UPDATE: There are some discrepancies on the figures from Bris for the Remsen; the Bris figures I used came from their Handicappers Edge page, while two other sources list different figures for both the Remsen and KY Jockey Club (Link1 - also from Handicappers Edge, & Link2 at Hello Race Fans!) I'll go with superiority in numbers and take the other figures for now. It doesn't make a big change expect that the Remsen is now the 2nd highest figure for a 2yo graded stakes race in 2012 from Bris, but it's something to watch going forward.
ORIGINAL POST: The Daily Racing Form's Mike Watchmaker wrote a piece on the Beyer figures from last weekend Remsen and Kentucky Jockey Club, which produced a very good number (Overanalyze - 99 in a Remsen) and a very poor one (Uncaptured - 75 in the Kentucky Jockey Club). In his piece he looked at the reasons why a figure might be better or worse than the raw number, or what factors influence final time.
I highly recommend reading the piece, especially If you like looking at the theoretical side of speed figures. I want to highlight a couple of points in his post along with a slightly different analysis of the strength or weakness of each figure.
Let's look at Watchmaker's discussion of the Kentucky Jockey Club and possible explanations as to why the figure was so slow (as compared to previous figures from the top finishers in the race):
"A change in the relative speed of racing surface would be another likely culprit, but that doesn't seem to be in play, either. There was no "weather" at Churchill on Saturday, and the main track there isn't like Big Sandy at Belmont Park, which dramatically speeds up, or slows down, depending on how heavily it is, or isn't, watered."
Watchmaker is essentially writing about a split-variant situation, where the track speeds up or slows down during the course of a card, usually due to rain, wind or some other type of meteorological phenomenon (my fancy word for weather). The weather at Churchill was, for the most part, constant on Saturday.
Earlier this week, Daily Racing Form's Marty McGee tackled the low Kentucky Jockey Club Beyer figure in his Churchill Downs Notes story. A couple of excerpts from that piece:
Moreover, that 75 figure was exceeded by the winners of all 10 of the other races run Saturday on the 2-year-old-only program - five allowances and five maiden special-weights.
"I know it somehow looks like an aberration," said Andrew Beyer, who devised and continues to publish the figures in Daily Racing Form. "But I just couldn't find a way to fudge it. I just had to accept the evidence. The fact is the maiden [Bradester] ran faster than the stakes at the same distance [1 1/16 miles]. It's right in front of you, so what do you do with that?
McGee then spoke with Frac Daddy's trainer, Ken McPeek, who proposed his own theory as to the cause of the seemingly out-of-line figure:
"It was frozen in the morning, just crust," McPeek said. "In fact, we canceled all our breezes. I've seen it happen a lot with that track. As the day went on, it just didn't warm up much at all and didn't start getting looser until much later as they kept harrowing it and working it.
"I think the distortion in the [Beyers] is because of how the track just slowed down as it got looser. That was a really, really good group of horses you saw in the Jockey Club. I think that'll prove itself down the road."
If we take all of these statements in their totality, we are left with a prime example of why a handicapper should never treat a figure as a complete, 100%, non-subjective description of the race. It's not.
Ken McPeek is essentially arguing for a split-variant situation, that while the final time of the Kentucky Jockey Club was virtually identical to the Golden Rod, run just an hour earlier for juvenile fillies, that the track was becoming slower as the day went on, thus producing the low figure. Watchmaker's argument is, essentially, the evidence doesn't suggest this was the case. Furthermore, the Beyer figures, Bris figures and Equibase figures are all in agreement in assigning the same number for the Kentucky Jockey Club as they did for the Golden Rod.
Let's dig a little deeper into Saturday's card at Churchill and perhaps we can draw some conclusions of our own on what the heck was going on. (Note: I'm using the Trakus data for all of the numbers generated for Churchill's Saturday card, not the chart timing.)
|Race||Off||Approx Temp (F)||Approx Humidity||Felt Like||Dew Point||Barometer||Wind|
|2||1:09pm||35||52%||30||19||30.25||W @ 6|
|4||2:07pm||37||50%||30||20||30.23||WNW @ 9|
|6||3:07pm||37||50%||37||20||30.22||W @ 3|
|8||4:05pm||37||48%||37||19||30.21||NW @ 3|
|10||5:05pm||36||52%||31||20||30.21||WNW @ 6|
|12||6:05pm||32||61%||26||20||30.20||WNW @ 3|
The first thing to take note of is the fact that every race on Churchill's Saturday card was a juvenile race. That right there is something that's out of the ordinary. Second, as the other commentators have pointed out in their pieces, there were three races run on the main track at the 1 1/16 mile distance: the KY Jockey Club, the Golden Rod and a Maiden Special Weight in race 6.
