Thursday brought 132 races across the cards, and the model's first pick won 30 of them. That is a 22.7 percent strike rate, a touch below where we have been running this month, so we will call it what it was: a workmanlike day rather than a headline one.
The first pick also hit the places another 60 times, which tells you the selections were consistently competitive even when they were not winning. Dogs finishing second and third all day is mildly frustrating to watch, but it is exactly what you want to see under the bonnet. When picks stop placing altogether, that is when a model has a real problem. This was not that.
The broader picture was healthier. At least one of the top three rated dogs won 101 of the 132 races, a 76.5 percent top-three win rate, and the frame was covered 123 times. The model is reading these races well, it just kept finding the right dog in the wrong finishing position on Thursday.
Days like this are the reality of greyhound racing. Bumps at the first bend, a slow break, a wide runner getting the perfect trip, and a deserving favourite ends up third. Over a week these things wash out, and the seven-day numbers say as much. The full race-by-race breakdown from Thursday is on the results page, and Friday's card offers 130-odd chances to do better. Onwards.