Chapter 5: What Am I Doing?
What exactly am I doing?
Well, if you have gotten anything out of this article, the truth is what I am doing probably won't matter much to you.
Chapter 1 suggests that each person has to find their own approach. So, if I give you my approach, it probably won't work for you. (Oh, I 'll give it to you, or at least most of it. Be patient.)
In Chapter 4 I laid out a way for you to develop your own strategy, but I understand that you'd like to start with a winning strategy. That would make it easier, wouldn't it? (Remember that personal responsibility thing? I meant that. It is the only way.)
Before I tell you what I am doing, I want you to consider the program we work in. The HorseStreet Handicapper has some great features. You can click this button or that button and get good answers instantly. Want to know who the logical front runners are? Click the ES button in the HQ.
And what about simply taking the output of the HQ and building a line? One user reported that he grew $200 into over $1,000 betting 1% of bank on the top horse from the Pct button (HQ) if the horse showed a $net of $2.00 or more. This is about 6 months, as I recall. (Didn't work like that for me, by the way.)
My point is that we have some very high-level answers in our software at the touch of our fingers. If we start with those high-level answers and add just a couple of meaningful questions, that should carry us to profit. By "meaningful questions" I mean building a list of "mistakes" as mentioned in the previous chapter. That is how you build a system.
As all of you know, I have chosen to avoid artfulness like the plague. Call it a personality flaw if you will, but that is who I am. Thus, I am totally systematic in my approach.
1. Contender Selection
First, I select contenders, using the Reynolds Prime 3 Ranks. (Note for readers that are not HSH users: The Prime 3 represents the best 3 factors for each horse out of our 6 "prime factors." Those 6 factors are- Overall Rating, Early Speed, Earnings Box Class, Level-based Class, Form, and Last Race. We then take the ranks for those 3 factors and multiply them together. Thus, a horse with 1st, 2nd, 3rd gets a "6" and a horse with 2nd, 4th, 5th gets a "30.")
Actually, I am still researching which of two approaches to use for contender selection. They both seem to work equally well.
The first system is a "negative contender process." That is, all the horses start as contenders then any horse that does not make it into the "half-the-field-plus-one" from the Prime3 column is eliminated. This should result in about 85-87% of the winners among the contenders, no matter what the field size.
The second system is a "positive contender process," meaning that we start with all the horses out and put them in when we find a reason to do so. The two places I am looking for reasons are a Reynolds Prime3 of 2 or less or a DK ALL score of 4.5 or more. (For non-HSH users, "DK ALL" is a form analysis rating where we compare the horses recent races against the average of all its races. A score of 4.5 or more is generally a ready-to-run horse.)
At this point I may have as few as 2 contenders or as many as 6 or 8.
The second system might be a little better at producing winners and a little worse at producing longshots because there are, generally, less contenders.
2. Tag the eliminated horses in HQ.
(For non-HSH readers, that simply means we have eliminated them.)
3. Assign a probability to each horse.
This is where Steve's book comes in. He uses "templates" to help us determine the odds line on each horse. HSH users don't need to do that because our system will create a probability. But let's understand how that probability is created.
Let us assume that we have set our "Cont Pct" (that is, contender win percentage) at 88%. (Edit> Preferences> Settings.) Let's assume a 10-horse field. If we have 10 horses, then HSH expects that we will have 6 contenders and 4 non-contenders. With the ContPct set at 88, it, therefore, assumes that the contenders will combine for 88%. That means the non-contenders will win 12%, or 3% each. (Make sure you get this before you move on.)
But what if we have only 4 contenders as Steve suggests? That means we have 6 non-contenders each winning 3% each. That means the non-contenders will win 18% instead of 12%. (Again, make sure you've got this part.)
If the non-contenders are going to win 18% then I guess that means the 4 contenders are going to split up the remaining 82%. (See why I set it higher than Steve's 80%?)
No, the beauty of this becomes apparent if you are handicapping a 6-horse field down to 3 contenders. The standard HSH configuration would say, as always, 88%-12%. (Contenders-non-contenders.) This means that the two non-contenders should win 6% each. Getting it down to 3 horses leaves us with 3 non-contenders at 6% each (18%). Therefore, the 3 contenders should win 82%. (Get it? A sliding scale.)
Then HSH uses the each horse's Overall Rating (RTG) and raises it to the ProbPwr (Probability Power) found on the Preferences page. Huh?
This is not complicated. Suppose the setting is 2.00. If the horse's RTG score is 85 then he gets 7225 points. ( 85 ^ 2 = 7225)
Thus, each contender gets a scores and the scores are totaled and "normalized" to properly split the contenders' total probability. In other words, they are splitting up the 82%.
But there is a catch. Suppose you have a horse that is given a 35% chance of winning but is 40/1. In other words, the public thinks it has a 2% chance and HSH it is 40%. The system will adjust that probability towards the public's opinion. The degree to which the public's opinion is used is on the settings tab in Preferences. It is called Tote Impact. The default is 55%. I am using "20." In other words, I am using only 20% of the public's opinion.
That button then ranks the horses from highest to lowest probability. We take the top 4 as our final contenders. (Top 3 in a short field.)
4. Play the highest-ranked horse with a $2.00 or more $net.
In other words, if the top horse is break-even or better, we play that horse. Otherwise we look onward to the 2nd, 3rd, or even 4th pick.
5. A few fine points
Suppose this is a 10-horse field and I have more than 6 contenders. I need to get it down to 6 before I do step 3. The first thing I do is use the "AOdds" button (Automatic Odds eliminations) which tosses out any horse over 39/1, and that 39/1 threshold gets smaller as the field gets smaller.
If this gets me down to the target contenders for the race, I will then apply step 3. Otherwise, I will look for other ways to eliminate down to 6. If AOdds doesn't do it, then I will use the Reynolds Prime3 Raw score to do it, simply taking the top 6 rankings.
6. The finest step of all is Steve's "filtering." While most people will see the templates as the central point of the book, in truth it is his "filtering" that is most powerful. Some of this I cannot give away here because you need to buy the book. After all, it was his idea.
But, his idea is only the starting point anyway. You must apply it yourself and "see what works for you." (That personal responsibility thing again.)
I use the PLF (Pass Legitimate Favorite) concept where it is applicable.
Summary of the steps for an HSH user
1. All horses are contenders. Determine TargCont (Target Contenders) to be half-field-plus-one.
2. Eliminate any horse not in the top half-plus-one for Reynolds Prime3 Rank.
3. Go to HQ
4. Tag the Elims
5. If too many contenders (i.e. more than TargCont) use AOdds to eliminate some.
6. If still too many contenders use Prime3 Raw to eliminate some.
7. Click "Pct" button.
8. If more than 4 contenders (3 in a small field) eliminate as many as you need from the bottom up. Break ties with the Reynolds, Prime3 Raw.
9. Once down to the correct Final Contenders, play highest probability horse that shows a $net of $2.00 or more.
Exceptions:
9. HSH underrates low-priced horses. If a horse is 7/2 and is the top horse, he is a play at $1.95. A 3/1 horse is a play at $1.90. Below that is a play at $1.85.
Maiden Races - Here is how I handle races with first-time starters. If all the FTS combined make up 29% of the wagering pool or more, I pass. That could be 1 horse or 5 horses. In HQ, click FTS button to eliminated the FTS. Click the Sum button (pool summary) to look at the percentage of the pool eliminated. Also, I will not play a race with FTS where I do not have a full complement of the necessary final contenders. That is 4 or 3 as a minimum.
Chapter 2: Finding My Approach
Chapter 3: Taking Responsibility
Leave a Reply