Over the last eight days, the prediction models have processed 856 races. The headline first-pick win rate sits at 24.6%, with a place rate of 43.7%. Those numbers are broadly in line with the long-term averages, which is reassuring given the volume.
The day-by-day picture is more revealing. Friday the 11th was the standout, with 38 winners from 142 races (26.8%). Sunday the 13th was close behind at 35 from 135 (25.9%). Tuesday the 15th was the toughest, producing just 29 from 143 (20.3%), a day where the second pick actually outperformed the first on raw winner count.
Monday the 14th deserves a closer look. Only 24 first-pick winners from 112 races (21.4%) and a place rate of 42.9%, both below average. But the top-three cluster still found 83 winners from those 112 races (74.1%), which tells you the models were identifying the right contenders but ranking them in the wrong order within the shortlist.
The broader top-three numbers are the real strength. 550 winners from 856 races gives a 64.3% win rate across the three selections, and the place rate of 88.4% means nearly nine out of ten races have at least one of the top three picks in the first two home. For systems that use rank-based filters, especially those that look at whether any of the top three selections has a standout composite or speed profile, that 88.4% place rate is the foundation everything else builds on.
Today's early results (five races so far) show four top-three winners already, so the week has started on the right foot.