Agenda of statistical studies
Dave
We're just getting started here - thus far we have a small one on
"Races Since Last Claim."
Any requests?
The Knacker
I'd like to see some of the form factors evaluated. I remember back
in the day, about 20 years ago, before computer handicapping took off,
any of us hard-working guys who were willing to do pace handicapping
and calculate feet per second velocities by hand (Tom Brohammer's
"Pace Handicapping") had a huge advantage. It was like having a
crystal ball and the prices were juicy! Now everyone has the numbers
and they have mostly lost their value. So now I like to look at other
data more like the various form factors, looking for some angles.
1.) How about crunching some numbers to evaluate days since last race
such as 5 or less, 6 to21 days 22 days to 30 days, 31 to 42 days off
and that kind of thing. Do this regardless of works.
Then using the same time frames, include workout info such as off 22
to 30 days with 1 workout, with 2 workouts etc. Off 31 to 42 days with
1 WO, 2 WOs, etc.
2.) Have you ever done a study to look at patterns in closing odds in
the last 3 races to look for patterns? I always thought this would be
interesting to evaluate. In other words you could look to see a.)
what percentage of winners drop in odds in their last race out
compared to two races back, b.) pattern of last 3 races final odds.
3.) Looking at the last race, some other factors to evaluate would be
running positions and lengths back at various calls. For example, I
think most winners are probably within 4 lengths of the front at some
point in their last race. a.) For example, how many winners were
within 4 lengths of the lead or closer at the stretch call of their
last race? b.) How many winners are within 4 lengths at one of the
first 2 calls in the last race?
4.) Finish position of last 3 races, looking for patterns. I happen to
like any horse that has just finished 3rd, 4th, or 5th in all of its
last 3 races. These often win at a nice price.
Many more ideas, but this is a start.
Jim
Dave
such as 5 or less, 6 to21 days 22 days to 30 days, 31 to 42 days off
and that kind of thing. Do this regardless of works.
Then using the same time frames, include workout info such as off 22
to 30 days with 1 workout, with 2 workouts etc. Off 31 to 42 days
with
1 WO, 2 WOs, etc
That's a good idea. I think what we'll find is that "days off" is so
trainer-specific that it will show very little. But "Days since last
work or race" combined with work patterns might have merit. The
drawback to these is that they are in total control of the trainer.
.
2.) Have you ever done a study to look at patterns in closing odds in
the last 3 races to look for patterns? I always thought this would be
interesting to evaluate. In other words you could look to see a.)
what percentage of winners drop in odds in their last race out
compared to two races back, b.) pattern of last 3 races final odds.
No, although I watched a guy in the racebook years ago manage such a
system. He went broke. Not saying it doesn't have merit but I think
there is just too much ambiguity.
However, couple this idea with "back at the same class or lower" and
we might find some interesting results.
3.) Looking at the last race, some other factors to evaluate would be
running positions and lengths back at various calls. For example, I
think most winners are probably within 4 lengths of the front at some
point in their last race. a.) For example, how many winners were
within 4 lengths of the lead or closer at the stretch call of their
last race? b.) How many winners are within 4 lengths at one of the
first 2 calls in the last race?
Now THIS ONE I really like.
4.) Finish position of last 3 races, looking for patterns. I happen
to
like any horse that has just finished 3rd, 4th, or 5th in all of its
last 3 races. These often win at a nice price.
Similar to #3, Like this one as well.
Good ideas, all.
pmannet
Dave
pmannet
On Apr 9, 9:29 am, Dave <daveschwr...@gmail.com> wrote:
> What would you like to add?
I'd like to add a couple things
For #1. How about adding two workouts within 10 days of today's race
or lengths of workout equal to number of days or number of days minus
1. For example a 5F workout within 5 or 6 days. Ed Bain looked for
the latter but I never saw any stats on it. I also pay attention to
works prior to last race if didn't win. That race may have been the
conditioning race so it may be good to test that also.
For #3. How about within 2 or 4 at the first or second call, faded in
the stretch and then didn't lose more than 1-2 lengths stretch to
finish. Or lost positions and lengths 2nd call to stretch and gained
stretch to finish. Not sure if you can also test this when chart
comments show the horse 3 or more wide. I've often thought Maidens
with these patterns were better within the first couple races when
they're learning to run.
