OK, let’s see if we can explain “Phase 2″ of our breeding rating functionality with relative brevity…by our standards. The irony, of course, is that our mission is to help horseplayers of all levels understand races faster, but when we go to explain what goes into the breeding ratings recipe, we have to go on a bit. At least the outputs are easy on the eyes.
Breeding ratings phase 2 explanation, here goes:
How we come up with these:
We use the horse’s family history, as told through TimeformUS Speed Figures, at different distances and surfaces over the last decade+. Even if a sire hasn’t raced in 20 years, his offspring will typically tell a story about pop’s ability to sire runners with a certain type of surface or distance preference, as well as class (meaning whether he sires many or ANY fast horses). The same holds true for a horse’s mother (the dam), and mother’s sire (the broodmare sire), and we algorithmically weigh the values of all of those factors to make our breeding ratings. As we’re fond of saying, voilà. Here’s a link to a piece on how we build those basic ratings. Enjoy the read. Welcome back:
And…here’s the standard caveat for both that basic rating and this “Phase 2″ functionality:
Once a horse has a history of efforts over a surface, its past performances are more important than whatever it was bred to do. Bred to go long but keep failing in routes and excelling in sprints? Sure sounds like you’re a sprinter to us. Etcetera.
But before a horse has clearly established its preferences, pedigree ratings can be invaluable tools for telling you what a horse is most likely to succeed doing–routing, sprinting, dirt, turf, synthetic (super-excited to have finished these up just in time for two major synthetic tracks to be well on the way to going back to dirt), dry, wet. And in TimeformUS PPs for any horse, you can now click on any of the green text in this area for a more comprehensive understanding of what a horse is best-bred to do. Phase two.
Once you’ve clicked it, this display will come onto the screen:
As you can see, the display has 4 tabs. And no surprise, just like in our PPs, brown means dirt, green means turf, and blue means synthetic. OK, let’s go through the four tabs:
The first tab is “This Horse.”
This Horse is…today’s runner. This tab is the comprehensive ratings for today’s runner at different distances and surfaces. As described above and at the related link, we consider the family’s speed figure history and place today’s runner on a scale for different distances and surfaces. The advantage of the click-thru is that it allows you to see what a horse is best-bred to do, rather than seeing only the rating in the PPs, which merely allowed you to compare This Horse’s rating for today’s surface/distance to other runners’.
Three terrific plusses of seeing all the ratings broken down like this:
1) We have mud ratings you can compare to dry track ratings.
2) If there’s a surface change (typically, a turf race that moves to dirt, you can compare today’s runner’s turf and dirt ratings).
3) You can contextualize the ratings, looking for a possible preference. Example: Horse is rated in the 60s on dirt but 80s on turf? And is debuting on dirt today? Maybe it’s best to wait for him to try the grass.
Other notes on “This Horse,” some of which will be relevant to the other tabs as well:
- Yes, “wet turf” means something other than firm or hard turf courses. While it’s true that US horses don’t run on “softer” courses as much, we can indeed show you if a horse’s family history on-track shows a particular like or dislike to a grass course with “give” to it. You should see some slant towards the negative here, as domestic horses tend to run less than their best on more yielding surfaces, as training and running on this type of surface is the exception rather than the norm in North America.
- The norm, of course, is dirt sprints. 2/3rds of all North American races are main-track sprints. So you’re likely to see pedigree ratings more tightly bunched for dirt sprints. Because these races are overrepresented, we tend to breed for them, and vice versa. Given the preponderance of dirt sprinting, a very strong dirt-sprint rating will have the effect of impacting a horse’s overall dirt rating. So we encourage you, when looking at “This horse,” to contextualize the dirt sprint ratings with ALL of the other ratings for This Horse.
The second tab is for the Sire of today’s runner.
The sire rating is a significant factor in today’s runner’s overall rating, but we wanted to break it out for you so you can get a deeper understanding solely of a sire’s relative strengths and weaknesses. In this tab, you’ll see the sire’s career record, his top TimeformUS Speed Figure, if available (on the Timeform Global Scale, which generally tops out in the 140s), and then the categorical ratings for ALL of his offspring, rather than just for today’s runner. Again, the ratings themselves are on a 100-pt scale. Sires are evaluated based on the top performances of their offspring (again, at today’s surface and distance), and they are also bonused if they have a history of siring horses who ran figures in the top 1% of our speed figure results over the last decade+. If the data on the sire is limited, such as when a sire is new to the breeding shed, or simply inexperienced, we rely on his father (the grandsire) as a proxy, but we proceed with caution here, slowly ramping up the influence of the sire as it becomes more apparent what the new sire’s true stature is. In these situations, we’ll put a white box around the pedigree rating on the surface to let you know the data is limited.
The third tab is for the Dam (mom) of today’s runner.
This gives you a quick glimpse at a dam’s career record, and lets you know the surfaces on which she typically ran. For younger dams, it also includes their speed figures, which we factor into our ratings for today’s runner more heavily when Mom has had few or no babies who have hit the track. The more she proves herself as a broodmare (successful mother of racehorses), the less we care if she was fast or slow when she raced. So if the dam is unproven, her race record and speed figures can be of particular interest. If she’s got a record with her babies, it’s best to rely more on the fourth tab.
The fourth tab is for the siblings of today’s runner–other racehorses who have the same dam.
While the data set is typically a small fraction of the data associated with the sire, we do indeed evaluate today’s runner via the lens of how his brothers and sisters have done in the various surface and distance categories. If his dam has produced a series of runners who have tried today’s surface and distance, we weigh these siblings’ results, and upgrade or downgrade today’s runner’s rating accordingly. Again, if no such data exists, we rely more on the dam’s results. If neither exists, we simply rely on the results of the sire and broodmare sire, but ONCE AGAIN employ a white box to let you know: the data’s limited here.
The fourth tab might be the most fun to go through…
It starts with a cumulative career record for all siblings (above), and tells you the single best TimeformUS Speed Figure earned by any sibling. Those cumulatives are then broken down by category, and finally, each sibling’s career record and top Speed Figures by category are displayed.
It’s a great way to learn about a family history for today’s runner, and by routinely looking at the siblings’ figures, you’ll start to see the influence of certain sires (all sires are listed on the siblings’ records) on the breed.
Things to do (phase 3, 4, etc):
1) More detail on the career records of older sires and dams at particular surfaces and distances.
2) Integration with Timeform global figures to pull in the records of offspring overseas, particularly in the case of siblings.
3) More work beyond just the closest relatives of today’s runner–there are some aunts and uncles who deserve to be accounted for as well!
4) More subtle categories, such as 8-9 furlongs vs 9+ furlongs. As ever, sample size is the issue but we’re unafraid to “bonus” a rating when there’s a strong trend based on a small sample size.
Questions or comments? Please let us know either in responding to this post or e-mailing us at firstname.lastname@example.org