Originally Posted by crash

I think what Stix is saying is useful, but could be refined a bit more maybe. From those horses that had finished between 1st and 4th at their 2nd last start [which ended up being between 50-60% according to Silver_and_sand's figures], if we check the race times of those 2nd last starts [and why not last starts too?] and ignore the horses that were in slow races, that 50-60% would probably narrow down to about 10%. Now that would be meaningful.