Wednesday's card covered 131 races across the day, and the prediction model returned a solid if unspectacular set of numbers. The first pick won 23 times from 131 races, a win rate of 17.6%. That is a shade below the long-run average, but comfortably within normal range for a single day's racing. Where the model earned its keep was in the place market: the first pick placed 51 times, a place rate of 38.9%, which keeps things ticking over.
The broader picture was stronger. Taking the top three predicted runners as a group, at least one of them won in 71 of 131 races, a hit rate of 54.2%. At least one placed in 104 races, which is 79.4%. Those are healthy numbers on a day with no obvious standout performances.
It was not a day of dramatic winners or spectacular collapses. The model did what it is designed to do: identify the most likely contenders in each race and get them into the frame more often than not. Some days will see the first pick win rate push into the mid-twenties; Wednesday was not one of those days, but the consistency of the place figures suggests the underlying ratings were sound.
The second and third picks contributed meaningfully too, combining for 48 wins between them. That distribution, where the second and third picks win almost as often as the first, is typical of days with competitive fields and tight margins. The data was reading the form correctly, even if the top selection was not always the one crossing the line first.