In this article, freelance writer Ynez Ybarbo completes her special two-part report, with some very surprising conclusions.

Of course, the better horses usually carry the higher weights, and one would expect the better horses to clock the faster times, but nevertheless the higher weights carried by these animals were supposed to be a penalty, and we had expected the higher weights to even things out.

In the end all three of us agreed that, in broad terms, weight was not a major factor in stopping horses from running fast races. Of course, we needed to investigate further, as other questions now came to the surface.
For example, did weight swings make a difference? And, did horses have a maximum weight that they could carry effectively? If so, was that maximum weight different for each horse, or was it about the same for all thoroughbreds?

In order to examine the data to see if weight swings made a difference, we first isolated races where the winner and horses that were no more than six lengths back met again at a later date. There were many thousands of such races, and in fact there were many races where foes from the first meeting met three, four or even five times.

Our program now recorded the result of the first meeting, the time for the race, the weight carried by the horses concerned, and the number of lengths involved. We also noted the running style of the horses - classifying them as frontrunners, pressers (a half to one length back for about the first two thirds of a race, without attempting to take the lead), stalkers (one-and-a-half to three lengths back) or closers (at least four lengths back).

We also noted those races where one horse was on an uncontested lead. The program searched for subsequent meetings between those animals and recorded the weight differences. If there was a weight swing of one or more kilograms between any of the runners, the race was flagged. If there was no significant change, or swing, in the weights between the horses being studied, the race was discarded.

We had to get to swings of 6.5kg or more before we began finding significant reversals of form. In other words, if horse A beat horse B by two lengths over 1600m, and horse A was penalised by 2kg next time out, horse A still tended to beat horse B.

Sometimes the margin of victory did decrease, but just as often it increased. If the penalty was 4kg, horse A still tended to win the repeat race (or, at least, beat horse B home). It was only where, in a subsequent race, horse A was penalised by 6.5kg more than horse B that horse A began consistently finishing behind horse B.

And, as can be imagined, the number of races with swings of such magnitude are small indeed (but they do occur, as in cases where horses move from handicaps to weight-for-age events).

There was another important discovery here. On many occasions, it took three or four meetings between the same horses, with the weight swing becoming gradually larger, before the cumulative weight penalty took its toll.

Therefore, it was a mistake to look at just the most recent two races, find a swing of a kilogram or so, and conclude it was that kilogram that caused the turnaround in form. It was the cumulative penalty that made the difference.

The same results were found over all common distances. The running style of a horse did not seem to matter much either, because the results were the same for frontrunners, pressers, stalkers and closers.

Where we did see significant differences was when stalkers or closers that had run behind lone frontrunners in the first race found a contentious pace the second time (or vice-versa).

Bearing in mind the other results, however, we figured that in these cases we were comparing apples and oranges.

In other words, when the pace was true, the stalkers and closers were better able to run to the best of their ability.

We now wanted to discover whether the scale of weights that was used in racing was sufficiently severe, at the top end, to effectively kill the chances of any runner from competing. We only considered Major Stakes and Graded races for this particular study, our reasoning being that horses that were unable to compete at that level may skew the results.

If they were not able to rise above the Allowance ranks, they were not top-quality animals, and an apparent inability to carry weight may well have been caused by any number of other reasons. The result was overwhelming.

Horses that regularly competed at the upper levels of racing did not appear to be affected by the top weights, which are seldom much over 58kg in the US. As with the time study, in fact, the horses carrying the heavier weights still won far more than their fair share of races.

Did that indicate, though, that there was a weight threshold over which the other competitors could not cross? Not really. After carefully studying the result charts of many of those races, we came to the inescapable conclusion that the reason those horses were not able to win was not an inability to carry weight, but simply that those animals did not have the racing ability of the consistent winners.

In other words, they were outclassed, even though their overall records suggested that they deserved to be competing at the best levels. The situation is rather similar to that of athletes like Linford Christie, Calvin Smith and Sam Graddy, who were all top-class human sprinters in the 1980s, and certainly deserved to be competing at the highest levels.

All the sprinters of the era knew, however, that when they were on the track with Carl Lewis they were outclassed. That did not detract in any way from their own ability; they simply had little chance of beating Lewis under any circumstance.

This conclusion tended to be confirmed by the fact that a good number of mediocre performers are converted to steeplechase and hurdle racing, where the weights that are carried are much heavier than for flat racing. Many of these convertees manage to win over the jumps, carrying perhaps 12kg more than they were able to win with on the flat tracks.

A thoroughbred racehorse that is in training probably weighs in the neighbourhood of 500kg. The average impost carried in a race is 55kg, which is 11 per cent of the horse's weight. If that horse is penalised by a further 5kg after turning in a top performance, that extra weight diffuses into insignificance compared to the other weight figures.