I've included some weather data from weather.com, including temperate, relative humidity, dew point, barometric pressure and win speed. (These are readings from Louisville International Airport.) I pulled the weather data just to make sure there wasn't something strange that occurred during the Saturday card - i.e. an big swing in temperature, wind speed or humidity. As Watchmaker noted in his piece, the weather was fairly stable during the entire card, with the exception of a minor change in temperature and relative humidity.
If the Churchill main track slowed down as the day went on, as suggested by Ken McPeek, we would expect to see some kind of clear change in the timing of the races (or perhaps something weather related), but we don't. While the final times are all bunched together, look closely at the fractions and internal splits from each race.
In the three 1 1/16 mile races, the opening quarter went in 23.60, 23.62 and 24.15, with the final split coming in the Jockey Club. At first glance it looks like perhaps the main track was the cause of the slower opening split just an hour after the Golden Rod. But then look at the internal splits from the second quarter until the finish: the Jockey Club tracked right along with the other two races, including the fastest 4th quarter split (24.38) of the trio. If the track was slowing down (or becoming significantly tiring), I would expect to see much slower splits from the Jockey Club than the other races on the card, which just isn't the case. That doesn't mean that the track didn't slow down (it certainly could have) but the data isn't clear.
We have no objective data that suggests the track was slower later in the day than it was early in the day with respect to the two-turn races (or any of the races). What we do have is subjective conclusions - the Jockey Club is a Grade 2 race while race 6 was a garden variety Maiden Special Weight. Based on our experiences and knowledge of the sport, we feel like the Jockey Club will be the faster race and, thus, something must have happened to the track. Of course, assigning a grade to a race is a subjective endeavor based on a mix of historical results and conclusions. We also have the prior figures of the horses themselves (a point Watchmaker highlights in his piece), something that indicates all of the top horses ran much slower than their prior efforts.
Let's take a different look at the issue from the perspective of several different speed figures calculated for the Kentucky Jockey Club. Below is a summary of the Beyer, Bris and Equibase figures for the race, along with the rank of each in relation to all other graded stakes races for 2yo colts in 2012.
Beyer: 76 (24th of 26)
Bris: 93 (17th of 26)
Equibase: 90 (19th of 26)
A couple of things we can see right off the bat: first, the Beyer numbers are the lowest of the three when compared to other graded stakes races for juvenile colts this season. Second, the Bris and Equibase numbers are pretty similar in terms of rank. [The same hold true with the Golden Rod which tells us that neither Bris nor Equibase felt the track slowed down after the Golden Rod.]
If we have all three systems apparently agreeing that the track didn't slow down (at least when comparing the Golden Rod to the Jockey Club), then what do we make of the differences in rank versus all the other graded stakes races? Well, I think one of two things (or a combination thereof) could be occurring. One reason could be that the Beyer numbers for Saturday's card at Churchill are off due to a variant issue or some other computational decision. Another reason for the different rank could just be the general differences in the overall figures throughout the year. Honestly, I think we're seeing a little bit of both; I don't expect all three systems to always agree but when I see two of the systems essentially describing the same performance and one a little bit off on its own, you start to look for holes.
Turning to the Remsen, Watchmaker makes the following observation:
"Visually, this Remsen didn't look anything like the breakthrough race the final time says it was. Yet Overanalyze's 99 Beyer was 14 points higher than his previous career top over four prior starts (and, notably, 23 points higher than what he earned in his distant third behind Uncaptured in the Iroquois)..."
The Remsen is essentially the opposite problem from the Jockey Club: the race came back really fast - especially when you compare it to the prior races of the top finishers - but the data suggests nothing significantly altered the speed of the track during the day. If we once again look at each speed figure system and where the Remsen ranks as compared to all other graded stakes races for juvenile colts in 2012, we see a consistent pattern.
Beyer: 99 (1st of 26)
Bris: 106 (2nd of 26)
Equibase: 116 (1st of 26)
All three systems ranked the Remsen as the fastest graded stakes race for juvenile colts this season, and it wasn't even close. So while the Remsen figures might look high based on prior races by the top finishers, at least all three of the above figure makers were in agreement.
So where does all this "analysis", charts and tables leave us? I think the major takeaway is three-fold.
- I think it's good to take a big pictures look at the conclusions of various figure systems. Sure, it's possible that they are all wrong (maybe the Remsen wasn't the fastest juvenile race of the year), but at least we know what a majority of the players are going to be relying on going forward.
- No matter how hard we try, there is no such thing as a completely "objective" speed figure because...
- This is just another example of the significant number of subjective interpretations and conclusions factor into calculation a relatively "objective" speed figure.
Going forward, the strength or weakness of the numbers will become apparent, or perhaps it's more accurate to write "should" become apparent. Ultimately, the only conclusion we might come to is that the Jockey Club was just an oddly run race run over an oddly playing surface.