5) I'd also love to see anything on dirt to poly or vice versa. Right
now I look at early speed on dirt when going to poly, breeding and any
dirt works if possible. I really don't have much of a clue as to
which dirt horses will like poly but it looks like they run better
second time on it just like second time starters in Maiden races
6) I've always heard that horses in an outside post ran better if
they got an inside post next out. So perhaps horses in post 8 and up
going to post 1-3 split up by ran well last and out of money or more
than 3 lengths back at finish. And then vice versa.
pmannet
On Apr 10, 7:28 am, pmannet <pmannet...@gmail.com> wrote:
> On Apr 9, 9:29 am, Dave <daveschwr...@gmail.com> wrote:
>
> > What would you like to add?
>
> I'd like to add a couple things
>
> For #1. How about adding two workouts within 10 days of today's race
> or lengths of workout equal to number of days or number of days minus
> 1. For example a 5F workout within 5 or 6 days. Ed Bain looked for
> the latter but I never saw any stats on it. I also pay attention to
> works prior to last race if didn't win. That race may have been the
> conditioning race so it may be good to test that also.
>
> For #3. How about within 2 or 4 at the first or second call, faded in
> the stretch and then didn't lose more than 1-2 lengths stretch to
> finish. Or lost positions and lengths 2nd call to stretch and gained
> stretch to finish. Not sure if you can also test this when chart
> comments show the horse 3 or more wide. I've often thought Maidens
> with these patterns were better within the first couple races when
> they're learning to run.
>
> 5) I'd also love to see anything on dirt to poly or vice versa. Right
> now I look at early speed on poly when going to dirt, breeding and any
Dave
single dimension.
Example: Days since last race or days since last race or work cross-
referenced with furlongs worked/raced in 14 days.
#1 is too difficult to do from my current database but I can comment
on it.
Remember the old X-furlongs in X-days? (i.e. 14 furlongs in 14 days)
Well, there is just nothing there. In fact, typically horses that
"over work" their days disappoint. However, it would be good to show
how it works. I'll get to this one soon.
#3 - I can do stuff like that, within the parameters of 2-dimensions.
That is another good one.
#5 Dirt vs Poly. Still another good one. In fact, I should add that as
a factor in my system.
#6-Posts vary so widely by track that it would be like the guy that
wanted to see the difference between odds from last race to today's
race. A good basic idea, but so much of that is based upon class rise/
drop. Same thing with Post Position. An inside-to-outside study could
be pretty misleading if the track boasts an outside bias. (Like
Turfway, for example.)
So, the study agenda now looks like:
*** Furlongs Worked or Raced in X days.
*** Lengths Back at 2nd call, then lengths gained/lost in stretch.
*** Post Position Analysis
Thanks for the great ideas.
pmannet
for wound up tight and had two different levels or qualifiers calling
them WUT and WUTT. I gave up on it but do pay attention if a horse is
working more or faster. But overall I have no idea if horses do
better if they've had a work in 4 or 5 days etc and if it matters if
it's 4f or 5f.
He worked with someone by the name of Earl "Sigloff"? who had Pops
which had to do with odds drops. I tried a manual study on that also,
really concentrating on Maiden races but didn't come up with
anything. Maybe you will. I tried to make sure odds drops wasn't
because of a big drop in class, a good speed rating prior, less horses
in the race, by finish etc but didn't find anything that helped. I'll
be curious if you do.
On the running positions, one gentleman said a large sample showed a
flat bet profit on horses that were up close 1st call, lost ground 2nd
to stretch and didn't lose more than a length or two stretch to
finish. Even better if then gained stretch to finish or made 2 moves
in a race after losing ground or ran wide during the race. I gave up
on most running position angles...ran even, bid and hung, z pattern,
early and faded. Covered about every horse there was....
good points on the inside versus outside post.
Dave
What I have found is that there are tracks/distances that favor
"Inside-ness," some that favor "outside-ness" and even some that favor
"middle-ness." To further complicate the issue, it changes based upon
running style. The most obvious is that an E horse generally has an
advantage in the inside post but only if he can get the lead out of
the gate. By comparison, any other type of running style being inside
is generally a good thing.
Definitely worthy of a study. Let's do one now.