Attached courtesy of Mark Coomes and Chris Johnson.
What we\'re looking for is the usual breakdown (new tops, pair, off, X), broken out by no CD works, one work, 2 or more, before the Derby.
Love it.
Here\'s another one I just read...
Horses who have earned $1mm prior to the Derby are 1-36 in the race, lone exception was STreet Sense.
You think that\'s comparable to the question of horses getting a work over the track?
Nope, just an interesting factoid I had no idea of, similar to this one. Since both have been going on for so long, kinda crazy they were not more well known.
Could be something at least somewhat interesting to it, however, as it suggests that horses that put forth enough effort prior to the race to win a million might be tapped out by Derby time (or are too brilliant in winning the shorter races).
That stat caught my eye too. Figured a horse has to start a fair amount to amass those earnings, so started looking at performance as a function of career starts. Since 2005, 13 horses with 10 or more career starts have entered the gate, with 11 running X\'s, one Off and one Top. Since 2009, it is 6 X\'s out of 6 runners.
Street Sense only had 7 starts before the Derby.
Jcip-- I\'m definitely interested in someone running THAT one going back further. Wow.
If you do it, please let us have a list.
If memory serves me correctly I beleive in 95 Tejano run who ran second had shipped in late.
Going back further with it tonight, will post what I come up with.
Any chances we can get the list of non-workers and off odds? Average is what 3 per yr?
Did a quick scrub of the data, found a couple 3.5pt Offs I had mistakenly noted as Xs. Invite others to point out any other errors you find.
Derby Performance by # of Starts (https://docs.google.com/spreadsheet/ccc?key=0Ahw9zHYCs7vAdFpBYld0d3lkWWdCUjl5dWtjX251ZFE&usp=sharing)
It\'s in there. Click the second tab, marked No CD Work.
Ok, did it manually, could only get pp\'s (to check where they last worked) dating back to 2010 (4 years worth) online. So ran those, for every horse in each derby. If anyone has older pp\'s or knows a site that has them, can run the older ones.
RESULTS 4 yrs - 78 horses
2+ workouts at CD - 32 horses
TOP 13%
PR 16%
OFF 25%
X 47%
1 workout at CD - 25 horses
TOP 12%
PAIR 24%
OFF 24%
X 40%
0 workouts at CD - 21 horses
TOP 0%
PAIR 19%
OFF 14%
X 67%
It\'s only 4 years worth but...that 0% tops for 0 works at CD does kind of stand out. But again, only 4 years worth. And actually, given what we know of how few run Tops in the derby, maybe it should be the other way we are looking at it, that having at least 1 work over the track seems to yield a better % of tops than the usual 8.5%
But again, really need some more years to confirm
One thing that did stand out going through these, and this is interesting for THIS YEAR\'S derby. All of Makers horses (5 in total, all OFF\'s or X\'s) before this year had everyone of their works leading up to the race at the Training Center. Not at CD itself. Whereas this year, he brought all three over, VIC and HH will have had 1, and if the weather allows tomorrow or the next day, AROD will have had 2. And more to the point, they will all have been on the grounds since last Tuesday (22nd) afternoon.
A couple of notes
-Backtalk, in 2010 had his 2nd last work at CD, but his last, for some insane reason, at KEE. I included him in the group that had exactly ONE work
-Giant Finish, 2013, brought in late by T. Dutrow, I am pretty sure he had 0 works over CD, that\'s what news reports indicated (was not in the pp pdf i had for it for some reason), if anyone knows if he did in fact have a late blowout a day or two prior to race over CD?
-Prospective, 2012, had his final 3 works over CD but with a race in-between, the bluegrass. So I took him as having just 1 (his lone work after the bluegrass) at CD.
Nice! This is great work. Still a small sample size, but bears watching if there is such a thing as \"too much foundation\".
I\'ll echo tread\'s props, nice work.
If we combine yours (10+ starts last 5 years yielding 0 Tops or Pairs for 6 horses) with mine (0 works over CD last 4 years yielding 0 Tops for 21 horses) it does sort of point to some possible questions for a certain horse who just shipped in today. Mind you we\'re playing with some seriously small sample sizes at this point.
