Two weeks ago, I won a nice bet when Pleasant Colony captured the Kentucky Derby. On Saturday, I lost a bet when Pleasant Colony rallied to beat my choice, Bold Ego, in the Preakness.

This, of course, is the way horseplayers spend their lives, going through an endless cycle of victories and losses. Yet during that cycle, there will be certain experiences that stamp themselves indelibly on our minds, that mold our view of the game and shape our future actions.

For me, the first two legs of this year's Triple Crown have constituted such an experience. To other racing fans, this may seem a rather routine 3-year-old campaign, but for me it has been an intensive education, a revelation. I will probably never look at races in quite the same way again.

I am a devout believer in speed handicapping; I judge horses' abilities according to how fast they have run in the past. Having calculated a figure that expresses a horse's ability, I would generally expect him to run that figure, barring bad racing luck or a chance in his physical condition. If a horse were capable of traveling 1 1/4 miles in 2:03, I would expect him to do it no matter how the race were run.

In particular, I rejected the old maxim that "pace makes the race," the notion that a horse might run in 2:02 and 2:08, depending on how the early stages of the race were run. In my book, "Picking Winners," I decreed: "Horses are never burned up by fast fractions. There is no such thing as a 'killing pace.'"

Of course there was no way to prove this thesis, since thoroughbred racehorses can't be used for scientific experiment quite so easily as laboratory rats. But if I were to have conceived an experiment, I might have designed it like this:

In the Kentucky Derby, let's have the early leaders run the fastest first quarter mile in 107 years. In the Preakness, we'll take roughly the same group of horses and have them run the slowest initial quarter in 17 years. Then we'll compare the results of the two races and judge the impact of pace.

This experiment was conducted at laboratories in Louisville and Baltimore in the last two weeks, and after witnessing it I would like to expunge a few lines from my book.

At Churchill Downs, the horses who were running 1-2-3 after a half-mile finished 10-18-19. All the speed horses collapsed; stretch runners were the first nine finishers.

At Pimlico, the horses who were running 1-2-3 after a half mile finished 2-3-4. Two longshots, Paristo and Thirty Eight Paces, were among the front-runners, and they managed to beat many rivals who had far superior credentials. The clearest illustration of the difference between the two races was Bold Ego, who collapsed in the Derby and lost by 11 lengths, but held on well and lost by only a length in the Preakness.If Pleasant Colony hadn't been in the two races, it could properly be said that pace had determined the outcome of both.

But, of course, Pleasant Colony was in both races and won them both; he was a superior enough racehorse to beat this group of 3-year-olds under any conditions. The Preakness result made this perfectly clear, since the colt encountered plenty of difficulties -- running disadvantageously wide on both of Pimlico's tight turns -- but still won with authority.

A year or two ago, I might have picked Pleasant Colony in the Preakness -- he was, after all, the top "figure" horse -- and have felt that all was right with the world when he won. But the handicapping issues involved in the Derby and the Preakness were far more complex than that.

In the Derby, Pleasant Colony had had a perfect trip -- coming from behind after the fast pace, saving ground on the turns -- while Bold Ego had a disastrous trip, engaging in that suicidal battle for the early lead. Before the Preakness, I thought (correctly) that their positions would be reversed. Bold Ego would have a perfect trip, getting the lead without much pressure, while Pleasant Colony would have a very tough trip, losing ground at both turns as he attempted to rally. The question was this: would all this enable Bold Ego to turn the tables on Pleasant Colony after losing by 11 lengths in Kentucky? How many lengths would the difference in trips mean?

The answer was 10. I would have been happier if it had been 11, but Bold Ego's performance incontrovertibly demonstrated how important are pace and the other factors that constitute a horse's "trip."

Understanding this does not make handicapping any easier; it makes it tougher. I know I must learn to correlate figures and trips somehow, to judge how much the trip may affect a horse's performance. I'd like to reach the point where I could have said on Saturday: Bold Ego is going to get a perfect trip and improve dramatically, but he still won't be quite good enough to beat Pleasant Colony. Learning to make such judgments won't be easy, but it is the glorious complexity of this game that keeps horseplayers hooked on it for life.