What Equibase Profiles Actually Show You
Equibase profiles are the public-facing race cards that come with every thoroughbred in North America. They include past performances, speed figures, class changes, pace data, and trainer/jockey stats. Most people look at them the wrong way. They scan the last five races and try to spot a pattern. That rarely works because the data is noisy. The real utility comes from understanding what each section means and where the numbers lie. Speed figures on Equibase are Beyer-based for most tracks, but they adjust differently depending on the track and condition. A 90 at Saratoga is not the same as a 90 at Belmont. Same figure, different reality.
The Power Of Equibase Profiles Unlocking The Secrets Of Horse Racing
Here is what actually unlocks value from these profiles. You learn to read pace scenarios before you read speed figures. A horse with a solid figure but no recent lead-time advantage is often a trap. Equibase shows fractional times in the past performance line. If the second fraction of a race is slower than the first by more than a half-second, that race likely had a hot early pace. The winner may have gotten lucky. The second-place horse behind that split might be the real play because it saved ground and energy. I spent years ignoring fractional splits because they looked cluttered. Then I started pulling the raw data and building a simple spreadsheet. Within a month I was spotting pace traps that the morning lines completely missed. One race at Churchill Downs in 2023. A favorite came in at 2-1 with a solid 87 figure from its last start. The split showed a :22.4 first quarter and a :26.1 half. That is a race that likely won going away early. The second horse in that race checked in at 3-1 but ran a much steadier split and finished second by a length. I backed the third horse in the field that day, one that had shown late speed in prior races with similar pace shapes. It paid 12-1. That is the kind of edge these profiles give you if you know where to look. Another thing beginners miss: Class changes are far more important than form lines suggest. Equibase flags class drops and rises with notes, but the actual drop from allowance optional claimer to claimer is where the biggest mispricings happen. A horse dropping from $50,000 claiming to $25,000 often runs significantly better than its last figure suggests because it was holding back at the higher level. The profile will show the previous race speed figure, but it will not tell you the horse was visibly saving effort. You have to infer that from the class line and the jockey change notes.
How To Actually Use The Data
Download the past performance files directly from Equibase. They are available for free in CSV format on their website. The interface is dated, but the data is clean. Filter for the track, date range, and min race distance you care about. Then build a working sheet with columns for speed figure, class level, pace split, track condition, and recent surface change. Do not rely on the default Equibase viewer for serious work. It is designed for casual bettors, not analysis. Export the data and run your own filters. A basic pivot table in Excel or Google Sheets will let you cross-reference speed figures against class changes and pace shapes in under five minutes. The manual review process takes two to three hours per card. The export method takes about twenty minutes. One practical tip: save your filtered sheets by track and date. Over a few weeks you will start seeing which horses return well off layoffs at specific tracks. Equibase marks days since last start in the header of each past performance line. Horses returning after 45 to 75 days show surprisingly consistent patterns at certain venues. I found this at Turfway Park and Keeneland. Horses coming off a forty-day break ran roughly eight to twelve points higher than their prior figure predicted, on average. That is not a guarantee. It is a lean you can use when combining other signals.
Get the Full Details
Where The System Breaks Down
Equibase profiles do not tell you everything. They do not show workout times for most horses unless the trainer submits them voluntarily. They do not indicate equipment changes after the card is released. They do not reflect track bias shifts that happen during the day. A profile printed on Tuesday is not the same as the profile at post time on Saturday if the track dried out or got soft from rain. Also, the speed figure methodology has flaws. Equibase uses Beyer numbers for many tracks but switches to custom regional figures for others. The adjustments are not always transparent. A horse rated at 85 at a track using the custom method may actually be performing worse than a horse rated at 80 at a track using standard Beyer. You have to check the methodology footnotes on each race card, and most people skip that step. My workaround for the figure inconsistency is simple. I track horses across multiple tracks and build a personal reference chart. When I see a 85 figure at one venue and a 80 at another, I compare the actual margin of victory and the pace shape to decide which number is more trustworthy. It takes extra time but it removes a lot of the noise.
For workout data, I pull from daily bloodstock software subscriptions or rely on independent tracking services like Brisnet or TwinSpires. Equibase sometimes shows workout fractions for stakes runners, but covering the entire card manually is not efficient. If you are serious about this, you need at least one secondary data source for late scratch and workout updates.
Advanced Nuances That Matter
Jockey changes within a short timeframe are heavily overvalued by casual bettors. Equibase flags a new jockey with an asterisk, and that alone drives price movement. But a jockey change is often just a trainer moving a horse to a different rider for scheduling reasons. The skill gap between available jockeys at most claimer-level races is smaller than the market assumes. I learned this the hard way when I laid off a horse because the regular rider pulled out and a replacement got booked. The replacement won three races that weekend. The lesson is to evaluate the replacement rider's actual win rate at that specific track, not just the name recognition. Another nuance: surface switches. Equibase profiles list surface history clearly, but the conversion between dirt and turf is not straightforward. A horse with a decent dirt figure switching to turf does not automatically struggle. Some horses perform better on grass because of stride mechanics that Equibase does not capture. The profile will not tell you this. You need to watch the horse race or rely on trainer turf records. Trainers who consistently win on turf with switchers are rare, but they exist. Finding them takes patience. Finally, understand that Equibase data is retrospective. It tells you what happened, not what will happen. The biggest mistake I see is people treating past performance lines as predictive rather than descriptive. They are descriptive. Prediction requires combining them with current conditions, scratch information, and pace analysis. Done correctly, Equibase profiles cut your analysis time by roughly seventy percent compared to reading full race programs. Done poorly, they give you a false sense of precision.
