When you teach yourself something, you are susceptible to picking up bad habits early on that can haunt you for years. I can attest to this in my tennis game � I�m tennis-challenged partly because of all the bad habits I developed while picking up the game for myself. When I�m taught and try to use the proper form, it feels awkward at first. But if I keep doing it, it improves my game.

One bad habit that most handicappers develop early on is trying to beat "the races." Unfortunately, in racing, there is little or nothing that works in all situations. However, if you break "the races" down into meaningful subcategories, it�s possible to find things that do work. There are all sorts of significant distinctions that you can use to create subcategories (track, class, surface condition, gender, age, etc.). I guarantee that you�ll find that what works in one subcategory won�t necessarily work in another. As a handicapper, you need to think of these different subcategories � dirt vs. turf, two-year-olds vs. older horses, fast track vs. off track, etc. � as if they were different universes.

I�d like to talk about the universe of maiden races � specifically, choosing contenders in three-year-old and up maiden races. (It would probably be better to separate maiden claimers from maiden special weights, but I don�t do that here.) I�m a proponent of quick, mechanical contender selection. One of my favorite mechanical contender-selection methods is to take the highest two of each horse�s last three Beyer numbers and average them. That gives you a contender number for each horse. (If a horse has only one ratable race, that�s the number. If he has only two ratable races, take the higher one as the number.) Cross out all horses but the top five. (Include any horses tied for fifth place.)

Another mechanical contender-selection method I use is to scan through the horses for a list of characteristics that make them automatic contenders. I base these characteristics on positive or near-positive Return-On-Investments in studies that I�ve done or collected. These ROI studies are most useful and plentiful when they focus on a subcategory rather than races as a whole. Please understand that just because a subcategory has been slightly profitable in the past, it isn't certain to continue that profitability.  However, it is likely to continue to be a higher-than-average ROI.  By picking contenders based on ROI rather than win percentage, you tend to get the value horses among your contenders and increase your chances of showing a long-term profit.

In maiden races, here are the factors that make a horse an automatic contender. (If any of you have any other legitimate ROI studies of maiden races they�d be willing to share, please feel free to email me. I�d be glad to include them in a future newsletter.) When I see a maiden with any of these characteristics, I circle him as a contender:

• If a maiden won his last race and was disqualified, he�s a contender (1.00 ROI).
• A maiden with 6+ speed points is an automatic contender (1.11 ROI).
• The best projected two-furlong time (1.01 ROI).
• A maiden was in-the-money in the last race and ran within one length of today�s par (1.01 ROI).
• A maiden was in-the-money in the last race and is running for the third time today (1.01 ROI).
• A maiden with post position 12 or higher is an automatic contender (1.01 ROI). (Without any empirical data to back this up, I usually include the outside horse in a maiden race as a contender.)
• Claimed four or more races ago (1.07 ROI).
• Within the last three races, a maiden has a Beyer rating within four points of the best last-race Beyer rating (1.00 ROI).