Much appreciated guys...happy to finally contribute something after sponging up knowledge here for years.
Sample size notwithstanding, the combined question marks are hard to ignore.
Here you go - pp\'s back to 1992
http://www.trackphantom.com/KyDerby.html
bb
That is an awesome resource, not just for this but in general. Thanks. Going to get some pizza and then dig in.
The 2nd tab of the spreadsheet I sent lists all the horses since 1996 that did not work at CD, with off odds and trainer.
The 3rd tab compiles the #s for Did Work vs. Didn\'t Work.
Sorry, don\'t see it. See the PDF that JB posted, and the Goggle Doc that jcipoletti posted.
10 years worth, 2004 (the start of TG noting Thoro Patterns on the sheets, which makes it much easier for manually typing them in as you don\'t have to sit there for 10 sec looking at the sheet to figure out whether the horse ran a top, etc.. in the race) to 2013.
195 horses
2+ workouts at CD - 67 horses
Top 7...10%
Pair 7...10%
Off 18...27%
X 35...52%
1 workout at CD - 84 horses
Top 9...11%
Pair 18...21%
Off 17...20%
X 40...48%
0 workouts at CD - 44 horses
Top 2...5% (the two that did were Tiago & Giacomo)
Pair 10...23%
Off 6...14%
X 26...59%
So the trend held from the more recent, smaller sample set. 0 works over the track puts you under the average % of new tops, and 1 or more works over the track puts you over the ave % of new tops.
EDIT - To really feel surety on this, kind of would have liked to have seen the PAIRS (for the 0 works) be markedly less as well. Because just a couple of PAIRS getting bumbled up by a half point here or there and you\'ve got a regular dispersion for TOPS.
EDIT (II) - Just going through all these really brings home how insanely bad most of the horses in the derby run, year in and year out. It\'s kind of hard to believe...except that it keeps happening every single year.
A few thoughts:
1-Like I mentioned before there\'s a few, I\'d say about 10, horses that worked at CD, then ran the BG, then worked again once at CD. I\'m taking all these as just ONE work. But I\'m wondering if that\'s really right. Our entire point here is to focus on horses training over the track for an extended period of time. I wonder if we should consider these as 2+ works. Either way, it does not change the 0 works numbers, which I suppose is the important thing.
2 - High Limit - arggg. I know I know pace casualty, fastest derby pace ever, blah blah blah...but f**k...he\'d been training at CD since late March! with paired tops, with tons of space to move fwd still of his 2yo top, by Maria\'s Mon, in Frankel\'s hey day of the White Mercedes. How the hell did he run that bad. (I still hate you Giacomo).
3 - Lion Heart - doesn\'t really matter but, just fyi, his posted THoroPattern on the sheet is wrong. It says OFF, but he ran a 0 in the derby, after a -1 in the prep, so that\'s a Pair. I noted it as such for above.
phil-- interesting stuff, thanks for putting this together. Not sure you can really say the trend held from the more recent sample though. On an overall basis you still do see the trend (albeit to a lesser extent), but when you extract just the subset of the older sample from the data you posted, you get an entirely different picture:
2004 – 2010
117 horses
2+ workouts at CD - 35 horses
Top 3...9%
Pair 2...6%
Off 10...29%
X 20...57%
1 workout at CD - 59 horses
Top 6...10%
Pair 12...20%
Off 11...19%
X 30...51%
0 workouts at CD - 23 horses
Top 2...9 % (the two that did were Tiago & Giacomo)
Pair 6...26%
Off 3...13%
X 12...52%
In this older sample, the percentage of Tops are around the same for all three groups. And when you combine the Tops and Pairs, the group with 0 CD works is actually the best (35%), while the group with 2 CD works is by far the lowest (15%).
So it looks like either something's happened in recent years that's made it more advantageous to ship to CD early, or the sample size is just too small to draw any statistically significant conclusions.
Rocky R
My conclusion is that number of works is not a relevant factor.
