Author Topic: TG Preakness Figures 2007  (Read 1213 times)

TGAB

  • Administrator
  • *****
  • Posts: 1223
    • View Profile
TG Preakness Figures 2007
« on: May 24, 2007, 02:06:52 PM »
The attched file contains Thoro-Graph figures and sheets for the 2007 Preakness.
TGAB

Caradoc

  • Posts: 258
    • View Profile
Re: TG Preakness Figures 2007
« Reply #1 on: May 24, 2007, 03:47:33 PM »
Relying on the studies described in the Preakness commentary, and presuming my math is right, there was only a 5.7% chance that a) Street Sense would not back up more than one point, and b) Hard Spun would not back up by more than four points from their respective Derby efforts.  Further, that 5.7% chance may actually overstate the chances of both occurrences because 1) as the Preakness analysis noted, the studies lumped all horses with the negative effort in prior start together; it did not segregate those, like Street Sense and Hard Spun, which were coming back on short rest and thus were more likely to react significantly, and 2) the effect of the hot pace on Hard Spun.  Isn’t it more likely, then, as a pure statistical matter, that both of those horses ran 2-3 points slower than these figures indicate?  And further, why isn’t it far more likely that Curlin, who fires four zeros right out of the box and is now racing for the third time in five weeks, and for the first time ever on short rest, pairs his zero rather than jumps out of his skin?

alm

  • Posts: 800
    • View Profile
Re: TG Preakness Figures 2007
« Reply #2 on: May 24, 2007, 04:12:54 PM »
Visually speaking, that\'s what SS and HS appeared to do.

As for Curlin, he was running with a \'full\' tank as my old (crooked) trainer used to say.

miff

  • Posts: 6008
    • View Profile
Re: TG Preakness Figures 2007
« Reply #3 on: May 24, 2007, 05:03:03 PM »
Caradoc,

I think the figs look very good in relation to what happened on that day and in that race.

Why would you think it was more likely that the horses ran 2-3 points slower. What science tells us that horses with two weeks spacing or multiple fast races will absolutely regress? Did it look like Curlin only paired or that SS regressed. Forget what the figures say, what did you see?

Beyer, Rags and TG generally agree on the Preakness but the tops given by Rags seem to indicate that they had the derby too slow.


Mike
miff

TGJB

  • Administrator
  • *****
  • Posts: 10868
    • View Profile
Re: TG Preakness Figures 2007
« Reply #4 on: May 24, 2007, 05:07:14 PM »
a) I don\'t look at patterns when doing figures, let alone studies. I would never start with an assumption that a horse should go back. And in this case, since two horses in question cleared the field, that would be a particularly weak assumption to make.

b) There are other horses in the race that have to be looked at as well.
TGJB

TGJB

  • Administrator
  • *****
  • Posts: 10868
    • View Profile
Re: TG Preakness Figures 2007
« Reply #5 on: May 24, 2007, 05:15:11 PM »
Miff-- I haven\'t seen Rag sheets for all the Preakness horses, but I have seen them for the 4 that came out of the Derby, and based on those I have no idea how they got to those figures-- looks like they should be adding a point or two, based on their OWN figure histories of the horses. Certainly they can\'t be looking at surrounding races-- first of all it rained before the Preakness, and second of all, even if you didn\'t know that, the track changed speed a LOT during the day, just based on looking at the horses and how fast they ran.

Ragozin blew the BG and Derby by a lot, Preakness by less.
TGJB

Caradoc

  • Posts: 258
    • View Profile
Re: TG Preakness Figures 2007
« Reply #6 on: May 24, 2007, 05:22:33 PM »
a) Of course not, but that is not the point.  The point is that for any given race you know or determine how the horses run in relation to each other.  In the Preakness, you know that Curlin ran about a point better than SS, and about three points better than HS.  The issue is whether, given the studies described in the Preakness analysis, and b), it is more likely that Curlin ran about a zero and HS about a 3, or that Curlin ran a negative three and HS a zero.

b) Right.  So how likely is it, given Curlin\'s pattern, that he pops a negative number? How likely is it that CP West, who bounces three points off a top as a 2yo (a bounce that in the past has been identified on this board as a sure sign that a horse is hurt by big efforts), runs a new top 2nd out as a 3yo and then pops a 5-point new top off three weeks rest in the Preakness?  Isn\'t that the biggest new top (by far) you ever gave out in the Preakness?

miff

  • Posts: 6008
    • View Profile
Re: TG Preakness Figures 2007
« Reply #7 on: May 24, 2007, 05:30:53 PM »
JB,

Honestly, I had a little trouble with the derby because I felt the track was pretty fast and \"whole\" race came up a little on the slow side. Looking at it raw, against history, and the track speed for the day, I was slighty slower than you.

Re the Preakness, Curlin moved up app 12 lengths on Rags and SS about 6 lengths. I find that very difficult to reconcile on any racing level but I\'m no expert on how Rags came to their figs.

