07/24/2001 11:00PM

Identifying dead-rail days no simple task


COLUMBIA, Md. - Saratoga is one of those special cosmopolitan race meetings. Horses come from everywhere to run in those grand, historic surroundings. Trainers and owners point for it, craving the particular pleasure and prestige of winning in an atmosphere of reflected ancient glory.

But Saratoga also has its more humdrum side. It's not all glory and atmosphere. After all, Saratoga is also part of the never-ending New York race year - just six more weeks on a 52-week marathon.

An understanding of that part of Saratoga could be the key to success in 2001. You can order all the specialty players' guides to past Saratoga meetings, which surely have their uses. But they will not tell you what you most need to know: what to do with all those bad rail days-or what appeared to be bad rail days-at Belmont Park.

The Belmont rail can play with your mind. Some days it clearly looked very deep. Other days it looked only slightly off, and if a horse raced in the two or three path he was probably okay. Some days the deeper inside damaged speed horses, but most days speed horses were not affected as long as they stayed a path or two off the rail - which most of them did. All this made handicapping more complicated. If you liked a speed horse from the inside, how could you be sure that the jockey would stay off the rail? Or that a closer breaking from an inside post would move outside, rather than try to take the shortest but deepest way home. It all made for an unusual amount of second-guessing.

That indecision also affected my judgments about just how many truly biased days there actually were at Belmont. Often the rail appeared to be bad, but a careful review of the races showed enough horses running well when down toward the inside (if not precisely on the rail), and speed horses doing just fine. Were those days biased enough to matter?

So, in preparing for Saratoga I wanted to get a firmer grip on the Belmont biases. But there were many problems:

* With its two turf course Belmont runs a lot of grass races - often as many as four per day. That leaves a limited sample of only five dirt races on which to make your judgment about biases.

* The Belmont dirt races typically have small fields, further limiting the number of horses who might race on the questionable rail. And on many days one of those dirt races will be a five-furlong 2-year-old race, where inside speed often wins despite any bias. And there might be cheap claiming races and weak New York-breds - not the best evidence on which to make firm conclusions.

* When the reports about bad rails made it to the jockeys' room, the riders started avoiding it - so almost nobody actually raced on the rail. That reduced the sample even further, making it impossible to make any reliable judgment about biases, and leaving the observer wondering whether, by consistently migrating toward the middle of the track, the riders were actually creating an visual illusion of a dead rail.

* This doesn't even take into account the amount of time a horse spends on the inside - runners sitting on the rail for most of the turn, for example, and then angling out wider in the stretch. In what category do you put that horse?

Despite all the problems, I tried to put together a list of horses who raced almost entirely on the rail on days when I thought the rail was most clearly negative: May 12, May 13, May 20, June 3, June 10, June 16, and June 30, just to mention a few. Then I recorded how much their Beyer Figures improved or declined in their next starts. The result? More confusion.

After all, rail runners return in all kinds of situations - in the slop, on the turf, after layoffs, at different distances, at different points in their form cycles -and that makes simple mathematical calculations very dicey.

But I still thought there might be some kind of pattern that would confirm some of my more doubtful judgments about bad rails.

There wasn't. On most of the questionable-rail days I analyzed, rail runners showed no pattern of higher Beyers next time out. Most often there was about an equal split between rail-runners who ran higher Beyers in their next starts and those who ran lower. (May 13 emerged as the deadest of dead rails, along with June 16 and June 22 - at least according to this simplistic method.) For those horses who ran the biggest Beyers as inside speeds, their strong tendency was to run worse next time out. Perhaps these bouncers, in battling hard against the bias, actually provided some counter-intuitive confirmation of the deep rails. But who knows? That might be too clever.

After all these efforts to determine the gradations of Belmont bad rails, I would like to be able to offer a full list of dead rail days. But I wouldn't dare. Too many days were just too ambiguous, too confusing - and the evidence was not fully convincing.

But here are some safe, if not overwhelmingly helpful, pre-Saratoga conclusions about the Belmont meeting:

* There were no big speed-rail days during the entire Belmont meeting.

* There wasn't a single day when the rail was an actual advantage.

* The rail in July seemed less of a problem, but still there were few if any horses wiring fields right on the rail.

* You should watch carefully to see if any pattern of improvement emerges for rail-runners from any particular Belmont day.

You have to be cautious in betting back horses coming off rail trips at Belmont. There's nothing more disastrous in handicapping than a mistaken or exaggerated bias diagnosis.