About 2 years ago there was an Angles thread started by a gentleman named Boxcar on the PA site. The thread received a lot of attention and stayed live for months. During that time I received scores of email requests from current (and potential) JCapper customers asking me to perform hard database testing of the angles presented in the thread.
Ever curious as to what I might find I dove right in and did some heavy R&D. If you believed the posts in the thread the angles were supposed to be the key to the mint.
But like a lot of things I've tested - actually, this is true of almost every idea someone else has ever given me - the angles in the thread didn't produce overlay winners anywhere close to the claims made in the thread. In fact, many of the angles produced long term results WORSE than what you would get if you opened up the Data Window and just hit the ALL button.
But the research wasn't without merit. Because some of the angles produced long term results WORSE than what you would get if you opened up the Data Window and just hit the ALL button, I added preset filters to the program to AVOID horses fitting patterns defined by the angles - and in so doing enabled a very non traditional way to improve overall results.
What does any of this have to do with chartcaller comments?
When I added the ability to analyze angles in the Data Window I created an Angles Field in the PlayList File. The field itself contains embedded "codes" to designate True/False as to whether or not a given horse fits the definition of each Angle.
At the same time I was looking at these angles, I was also looking at chartcaller's comments. To make a long story short, my R&D into chartcaller's comments revealed several relevant areas (roi-wise) out of the the myriad of different things a chartcaller might try to tell you with a comment. Again, hard database testing dispelled many of the myths about chartcaller comments I had heard from other "handicappers" almost my entire life.
I added "codes" to the Angles Field to enable the Profile Marker and Data Window to identify horses fitting the relevant categories turned up by my R&D.
I then added Preset Filters to the program - enabling the user to write UDMs to either "grab" or "reject" horses having eventful races in their records as described by the chartcaller. I created the filter codes in a way that enables the user to further break things down by the number of starts back that the incident took place.
The Preset Filter Codes covering chartcaller comments are:
BORE= require bore in/out LR or 2bk BORE1= require bore in/out LR BORE2= require bore in/out 2bk BORE3= require bore in/out 3bk BOREX avoid bore in/out LR or 2bk BORE1X avoid bore in/out LR BORE2X avoid bore in/out 2bk BORE3X avoid bore in/out 3bk TROUBLE1= require altered trip LR TROUBLE2= require altered trip 2bk TROUBLE3= require altered trip 3bk TROUBLE1X avoid altered trip LR TROUBLE2X avoid altered trip 2bk TROUBLE3X avoid altered trip 3bk WIDE1= require wide trip LR WIDE2= require wide trip 2bk WIDE3= require wide trip 3bk WIDE1X avoid wide trip LR WIDE2X avoid wide trip 2bk WIDE3X avoid wide trip 3bk COMMENT= require favorable chart comments COMMENTX avoid favorable chart comments
As TPW suggested, if you do the work, you just might find that some of these Preset Filter codes have the ability to improve the strength of (at least some of) your UDMs.
-jp
.
~Edited by: jeff on: 5/10/2008 at: 3:59:31 PM~
|