Exactly ! I would also choose the recent race, IF I had to choose only one. But I don't. I can have both, (AND NP is in it's early phase of development. It will no doubt improve). The recency thing I discussed above. The NP calculation could be weighted towards the recent end, if the research were to be done to support it, ((and it, (the standard deviation) IS computed from ALL the last 10 races)).
There remains the fact that while it is very unlikely that a horse will "break out" from it's established behavior patterns, I still have to look at the recent races to see at what position in the range the hoss will likely run today. I need all the help I can get.