Sunday delivered a solid set of returns. From 156 races, the top-rated pick won 40 times for a first-pick win rate of 25.6%, right in line with the model's long-term average. The first pick placed in 75 of those races (48.1%), which is a comfortable number for anyone following place strategies.
The real strength showed in the broader top-three selections. At least one of the top three predicted dogs won in 98 of 156 races, a 62.8% strike rate. And for places, the coverage was excellent: 137 of 156 races (87.8%) saw at least one of the top three finish in the frame. That's the kind of consistency that makes combination bets viable.
The second pick contributed 31 wins and 66 places, while the third pick added another 27 wins and 52 places. What's encouraging is the spread across all three picks rather than the top pick doing all the heavy lifting. When the model distributes its accuracy like this, it suggests the underlying ratings are separating the field well rather than just getting lucky with one obvious favourite.
Sunday's card featured 156 races across multiple venues, which is a big enough sample to read the numbers with some confidence. A 25.6% top-pick win rate on a sample this size is a genuine indicator of form rather than noise.
You can dig into every race from yesterday on the historic results page, where each prediction is matched against the actual outcome. Good days like this are worth reviewing to see where the model's strengths lie and where the value sits.