way also reveals if
one factor alone comprises most of the difference in the ranking for the whole line score, which is the case for the 3rd line in the above example: the EPR 21.5 if by far the biggest (i.e. worst) member in the resulting line score. If this had resulted (in a different example) in ranking a horse (or a single horse) best but was predominately due to ONE factor, we'd like to know that, in case that factor is actually a
negative in today's race (e.g. stong LPR value in an Aqueduct Inner dirt sprint, or stong EPR in a Turf route (at least typically). That's the fine-grained part of Perceptor, as mentioned above.
The above example refers to Perceptor numbers as shown for a single horse, and I would recommend its use more for
selecting a representative line or lines (
along with Total Energy) on a horse by horse basis than for making final elimination decisions among compared horses. Eliminate to the top 5 Total Energy horses and you will almost always have the top 5 ranked Perceptor Totals as well.
However, there is a bit of magic in how BL/BL tweaks the results of that Top 5, as shown
very typically here when the #10 horse is elevated from 4th ranked Perceptor Total to #2 ranked BL/BL, and at 20-1 is a bet in your sleep (and a
$41.00 winner
)
Moral: use Perceptor figures moreso to identify a line or lines to represent a horse in today's Analysis set, and less so to identify wagers - leave that final decision making to BL/BL rankings and/or energy disbursement screens like the Segments screen and the Early/Late graph (and of course the offered odds, and your knowledge of factor bias for today's track, surface and distance).
Hope that shines some light on derivation and usage or Perceptor figures.
Ted