There is an old, old story about an owner who maintained that the way he won so consistently was to keep himself in the best company and his horses in the worst. He moved in the very best circles and, as far as he was able, he would hold his horses in their lowest possible classes. An American, he occasionally faced the possibility of losing his horses to a claim.

In a claiming race, any horse may be purchased at the set Price before the race. After the horse has run, it is safe, although a rival trainer or owner might still make an offer, which can be accepted or rejected. It is a risk entering your horse in such an event, and one that can backfire if you have a promising young horse, or one about to strike the peak of its regular form cycle.

I once saw a pacer called Len Way claimed for $1500 in Toronto, Canada. This chap immediately went on to win, from memory, seven or eight races from his next ten, graduating to open class "1.58 and better" level, against the cream of North America's crop. It happens.

But claiming also offers riches in the form of excellent prizemoney, and often also creates good odds about a horse. Some trainers try to beat the claiming chance by running their horse in a high-class claiming event, where the high cost of claiming the horse would prohibit a purchase, and then slotting it back into its correct claiming level, and winning, without losing the horse. It's a very tricky business, but running your horse in the low classes can be fraught with dangers (not the least of which, in Australia, are huge weights and lousy prices).

The American fellow who made the assertion I started out with (there is some debate as to which celebrity actually made the statement) made no secret of his life's ambition to stay with the best, while his horses remained in the worst possible classes. It could pay off, and for us as investors it is worth investigating any horse that seems to be racing below its logical class level.

This is harder to get away with m Australia, as we have no claiming events. Trainers have to enter 'out of town' horses according to their proven ability. For example, if a horse has won five races on professional tracks, it cannot compete in a Class 3 race. If it has won, say, seven races, it is ineligible for any Restricted events out of town. If it wins three city races, it is usually heading towards Open class, with no protective coating on the conditions of its

Everyone has such excellent information these days. Gone are the times when you could have a feeling for a horse at, say, 10/1 and expect it to start at those odds. Now, with the incredible coverage of SKY allowing punters such access to up-to-the-minute information, if that horse is backed on its home TAB, it will be backed on all TABS.

So, the question I want to put is this: do we focus on the BEST horses, or on the WORST horses, or on a blend of ALL horses (or maybe stick to the middle, where they win and lose in turn)?

There is an easy way out of this: take the blend. That way you can just stop reading now, and go back into your cocoon. The blend will mean that you can bet on any horse, any time, anywhere. Just so long as you believe it has a good chance (meaning a better chance than its odds are likely to suggest) of winning.

And it's not as though this is a stupid, or even a head-in-the-sand way of dealing with the problem. Because you could well argue that there is no problem. Why shouldn't any sensible punter just shop around all races for value and bet accordingly? I'll tell you why, so far as I am concerned.

There are too many races. 

There are too many races for the average punter to be able to take them all into account. He has to specialise, he  as to make annual, but not lifetime, decisions as to what kinds of races he is best at unravelling. We learn by experience.

I seem to be best at the top end, so far as my win bets are concerned. My best single and doubles results are more often than not obtained by selecting the best horses in the best races. I take so many factors into account, factors that I am hard-pressed to evaluate in lesser events.

For example, I am very strong on certain riders, and I am impressed by trainers who place their horses to win good races on a regular basis. And I am equally, or maybe more, keen on horses that have 'done it' before. Regular readers know how much emphasis I place on repeaters; horses that have won over the particular route. Proven winners. Horses with the form on the board.

This means that I will gather around me a small group of class animals that I will tend to follow, supporting them when I believe they can repeat a former victory, or even a near-victory. The most important thing here, I believe, is your memory.

And in that regard you have two options. I am lucky that, so far as my racing is concerned, I tend to be able to recall how certain horses ran, and why they won, or didn't. But it doesn't matter any more, if that isn't your strong suit, as we all have access to something that has a memory which makes ours seem almost redundant.

I am talking, of course, about computers. A good one costs less than $2500 these days. OK, not a state-of-the-art professional one, but one that will do anything you ask of it. With two years' Net connections included, you can pay as little as $1700. Add a printer for $200, and you're away.

Or you can get a good laptop for under $3000. They are usually one stage below state-of-the-art equipment (for example; the majority of the cheaper breed are selling today with Pentium or Pentium II-level gear). I recently purchased a new IBM, with an active screen (the essential feature), more memory than I could ever need, and all the gizmos, including a 56k card modem, and it cost me $2700, because it was withdrawn, discontinued or superseded (take your pick of the terminology).

How old does "withdrawn" or "discontinued" make it? This will surprise some of you. When I checked the IBM info sheet, I discovered my fellow was released in the USA in very late 1998, was built in January 1999, and was then withdrawn in February 1999, as a newer and more powerful model appeared.