Good work. 89.3% of those who had 10+ starts ran Off (17.9%) or X\'d out (71.4%). However We might see these numbers change going forward with the new qualifying system. I\'m guessing most of these 10+ start horses were ones who qualified early as 2 year olds and were out of form at 3 but their connections used them as their Derby ticket anyway.
Will 2014 be remembered as the year of the diluted Kool Aid Derby?
Extensive discourse on workouts and dosage on the TG board?
Seriously, great to hear from some new voices. The work on the EXcel charts is top
drawer.
Now all we need is a report from the Herd Whisperer, or one of his/her followers...
Indeed we do, emotional profile may be more important this year than ever with such a closely grouped field.
Here is the extended doc with the multiple tabs, someone sent it to me on twitter (thanks to @o_hoolix on twitter)
https://drive.google.com/file/d/0BxOE5bcnZPo6Yjh6ZlJiQ081T3M/edit?usp=sharing
An interesting stat, though likely apropos of nothing outside the realm of anecdotal, is that, from 2002-2013, the winner of the Santa Anita Derby only hit the super two out of twelve times. Now, since only four of twenty hit it every year, in raw numbers, that\'s not very disproportional. Nonetheless, one might assume that there would be a greater likelihood than pure random probabilities for the winner of a Grade 1 \"major.\" The betting public would concur as eight of those twelve horses went off at single digits, usually giving them \"top five\" consideration in most races. This becomes even slightly more surprising when you see the names O\'Neill, Baffert, Sadler, Mullins, Shireffs, etc., but then again all these numbers are superior to TAP\'s, thus so much for star power. I\'ll Have Another had been the longest priced entrant since Buzzard\'s Bay went off at 46-1 in 2005, a race in which he ran fifth. One can only wonder what that super might have paid with Giacomo on top, had Afleet Alex encountered a worse trip, and Buzzard filled the super with 50-, 71- and 29-1 shots. It paid $864K for a dollar, even with Alex. Of course, Giacomo did come out of the Santa Anita Derby, as well.
Good points Rocky and yep I hear you. Sample Size definitely does make it tough to say we\'ve proven anything.
Thanks, got it.
First of all, yes, if a horse is stabled at CD and has multiple works there, it doesn\'t matter whether there was another race in between or not. The issue here is familiarity with the track. They should be in the two WO category.
I haven\'t looked at this closely (a little busy), but I\'m having trouble understanding how the original reported results by finish (only 6% of horses with no works hitting super, 25% of those with works doing it) has not much correlation with how they run in figure terms. Seems to me for that to be true horses working at CD would have to start out faster than the other group.
Just took a quick look back to 2007, 31 runners total.
Have to go back to Tiago in \'07 to find even a small new Top among the Non-CD work group, moving forward 1.75pts. Master of Hounds (0.5) and Any Given Saturday (0.25) the only other two to move forward at all. I\'ll Have Another (0), Nobiz Like Shobiz (-0.5), Lines of Battle (-0.75), Circular Quay (-0.75) and Cowboy Cal (-1) are the other pair ups.
That\'s 1 small Top and 7 pair ups out of 31 runners. Only two - Dialed In and Great Hunter - ran Off, while the remaining 21 X\'d by an average of 13.5 points off their top.
Again there is the issue of sample size. I tell the basketball coaches I work with that findings from a small sample will not be conclusive, but they do fine tune the lens and give you something specific to look for in the film room.
I agree, would have thought we\'d see more correlation. Don\'t have a good explanation. Double checked each last night before posting but it is a manual type in thing, pretty sure all correct, but user error is not impossible here.
TGJB Wrote:
-------------------------------------------------------
>
> I haven\'t looked at this closely (a little busy),
> but I\'m having trouble understanding how the
> original reported results by finish (only 6% of
> horses with no works hitting super, 25% of those
> with works doing it) has not much correlation with
> how they run in figure terms. Seems to me for that
> to be true horses working at CD would have to
> start out faster than the other group.
The Non-CD Work group from 2007-2013 had an average top of 2.30 (std dev 1.88) coming in. Any idea how that compares with those that had a work over the track over that same timeframe?