Saratoga - 2004

[dehere]

I thought I would start a thread for the this year's version of what is IMHO the greatest meet in all of horseracing or, as Tom Durkin enunciates it, the "spaaaah". I'm posting a couple of charts that I've done for the meet so far. The one that I'll post with this posting is a printout print out of some things that I have found regarding juvenile races, again based on last year's meet. I think its pretty self-explanatory.

REMOVED

Finally, before I go let me mention another interesting statistic. Using the new Robot feature in HTR with last year's races, I ran a little test using K-line of 2.0 or lower and got some interesting results. Essentially, based on all races run at the Spa last year, there were 72 plays and of those 72 plays 46% were won by horses with a k-line of 2.0 or lower and the ROI for all of those plays was 1.03. Now, when you narrow that down to all claiming races at the spa, there were 22 plays, 50% winners and an ROI of 1.24. The interesting factor is that, when you compare that data with all claiming races run during the month of June throughout the country (514 plays) you get a win percent of 35% and an ROI of 0.85. That tells me I kinda wana look out for that angle this summer.

Anyone else with anything interesting to share?

[dehere]

I said I was gonna post a couple of things regarding the upcoming meet. The second post was too big to get up there though so I'm only posting one page. If you are interested in the rest let me know and I'll email you a copy.

The post which I'm talking about here is a chart for last year's meet looking solely at post positions for fast dirt sprint, fast dirt routes, wet dirt sprints, wet dirt routes and turf races. I have left two year old races out of all of these charts.

The three lines at the bottom of each chart are, in order of top to bottom, the number of winners at each post position, the number of "available races" at each post position and the numbe rof the post position itself. The number of "available races" refers to the number of races run in the specified category for which that post position was "available". For example. there are fewer available races for post 11 than there is for post 5.

[Huguenot]

Careful,
last year's meet was extremely formful with 41% favorites winning. The year before it was different with more longshots so your models would have been far different with likely worse results.
Saratoga does not fall into an orderly pattern. You've got to adjust on the fly based on what's winning. Sometimes early speed is just huge -- the overriding factor. At Travers Day last year, you had to bet F1 in all the dirt races -- blindly.

Other times, it's not and some hot trainer is a key element. And it's not always the same.
I recall a few years ago Dale Romans came into Saratoga and couldn't lose. The last couple of years he's struggled more.

I would look at several years worth of data and be alert to short-term trends at the Spa.

[dehere]

Huguenot,

I am in the process of doing so and you are so right. For example, at first glance the k-line data is woefully different for 2002 than for 2003 (although the data seems to be pretty darn close for 2001 as compared with 2003). I'll be posting more stuff as I look at history a bit more and I certainly appreciate anything else anyone else has to offer.

[Huguenot]

I do recall that if the track is fast, every year, speed would hold up pretty well the first week at Saratoga. I'll be looking at that tomorrow. Doesn't mean wire to wire types win every race (closers still win heavily contested, very fast paced races), but strong F1 types win enough
at long prices to make it more than worthwhile.

Track the strong F1 types the first day or two and see what happens. (But tell me if you notice something) :)

If it does occur, it could disappear just as suddenly.

One trend from last year I want to track is that heavily bet Lukas maiden horses fared poorly.

I'd love to track whether good shipping trainers who had strong Julys did worse at Saratoga and trainers who had weaker Julys did better due to the normal form cycle of horses in their barns or they were starting to outrun their classifications. In other words the year Roman just killed the Spa out of the box, had he been hot the previous weeks or had he been keeping a barnful of tigers under wraps aiming for the meet?

[dehere]

Huguenot - what i can tell you about your Fr1 thought is this.

For 2001 there were 37 sprint races and 14 dirt route races run during the first week. For the sprint races 4 were won by the top Fr1 horse and 20 were won by the top 4 Fr1 horses. For the route races 3 were won by the top Fr1 horse and 11 were won by the top 4 Fr1 horses.

For 2002 there were 26 sprint races and 13 dirt route races run during the first week. For the sprint races 3 were won by the top Fr1 horse and 17 were won by the top 4 Fr1 horses. For the route races 3 were won by the top Fr1 horse and 9 were won by the top 4 Fr1 horses.

For 2003 there were 30 sprint races and 13 dirt route races run during the first week. For the sprint races 4 were won by the top Fr1 horse and 18 were won by the top 4 Fr1 horses. For the route races 1 was won by the top Fr1 horse and 7 were won by the top 4 Fr1 horses.

That means that for this three year period of time 12% of the sprints were won by the top Fr1 horse and 59% of the sprints were won by the top 4 Fr 1 horses. For routes 18% were won by the top Fr1 horse and 68% were won by the top 4 Fr1 horses.

By comparison, if you look at all 529 sprint races run at the Spa over the past three years, 92 (17%) were won by the top Fr1 horse and 320 (60%) were won by the top 4 Fr1 horses. For the 237 route races which were run at the Spa over that time span, 46 (19%) were won by the top Fr1 horse and 69% were won by the top 4 Fr1 horses.

That tells me that this indicator of early speed either does not truly reflect early speed or that early speed is not as clear cut during the first week as one might expect. To test the first part of this conclusion I also looked at beaten lengths data for races run during this first week. Using a criteria of being within 1.5 lengths of the lead at the 2 and 4 furlong marker in the race I noted that for sprints, 58% of the winners were within 1.5 lenghts of the lead at the 2 furlong marker and 64% of the winners were within 1.5 lenghts of the lead at the 4 furlong marker. Overall, for all sprint races run at the Spa over the past three years, 65% of the winners were within 1.5 lengths of the lead at the 2 furlong marker and 76% of the winners were within 1.5 lengths of the lead at the 4 furlong marker.

I did the same test with route races. In doing so I noted that 33% of the winners during the first week were within 1.5 lengths of the lead at the 2 furlong marker and 43% of the winners were within 1.5 lengths of the lead at the 4 furlong marker. Overall, for all route races over the past 3 years 41% of winners were within 1.5 lengths at the 2 furlong marker and 45% of winners were within 1.5 lengths at the 4 furlong marker.

This seems to demonstrate that there is no real advantage to early speed during the first week above what applies for the entire meet. So, is the factor you talk about Huguenot (which I certainly believe to be the case) all a myth or is my approach missing the point in one way or another?

[Huguenot]

Thanks for running the info.
This was several years back, not recently (it's been really rainy in the first week the past couple of years).