Computer Handicapping: “In-the-money” Ratios
by Joe Mainardi

After some time off -- for “good” behavior and a nice PRN review -- I’m back to stir up the pot!

Actually, I'm hoping that my explanation of how Hor$ense treats "in-the-money” ratios will make your lives easier.  Many people simply look at the number of “on-the-board” finishes and divide by the number of races.  This is not a good idea!  The comment "this horse has been in-the-money in his last 9 races." sounds impressive, but it doesn't tell a lot about the real value of the horse's recent record.  I use a relatively simple set of formulas to determine a ratio that is very accurate.  Simply stated, an “in-the-money” ratio (or in-money ratio, as I call it) is a weighted calculation for how well a horse, jockey or trainer is doing.  I’ll start by discussing the formulas, and then I’ll present a simple example.

The key to the formulas is deciding what win, place and show finishes are worth.  In Hor$ense, I use California’s purse distribution values, where a win is worth 55% of the total purse value, a place is worth 20%, and a show is worth 15%.  Notice that the other 10% is missing... that’s because they pay out five places.  Since horses (actually owners, but we don’t bet on them... do we?), jockeys and trainers all get paid the same way in each state, you can use any ratio; the results will always work out the same.  My scale produces a 20-point system, since 100% divided by 5% (the lowest common denominator) equals 20.

Using the Hor$ense formulas, I arrive at the value of a win being worth 11 points (55%), a place finish is worth 4 points (20%), and a show finish is worth 3 points (15%).  Since winning is the best that you can do in a race, a race is worth 11 points.  Now that we've established the rules, let's run an example!

Horse A has a record of 16 races, 4 wins, 4 seconds, and no thirds.  Horse B has a record of 22 races, 2 wins, 14 seconds, and 5 thirds.  Let’s assume that these two are the “best” the race has to offer... there I go dreaming again.  Which horse has the better in-money ratio?  It all depends on what you want to know.  With all of the exotic wagers available today, it’s important to determine what you want to know BEFORE you go looking for it!  Let’s look at the numbers, shall we?

The ratios for the win only analysis -- horse A = .25 [(4*11)/(16*11)], horse B = .09 [(2*11)/(22*11)] --  show that horse A is clearly the better horse... and that’s where my exacta wheel would start.  For the   win-place ratios -- horse A = .34 [((4*11)+(4*4))/(16*11)], horse B = .32 [((2*11)+(14*4))/(22*11)] -- horse B is getting closer to horse A, solely on the strength of all those runner-up finishes... and my exacta combinations have just gotten a lot smaller.  For the win-place-show ratios -- horse A = .34 (again) [((4*11)+(4*4)+(0*3))/(16*11)], horse B = .38 [((2*11)+(14*4)+(5*3))/(22*11)] -- horse B now looks like the better bet... and maybe a trifecta key is in order.

That's all there is to it!  The next burning question should be: "What criteria do I use to decide what information to use?"  This is going to sound like a cop-out, but you have to decide.  For horses, Hor$ense recommends that you go back six months, only using races that are on the same surface as the current race.  This simply means that you don't use turf PP's in dirt races, and that you don't use "off track" PP's on "dry" days.  The way that I get the numbers for the horses is that I look at the PP and use the final position... I ignore disqualifications, as a horse's ability isn't affected by the stewards' decision.

For jockeys and trainers, you should use the most detailed data you can get.  For example, the Daily Racing Form publishes jockey and trainer standings that are broken down into sprints and routes, turf and dirt, and so on.  If you have access to good data, or you keep track of it yourself, these   in-money formulas will make your analysis a lot easier.

As you can see, these are simple formulas.  But they are not very accurate if you don’t have at least three races to go on.  If this is the case, you may be better off avoiding in-money ratios to help you handicap that particular race.  While Hor$ense uses some extrapolation formulas to estimate the ratios, they are much too cumbersome and boring to detail here.  Besides, I can’t give away all of my secrets... can I?

You can now answer the question "How good is good?" (or "How bad is bad?") with numbers.  As long as you use valid data, these simple formulas allow you to see the results in a way that can be easily understood... even if you don't have a degree in mathematics.  Now, that wasn't so bad, was it?!?