Just how important is a last-start performance ?
Top class galloper Beauty Generation (left) will head into the G2 Chairman's Trophy at Sha Tin on Sunday as one of two last-start winners and he will be getting plenty of respect from the public for that but how important is that form number that appears next to each runner?
The first place that most punters, even jockeys and trainers, look for the winner of a race is by looking at where they finished the last time they ran. The horse that has a 1,2 or 3 next to his name is considered "in-form" while the horses with bigger numbers in their last run are considered to be out of form.
That is the simplest form assessment shortcut possible, overlooking even the fact that some horses with a 7 or 8 from their last run might have had a 1, 2 or 3 there but for bad luck or a poor ride or that one horse might have been fit and another not.
Yet, as simple as that is, figures show over time that punters are right to give plenty of weight to the last-start finish. If we look at the table on the right, with figures from the last completed Hong Kong season in 2016-17, placegetters proved the horses to beat when they stepped out next.
Not only did last-start winners provide the highest percentage of winners, there is a pattern down the line as we look at different finishing positions. (For our purposes, horses which fell or for some other reason didn't complete their race are included under last place, depending on how many runners were in the field.)
Last-start winners performed better than last-start seconds, which in turn performed better than last-start thirds, and so on. With one or two exceptions, the trend fits beautifully and, as we will see later, the bumps get smoothed out the more data we use.
So, why would winners perform better?
In some jurisdictions, it would make total sense for the best one day to be the best the next day as well, but in Hong Kong, where races are 96 % handicaps, it seems less likely. In a pure handicap regime, every race should start with all runners having even chances, but we can see that the winner last time has a better chance than the horse which finished somewhere worse.
The reality of handicapping is that the winners are rarely penalised in keeping with their dominance - there is a generally-held rule of thumb that the horse who does enough to win is usually value for a little more than the margin - and the handicapper cannot account for the horse who wins well within his own capabilities. Racing is adversarial - you only have to win, you don't have to establish a true difference between horses and there is no benefit for a wider margin. So the progressive horse, maybe even a future Group One horse, going through the classes can win nicely by a length or two and not by far enough that he can be appropriately penalised to bring them back to the field next time.
And the handicapper is also over-cautious about lowering handicap ratings for beaten runners. Unfortunately, that means that by lowering the beaten horses only in small increments - and there are many of them - and by being unable to sufficiently raise the rating for winners - and there is usually only one a race - the whole equation leans towards the ones performing well, like the unbeaten Lean Perfection (above) who lines up on Sunday looking for 4 from 4, even though a pure handicap system is not intended that way.
In modern racing everywhere, too, there is an almost unconscious trend to not being hard on winners because it is perceived as good for the sport if winning horses continue to win. You'll hear trainers complain that such and such a horse is punished for his consistency but that is the essence of a handicap - the horse performing well is getting prizemoney at the expense of those not performing well, so some evening up should constantly be occurring.
As our tables show, it's insufficient. But, as we will see now, it may be a legitimate part of the puzzle in working out the winner but slavishly following the last-start performance figure is not a profitable betting strategy on its own. This table uses a lot more data, going back to the start of the 2012-13 season, and if we look at the win percentages in the fourth column, they are a smooth descending line without the little aberrations of the earlier table.
But the final column shows how you fared in a straight up $10 bet on each finishing position. You lost 16.4 % on last-start winners and your best result is losing 12.99 % on your outlay by following last-start fifths.
Maybe the slight surprise is how poorly you fared backing last-start seconds and thirds - the perception is that these horses are the ones ready to win next time and the win percentages support that as far as the result is concerned but clearly they are being overbet and not converting their placings into wins at the rate expected by the public.
So the answer to our headline question is that last-start performance is important, a building block to finding winners in Hong Kong, but making a profitable bet is going to need consideration of more factors than the simple form numbers.