Sometimes the differences in strength between horses is razor thin, that only a computer with the help of sophisticated software can successfully pinpoint who’s best. That was true today in race 3 at GP, a $16K claiming race for four-year olds on upward over 7 1/2f on the turf. The Handicap Wizard rated Nileator at 7-2 as its top selection, albeit barely. Two other horses, namely, Rompin Red at 7-2 and Tipsy Kitten at 6-5 had comparable winning probabilities around 25%. In such situations, it often pays (quite literally) to defer to the speed estimate instead of winning percentage (WP) to decide what horse to play, because speed is the most significant explaining variable. Nileator with a winning speed estimate (SPD) of 37.6% looked then to be the best choice. This was how the race went, with Nileator taking the lead out of the gate and never looking back to win wire-to-wire. Nileator was also a fair bet, having a positive expected ROI of 11%.
According to the live odds, race 5 looked to be a contest between Teak and Big Possible with odds of 3-2 and 2-1, respectively. However, this $16K race for three-year olds and up was very short at only 5f on the turf. The data in the Handicap Wizard revealed that Teak was indeed the speed horse in the race but only as far as E1 with an estimate of 54.4%. He had nothing to show for E2 and only the 5th strongest LP rating. Another horse, Salambro at 11-1, had the 2nd highest E1 estimate behind Teak and the highest E2 estimate at 65.4%. No one else came close. Given the short distance, it could be deduced from that Salambro had precisely the right speed where it was needed to win at 5f. The race played out with Teak going to the early lead, while Salambro stalked from about 3 1/2 lengths back. Heading into the final stretch, Salambro turned on his speed in a five wide bid, overtaking three other horses inside him, to win by under a length. Salambro paid $24.20 for the win. The horse had a positive expected ROI of 94%, while the two favorites had negative expected ROIs.
Pro-Handicap Analytics 