Is it really likely to make much of a difference? Let's return to the Carl Lewis analogy for a moment. Lewis weighed in at about 80kg in his heyday, and if human athletes were handicapped in the same manner as horses, his penalty would have been 9kg. If, after winning an Olympic Gold Medal in 1984, he was penalised by another kilogram, would that have prevented him from winning again? (Remember, if Lewis was carrying weight, so would all the other athletes at his level.)

We do examine weights when handicapping a race now, but only to ensure that no one runner has received a big swing for any reason. Thanks to the computer, it is a relatively simple matter to keep track of punitive penalties, and once an animal reaches 6.5kg in excess of what it was carrying the first time he beat similar horses, we start taking notice.

Otherwise, for the race-to-race changes of two or three kilograms, we ignore weight as a factor.

Moving on, our database has allowed us to study any number of other factors. Post-position can be very important on the dirt, especially if the track has any degree of moisture present.

We found that on the turf, post-position was the least of our worries when handicapping.

Mile races, for example, start very close to the first turn on most grass tracks here, and that would seem to indicate that horses drawn on the outside of the field are at a disadvantage.

The statistics say otherwise: horses that start in the ninth gate and higher do not win at a rate that is significantly different to horses drawn closer to the rail.

The reason is, I suppose, that horses with speed are able to clear most of the field on their inside, short distance to the turn or not. Thus, by the time the turn is reached, they are able to move over towards the rail and not lose too much ground.

Conversely, closers simply let the fast horses get away, then usually have ample room to move around and position themselves. The only time I become overly concerned with post position now is if my fancied horse is drawn way wide, is a speedy go-to-the-front type, and there are several others with speed that have also been drawn in the outside positions.

Then the traffic can get hectic, and unless my horse has won in similar situations before I would probably pass the race.

We have found a wealth of information concerning trainer habits in the database, and in fact the intention of the trainer is now our prime focus when handicapping any given race.

Because of the fact that horses run on both dirt and turf in the US, there are any number of experiments being conducted by trainers all the time - and that opens the door to some lucrative betting opportunities. Some trainers are very adept at switching horses from dirt to turf, but of course those statistics are readily available these days, and runners from such stables are always overbet.

Harder to find, however, are statistics that relate to a trainer and just one owner, or circumstances such as switching from a dirt sprint to a turf route.

We can track these manoeuvres with ease, and have found some amazing betting opportunities as a result. For example, a trainer who wins at a rate of 13 per cent with his entire stable, but who wins at a rate of 46 per cent the third time his charges go on the turf, at an average starting price of 11/1.

Or, the trainer who gives his horses their turf debut on a minor track, hardly ever wins, then ships them to California where they win their second start at average odds of 6/1.

I realise that the US is practically unique in the dirt to turf department, but the point is that by having a comprehensive database a punter can find all kinds of betting angles to work.

Some trainers are very good with two-year-olds, others better with three-year-olds, but almost all have other hidden data which can be very beneficial when betting.

Simply tracking trainers that regularly ship horses from track to track can be most rewarding. The mere act of sending a horse from a better track to a lower-class one does not necessarily mean instant riches for bettors, because it is quite usual for such runners to be heavily bet. But what about the trainer who races his young horse at a minor track a couple of times, shows nothing, lays the animal off for six months or more, and then enters him in a field at a major track against what appear to be much better animals? We know of several trainers who are very good at doing this.

On average, they only win 10 per cent of the races they attack in such a manner, but the median starting price is an astounding 44/1.

The betting public as a whole never picks up on the situation, because all they see in the form is two miserable races at a small track, and now the horse is in against the big boys. Many of the more astute punters will realise that the trainer is up to something, and look no further than the fact that the move seldom works. You can be sure that there are many such situations and angles available to you.

Your work is to root around until you find them, and then capitalise on them when they are uncovered. One fact is inescapable: if you bet the same way that everyone else does, you will, sooner or later, lose your bankroll. Find the right spot and, after testing it, bet it.

And don't make the beginner's other mistake that I made of betting $500 on the sure thing at 11 / 10 and then betting $20 on the live longshot that I uncovered. Invariably, the favourite found a way to lose, and the longshot won. Even if the favourite wins, the next bet may well lose, and you are back to square one.

If your longshot is 10/1, it means you have ten opportunities to reinvest the profit and, even if you lost all ten bets, you would still be no worse off than the one loss on the favourite. I did not say you should bet indiscriminately on longshots, however - do some research, make sure that you are not misreading the situation and data, and find a convincing reason first.

More thinking, less betting.

Ynez Ybarbo, a native of Las Palmas, has a BA degree in Communication Arts and Sciences from the University of Barcelona. She was a staff reporter for the English-language Costa Blanca News for three years, and was then, a business reporter for Diario Informaci6n (Daily Information) in Alicante for about four years. Since moving to the US, Ynez has been a freelance writer, with assignments at various large companies such as Ford, Boeing and Du

Ynez runs a daily jackpot based on four American races with an entry fee of 25 cents per line. She also acts as a betting agency if you wanted to bet on US racing. Her website is:

Click here to read Part 1.

By Ynez Ybarbo