Mike
miff

TGJB

  • Administrator
  • *****
  • Posts: 10868
    • View Profile
Re: TG Preakness Figures 2007
« Reply #8 on: May 24, 2007, 05:33:25 PM »
The studies give insight into what is more likely to happen, not what happens in a specific case, and on Curlin, I already said I don\'t look at patterns. If you add 3 to the whole race it looks worse for more horses-- you would basically have most horses running well off their tops, except Curlin pairing, and Zito\'s horse moving forward. This way looks better, and I would be surprised if it doesn\'t hold up.

Incidentally, what you are saying I should do is exactly what Miff sometimes accuses me of doing-- fitting figures to preconceived ideas.
TGJB

miff

  • Posts: 6008
    • View Profile
Re: TG Preakness Figures 2007
« Reply #9 on: May 24, 2007, 05:35:12 PM »
Caradoc,

A trap in purely using stats in racing is that just because a certain outcome is highly unlikely,it is not impossible. Sooner or later much of what seems  highly unlikely or impossible, happens, albeit infrequently.


Mike
miff

Caradoc

  • Posts: 258
    • View Profile
Re: TG Preakness Figures 2007
« Reply #10 on: May 24, 2007, 05:40:28 PM »
No, as my original question below states, I am asking you why another outcome is not more likely, particularly in light of your own studies and approaches, and the figures and patterns of the individual horses involved.

Chuckles_the_Clown2

  • Guest
Re: TG Preakness Figures 2007
« Reply #11 on: May 24, 2007, 06:56:31 PM »
For what its worth, I\'m in complete accord with TGraphs work on the Preakness. They ran fast early and good horses ran fast late.

The only issue I\'d broach is the lack of an off rail notation.

On those numbers I\'m certain TGraph wants all three to run in the Belmont.

CtC

TGJB Wrote:
-------------------------------------------------------
> The studies give insight into what is more likely
> to happen, not what happens in a specific case,
> and on Curlin, I already said I don\'t look at
> patterns. If you add 3 to the whole race it looks
> worse for more horses-- you would basically have
> most horses running well off their tops, except
> Curlin pairing, and Zito\'s horse moving forward.
> This way looks better, and I would be surprised if
> it doesn\'t hold up.
>
> Incidentally, what you are saying I should do is
> exactly what Miff sometimes accuses me of doing--
> fitting figures to preconceived ideas.

MO

  • Posts: 1080
    • View Profile
Re: TG Preakness Figures 2007
« Reply #12 on: May 24, 2007, 10:42:55 PM »
According to the study of 3yos running a neg # before June 30, Curlin should be a huge bet against in the Belmont and at likely very short odds.

Street Sense wont be entered.

Hard Spun won\'t get the distance. CQ\'s negative 1 surely has knocked him out.

Belmont is ripe for a big score and I look forward to seeing the sheets of the other entries.

SoCalMan2

  • Posts: 1048
    • View Profile
Re: TG Preakness Figures 2007
« Reply #13 on: May 25, 2007, 04:49:56 AM »
I appreciate all the work that went into the studies and I very much enjoyed reviewing it.  However, I do not think it was terribly helpful as a handicapping tool as I mentioned before the Preakness --

http://www.thorograph.com/phorum/read.php?1,35349,35357#msg-35357



To reiterate, by using minus 1 as the barometer, we are limiting ourselves to a fairly small and recent sample.  The issue is not minus 1 which is just a number but rather what happens when yuong 3 year olds run a huge figure.  Recently, a huge figure for a young three year old has been minus 1, but 5 years ago, that same concept might have been a plus 1 and ten years ago it might have been a plus 3. (I am just pulling these numbers out of a hat -- the idea is that minus 1 means something in relation to current 3 year olds, the equivalent number for 3 year olds ten years ago is going to be differnet and, whatever the right number is, it should be included in the study in relation to that crop).  I am concerned that, if the same study were done over a longer period of time using a sliding reference for outsize performance, the results might not be as clear as the smaller sample showed.

Again, I do not want to seem like an ingrate.  I very much appreciate the study, and it is instructive.  However, I think it is important to consider context when reviewing such things

flushedstraight

  • Posts: 270
    • View Profile
Re: TG Preakness Figures 2007
« Reply #14 on: May 25, 2007, 11:12:45 AM »
MO Wrote:
-------------------------------------------------------
> Curlin should be a huge bet against in the Belmont and at likely very short odds.
> Hard Spun won\'t get the distance.

Agree regarding Curlin, looks like great value going against him as he\'ll be 3/2 or lower at Belmont. He finally ran what can be considered now as an effort for him, and get\'s back to zero or worse off 3 weeks.

Disagree about Hard Spun, who likely reacted last week to his big Derby but maybe gets back to neg # at Belmont, which should more so resemble his Derby, except with no Street Sense charging and a significantly slower 1st 1/4 with little speed signed up so far (at worst Tizzy? Chelokee?). Questions about distance limitations will hopefully be offset by decent price, which remains to be seen. I believe any questions about him being unrateable were squashed in the derby, but maybe this issue now resurfaces again after the Preakness try, resulting in a possible overlay situation.