[tompkins]

I was one of the original testers for the first version of Speculator...IT is a great program, but it takes TIME to understand which screens are the most predictive.

Play several tracks several days on PAPER to find out which screens are predicitive for you. Remember that the longer the race, or when it is on the lawn, the more likely the incremental deceleration, OR True speed/deceleration screens are predicitive. ALWAYS look at the longshot detector screens to see what kind of pace the horse ran against. DON'T be timid about putting a SECOND line in the program to see if your first choice is SAMPLE ERROR. This maneuver alone saves many a "one race wonder" bet.

Let the program tell you form cycles by using the last three lines to see if this horse is improving, staying the same or going off form. REMEMBER never depend upon a SINGLE race out of context to the form cycle or it will be fraught with sample error. If you get your choices down to let's say A B C D, then go back and look at A1 A2 A4 (for example) against the B C D lines. Then put in B1 B3 B4 against A C D etc. to find out if there is a dominant animal here. Don't kow how often a third line back puts a horse on top BY ITSELF, but then when you put in the 1st and 2nd back you find this horse is tailing off bigtime. Infer to what the horse is going to do TODAY, not two races back.

PRACTICE with theory before you practice with real money.