Ratings Update: October 2026

Horstradamus says: "I just got the biggest upgrade of my career."

Horstradamus®, our rating engine, was upgraded on October 9, 2026. Here is what changed and what you might notice on your E-Z Win® Forms.

At a Glance

• Sharper on every surface. In six years of backtesting, the eventual winner on top more than 10% more often, and over 20% more Big 'Uns.
• Tested on live races. Side by side with the previous engine this summer, the winner on top about 10% more often.
• Fairer comparisons across tracks. Horses shipping in from stronger circuits are rated more fairly.
• A new first-time starter model. Built on workouts, auction price, and the trainer's record with first-timers.
• Turf judged on performance. Class is now built into the turf rating, with no separate number in parentheses.
• One rating per surface. Ratings no longer change with the track condition, so the condition dropdown is gone.
• Nothing to do. It is already live on every E-Z Win® Form.

The questions below go into each change in more detail.

Why do my picks look different?

We upgraded how we rate races on every surface. On dirt and synthetic, our speed figures now do a better job of comparing races run at different tracks, first-time starters are rated with a new model, and ratings no longer change with the track condition. On turf, horses are now judged on performance the same way as on dirt, with class built in, so turf is where picks and tiers look most different. This is the new, improved ratings at work.

Are the new ratings better?

Yes. In backtesting across six years of historical races, the upgraded engine put the eventual winner on top more than 10% more often than the previous engine, singled out a standalone top pick in more than 10% more races (and was nearly 30% more accurate when it did), and flagged over 20% more Big 'Uns, including nearly 30% more $1,000+ scores. It then ran side by side with the previous engine on live races this summer, and in that live testing it put the eventual winner on top about 10% more often. As always, past testing does not guarantee results on any single race.

Where did the class rating go on turf?

Turf ratings used to show a separate class rating in parentheses, like (12.7), in front of the figures. A horse's class is now built directly into its turf rating, the same way it always has been on dirt, so turf ratings now read like dirt ratings. Nothing was lost: class still counts, it is just part of the main rating now instead of a figure on its own. This applies to standard thoroughbred turf races. The change is forward only, so older results still show turf ratings the original way, with the class number in parentheses. Only races from October 9, 2026 onward use the new format.

Where did the track condition dropdown go on the E-Z Win Form?

Ratings no longer change with the track condition, so the selector had nothing left to do and we removed it. Each horse now has one rating per surface, whatever the condition of the track, and it holds up: in backtesting across six years of historical races, the new engine put the eventual winner on top more than 10% more often on fast dirt tracks and more than 15% more often on off-track dirt than our previous engine. The form still shows the current track condition, and on turf races you can still switch between dirt and turf.

Why are there four horses in the top tier now?

When the top of a race is a genuine near tie, our new tiering can seat a fourth horse in the top tier instead of forcing a cutoff at three. It only happens when the numbers are that close, and it is intentional.

Why do first-time starters rate differently now?

First-time starters have no races to rate, so we estimate them. The new engine uses a more detailed model built on debut signals such as workouts, auction price, and the trainer's record with first-time starters. You may see fewer first-timers in the top tier than before. In backtesting across six years of maiden races, where most first-timers run, the new engine put the eventual winner on top more than 15% more often.

Why did a horse shipping in from another track change ratings?

We improved how we compare races run at different tracks, so horses coming from stronger circuits are rated more fairly against local horses.

A turf horse shows "Not Rated." Why?

Some horses do not have enough relevant history to rate for a given race. Separately, we fixed a case where certain first-time-turf horses that did have a history were being left unrated, so you should see fewer of these on turf now.

Why did the track variant change on an old past performance?

Along with the ratings upgrade, we improved how we calculate track variants. Past performance lines now show the improved variant, the same one the new ratings use, including for races run before the upgrade. The race itself has not changed, only how we measure the speed of the track that day.

Do I need to do anything?

No. The upgrade is automatic and already live.