Translate into your language!

Projections Parameter Tab

Overview

This is the second tab in the Main window, and is used for choosing a specific strategy for the lottery number suggestion process. You have a wide selection of choices of parameters.

How to Invoke

Click the "Projection Parameters" tab in the Main Window.

Basic Procedure

Window Controls

"Generate" (number of suggestions) dropdown

Use this dropdown to select the number of suggestions you want. You have a choice between discrete (non-wheeled) and wheeled suggestions.

For discrete suggestions, choose from the dropdown menu the number of suggestions you want (from 1 to 25). If you want more than 25 suggestions, choose the "(other discrete suggestions)" selection from the dropdown menu. When you start the suggestion process, you will be asked for the quantity of suggestions you want (up to 999,999 ).

If you want wheeled suggestions, choose the wheel you want from this dropdown menu. If there are no wheels showing up, make sure that you have the "Hide Wheels" unchecked in the "Preferences" window. If there are still no wheels shown, then there are no wheels available for this lottery.

"Scope" date control

This control lets you select the starting and ending point for the analysis… that is, how far back in you want the starting analysis point to begin, and the ending analysis point.

It is generally recommended that you should start with the earliest drawing, so that Lotto Sorcerer can try to find longer patterns. However, there are some circumstances where you may want to choose a later date. One circumstance would be if your lottery changed its parameters at some point… for example, the lottery went from 1 to 48 balls to 1 to 52 balls. In this case, you may want to change the scope to the date where the parameters changed to the current settings. However, doing this will decrease the number of drawings available, and may lower the maximum Neural/Analysis Depth setting.

For generating actual suggestions to play in the lottery, it is strongly recommended that the ending Scope setting always be set to the last drawing in the dataset. This option to change the ending date is made available for experimentation purposes: if you set the ending Scope setting to the penultimate drawing, you can experiment with different settings and see if the suggestions generated match the final drawing in the dataset.

To change the Scope, use the date selector. To quickly reset the beginning Scope to start with the first drawing, click the "All" button; to quickly set the ending Scope to end with the last drawing, click the "Last" button.

"Neural/Analysis Depth" Slider

Use this slider to select the starting neural/analysis depth. The larger the start depth, the more accurate the suggestions become, because Lotto Sorcerer will look for longer patterns; however, the greater the start depth, the longer the program will take to generate the suggestions. Also, the greater the start depth, more past numbers are required to run the analysis. Lotto Sorcerer will alert you if you do not have enough past drawings entered. In this case, either add more drawings to the database (if possible), or decrease the neural start depth.

The evaluation version is limited to a maximum start depth of eight (8). The registered version can use the maximum number, 256 (if there are enough past drawings in the database).

Lotto Sorcerer will automatically set this control to the maximum setting, which is based on the Scope setting. If you are experimenting with the slider, and want to quickly set the slider to the maximum setting (for the number of drawings you have in your dataset), just click the "Max" button.

"Neural/Analysis Mode " dropdown

You can choose three types of engines for the suggestion generation process: Pool Temperature, Parity, Distribution and Custom.
  1. "Pool Temperature" : this looks for patterns with hot numbers (numbers drawn more often than normal), cold numbers (numbers drawn less often than normal) and everything in between.
  2. "Parity" : this looks for patterns in Even/Odd numbers.
  3. "Distribution" : this looks for patterns in the distribution of the numbers: high numbers (at the high end of the spectrum), low numbers (numbers at the low end of the spectrum) and everything in between.
  4. This lets you run your own function (created by the Scripting Laboratory). Your script is expected to take complete control over the suggested generation process, from acquiring the drawing data to posting the suggested in the Projections Results tab. Note that when the script runs, it will not be run in its own thread.

"Sectors" dropdown

This is an extremely important parameter (especially for lotteries which draw from small pools, as in Pick-type lotteries). The candidate pools should be of equal size as possible, otherwise undue weight is given to the largest pool.

In other words, the sector size should be an even multiple for the numbers drawn.

For example, suppose you are playing a 'Pick 4' type lottery, which draws four numbers from zero to 9 (10 numbers). If you use our guideline, that "the sector size should be an even multiple for the numbers drawn" , then you only have two options: 2 or 5 (because only these two numbers go into 10 equally).

What happens if you ignore this? What happens if you choose a sector size of '3', which does not go into 10 equally? The pools will not be equal size: two pools will have three numbers and one will have four. So the engine will favor the larger pool, giving you an erroneous result.

The rule of thumb is, 'choose the largest sector size which is a factor of the number of numbers drawn'. Of course, sometimes this is not possible (in the case of the 'number of numbers drawn is a prime number'). In this case, just choose the largest sector available.

Note: in the case of the Parity engine, you are limited to two sectors, because of the nature of parity: there are only two choices (even and odd).

"Sampling Size" dropdown

This parameter effects only the Pool Temperature engine and/or the Limitation filters, which rely on calculating certain statistics, which, in turn, rely on 'how far to go back'.

This value is determined in units. A 'unit' is based on the parameters of the lottery itself, where one unit is calculated as the number of drawings required for each number to be drawn at least once (if they were drawn equally).  For example, for a lottery drawing five numbers between one and 35, the unit would be seven. So, in the case of this lottery, a Sampling Size of four would mean 28 past drawings from which the sampling is gathered.

"Analysis Method" dropdown

Choose the analysis engine. There are ten choices:

  1. Pattern Recognition (neural network)
  2. Deep Pattern Recognition (neural network)
  3. Forecast 1 (small segments)
  4. Forecast 2 (large segments)
  5. Gaussian Prediction (±1 sigma)
  6. Gaussian Prediction (±2 sigma)
  7. Gaussian Prediction (±3 sigma)
  8. Sequence Projection
  9. Commonality Analysis
  10. Trend Evaluation

"Pattern Recognition" uses a neural network to find a pattern in past drawings.

"Deep Pattern Recognition" also uses a neural network, but goes to a far greater depth. This is processor (and time) intensive, especially on older computers, and moreso on "pick" type lotteries, because, in pick lotteries, each number is treated as a separate lottery.

"Forecast 1" and "Forecast 2" do not use a neural network; it uses linear regression, and is identical to the algorithm used in the FORECAST function of Microsoft Excel. Forecast 1 uses small segments in its calculations, and Forecast 2 uses large segments.

Note that it is not uncommon for the results of both Forecast 1 and Forecast 2 to be identical, depending on the lottery, its drawing history, and other settings. Forecast 1 is more sensitive to short-range trends, and Forecast 2 is more sensitive to long-range trends. 

The three "Gaussian Prediction" methods use the same technology that is in the Lottery Number Oracle function (Lotto Sorcerer menu item "Tools > Lottery Number Oracle"). This method has three settings, chosen from the dropdown menu to the right of the Gaussian Prediction selector, from which you can choose from one to three standard deviations (sigma). Note that it is not uncommon for the results of these three sigma settings to be identical, especially for lotteries that are very close to true random.

"Sequence Projection" looks at every possible pool combination, determines how often each specific pool combination occurs, and calculates which pool combination is most likely to be drawn in the next drawing.

"Commonality Analysis" chooses which pool combination has occurred more often than any other.

"Trend Evaluation" is similar to the "Commonality Analysis" method, except that far more statistical weight is given to more recent drawings.

Which analysis type should be used? Because all lotteries are different, only your experimentation can determine which to use.

"Notes" text box

This box lets you save a short note for this lottery's parameter settings. You are limited to 254 characters.