"Superseded" may not mean anything, as it often doesn't with cars. My machine makes rainbows. It flies along. Unless I were a professional programmer, I cannot imagine anything I would miss on this little darling. Plenty of zip, plenty of spare space, and easy to tote around. Frankly, it's a bit heavy for a 'lightweight', and I'd like more battery life, but it frees me up to get around while taking the office with me.

And that memory! If you think yours is good, wait till you find out what a decent data base can do for it. Say you want to examine your horse's performances at Flemington. Sure, the formguides tell you the wins and places and the percentages, but you may want to make various comparisons. The little wonder has it all at your fingertips (literally).

And that brings me back to the 'best' horses. I seem to be able to make comparative form lines between them all, and make them mean something for me. You might not find this suits you, but it's best for me. The only way is to consciously try each level, or examine your betting over the past, say, three years. It doesn't mean that I restrict myself to them, but it is possible that I should.

A classic example is that when I started to really concentrate on this form of selection for my Pinnacle followers, we started getting really great results again. The latest I can recall is the special bet on Sheer Kingston in the Brisbane Cup, where he streeted the field, repeating what he had done in Adelaide the month previous. As to the Stradbroke, I thought it too hard, but I advised those who wanted to bet that it was out of two horses, one of which was Adam.

That is what we are trying to discover: what's best for us as individuals. I am pretty confident that my single selections are best when I am trying to examine the top end of racing.

I cannot imagine anyone in Australia betting on the worst horses. Well, I can actually. I have to, because trifectas pay $10,000 and more, so obviously a lot of 'someones' DO bet on these events. They are the same 'someones' who wander into the TAB, turn to the nearest person (whom they don't know from Adam) and ask "what do you like in this, mate?".

On second thoughts, they'd probably know Adam. Let's just say that they are the ones we are trying to beat. We can, too, on the races that matter. They bet on all of them, too. But in races that are paying less than, say, $5000 in prizemoney, we are all in the same boat unless we happen to know something.

That does happen, you know. People do know something. But there's a problem. They don't tell us!

Can you imagine someone knowing that a horse is a good-un at its first start, and telling you? If a trainer has been clever in his placement, the horse will start at good odds and the connections will be very pleased with the man they are paying big money to. If they get to the TAB or track and their horse opens at 4/6, they are going to be less than impressed. Well, wouldn't you, if you paid the bills?

So next time someone tells you he/she knows something, that his or her son or daughter has been talking to, or working with, the owner of Miss Lovelylegs, a young maiden filly, and that she is set to show a clean pair of heels this afternoon, well ... I leave it to you. Be warned, though. There's one born every minute.

Use your imagination and imagine for a moment that you trained or owned the horse, and then imagine if you would tell someone who was going to blab it all around town. Me thinks you would not!

So what about the middle of the range? I am going to get offside with some of you by suggesting that what I already perceived in an earlier issue of PPM is a fact, not an opinion. Marble Halls and Yippyio are your average good racehorses. We got it wrong in 1998. I did, any-way, along with anyone else who believed in Marble Halls as a worldbeater.

They beat each other to a frazzle in the Queensland Derby, those two, and the mistake we made was in thinking they were two potential champions. They are, I suggested earlier this year, and I stand by now, top gallopers on their day. Pick their day, though ... ah yes, that's the trick.

I am suggesting here that they represent great value when they win (Marble Halls has won at some amazing prices), and that this might outweigh their efforts when they fail. For instance, a bet on Marble Halls every start would be way, way out in front. He could never win another race and you'd still be making a profit. Not true of Yippyio, maybe. But he can turn it on, too.

If we can predict with some confidence when horses like this pair are well placed, there is a strong argument for looking at all those horses from the high middleground. These are the ones that people sometimes refer to as 'taking it in turns' on the city circuits, in good quality races, even in some of the Listed events.

They are often at very attractive odds, and they will win their share of races. Over the past two years, would you rather have had a dollar on Might And Power every start, or on Marble Halls? Or better still, a dollar eachway? And then there's the thrill side of it all, too. A horse that can win (it might not, but I am saying that it can) at 10/1 is a big thrill. Only the self-important strutters who simply have to be seen as backing the winner can afford to make
a noise about backing champions at 4/7 or thereabouts.

So it's the chicken or the egg, is it? No, that's not my point. I think we, each one of us, have to sit and think this one right through.

Where do we make our money? Where do we lose our money? What are we best at in this game? We cannot take in all levels of racing and win. We have to be selective.

I wouldn't be at all surprised if, after a few hours' checking, you discovered that virtually all your best wins came from one of the levels we have been considering, and that (just as vital) most of your losses are racked up in one other area. If that is so, this has been a worthwhile exercise.

By The Optimist