HOW TO FIND A TREND

TREND PLAYING

Horse playing world seems to be divided between handicappers and players. The first group finds pleasure in the search for a perfect formula – a perfect method which will make them winners. The other group is more interested in making profit than picking horses. Players will bet their shoe number provided that on a hundred wagers they make 15 cents on a dollar.

In the simplest terms possible “trend playing” comes down to finding a positive pattern and playing it as long as it lasts. The process begins with analysis of the first five to eight racing days on a track of interest. If a pattern can be found the player commits to it and plays systematically every race. Obviously it is possible to make subjective evaluation of the race and concentrate only on the value bets, depending on personal preferences. Continuous analysis of the pattern provides additional indicators supporting decision making. “Profiling” allows discovering sub-categories where the selected method performs exceptionally well or where the method is nonproductive. This knowledge can substantially improve overall performance of the trend.

INITIAL ANALYSIS

Let’s make a simple logical exercise. Let’s imagine that one handicapping method applied to races of the first five days at CRC produces result A. Let’s apply another method to the same set of races; it is bound to produce another result B. One of those results will be better (almost certainly). Now it seems logical that if we keep applying different methods to the same set of races one of them will be the best. The initial analysis comes down to finding the most effective method for the first five days of racing (or any number deemed to be sufficient by the player) at CRC (or any other track of choice). If the best method is producing positive results then a trend player, after verifying distribution of winners and winning percentage, can assume that the method indicates a trend.

Combination of speed and workouts represents a handicapping method. Setting influence factor of the Speed component to 1.0 and influence factor of Workouts to 0.3 makes the Speed to be an anchor and the Workouts to be a 30% supplement.

Changing the influence factor of Workouts defines another evaluation method. If the influence factor is set to 40% the result will be different than it was in the first case. Reasoning this way we can quickly define a set of methods where Speed is an anchor and Workouts are set to values from 5% to 100% at interval of five. Next we define a test evaluation method which contains all of the previously defined ones. That represents a computational tree where each leaf is a combination of Speed and Workout with workout’s influence factor varying over the full domain of possible values (0.05 – 1.0).

Executing this combined method over the initial sample of races will produce a statistic containing the number of winners indicated by each leaf of the computational tree (the FI button in the Wagering Profile view displays the computational tree and influence of all its leafs).

Here is our example

Computations are done for five racing days at CRC form 10/19/2006 (That day I was able to import first BRIS data files). The table shows that the combination of Speed and Workouts with low influence factors for Workouts produces the best results. Experienced players will have enough confidence to start playing (that would be an entry point) or at least start covering high value opportunities. The playable method would be S – W (1.0 – 0.15) on a first pick.

This table shows the trend remaining stable after ten racing days.

Here you can see the result of a month of playing assuming that the entry point was chosen after first five days.

The combination of Speed and Workouts (1.0 – 0.15) produces $45.80 on 230 races with profit of $0.10 on each dollar bet. Further analysis of the profile might provide additional information about types of races, which produce best results. In our example turf races showed very high value already within first five racing days. Once the player becomes aware of that kind of regularity he can easier capitalize on value plays and save some costs on poorly performing types of races. In other words, betting blindly all distance turf races and taking chances on short-odds dirt sprints would have been a rational way of playing.

CONTRIBUTING FACTORS

We started this discussion with assumption that combination of Speed and Workouts is the best method for CRC. Most of the time it is a rational assumption, after all basic handicapping methods have been validated many times in many places. However, it is not a universal rule. Example above illustrates how to balance two factors for maximal profit. There is still an unanswered question: “Should we use those two factors in the first place?” That brings a general question: “How to decide which factors should be used in a method?”

The answer is relatively simple. Compute statistics for all basic factors on the initial sample of races. Once you see the results they produce, select two most effective ones. Combine those two factors into a method and balance it the way presented in the first part of this article. Then descend another level in the tree (+ button) and take a look at contribution of each factor. If the combined result is better or equal to the best result of an element then we have contributing factors.

Table above shows the computational tree of the method selected for CRC calculated on 22 playing days. The “Root” shows the combined result while leafs underneath show results produced by each factor separately. Obviously, this method is composed from contributing factors because combined result is considerably better than each of its elements.

That is not always the case. Quite often long racing meets present trends, which offer solid profitability, yet considerable fluctuations of the balance point make playing those trends quite challenging.

Tadeusz Kulacz

NOTE: Users of MAXIMUS Play who would like to perform calculations presented above remember to use BRIS data files, enter all scratches and use version of the Speed where:

“Running time at current distance” set to 1.0,

“Running time plus-minus ½ furlong from current distance” set at 0.9,

all other variables in Speed set to 0.

Workouts have:

“Running time at 2 furlongs” set to 0.2,

“Running time at 3 furlongs” set to 0.3,

“Running time at 4 furlongs” set to 0.6,

“Running time at 5 furlongs” set to 0.8,

“Running time at 6 furlongs” set to 0.8,

“Running time at 7 furlongs” set to 0.9,

“Running time at 1 mile” set to 0.9,

“Running time at 1 1/16 mile” set to 0.9

“Running time at 1 1/8 mile” set to 0.9

and your results should be identical.