Posted on May 26, 2001.
• Finally, I put an end to all errors in the pick-3 software package (filename on the FTP site: LOTWON3). I made a similar statement before, didn’t I? That’s why the proverb “Trust but verify!” sounds loudly inside my skull. Ronald Reagan used this proverb every time he met with Mihail Gorbachov. I follow myself that saying as a rule. I do verify myself to the extent equal to verifying other people’s ideas. I only looked arrogant when I rejected the idea of errors in my software.
As recently as this month, May 2001, a user named DEZ reported a serious error in the pick-3 package*. I did check the error, but with my in-house software. There was no error! Normally, I should have checked the report using the same package as DEZ used, the freeware pick-3! Then the aforementioned proverb sounded louder and louder in my mind. DEZ sent me also his data files and strategy files. He handled the matter professionally. I unzipped the freeware package and tested the error. It was real! It took me along time just to find the source code. Then the painstaking action of hunting the bugs! There is a total of 159 filters in the pick-3 package. I ran the W3 reports. I checked the first draw in the file. I created a strategy file ST3.x, writing one filter at a time, the minimum level, then the maximum level. I ran POWER-3 for every iteration of ST3.x. I discovered the erroneous filters: Min_Val1, Max_Val1, Max_Tot6, Max_Pot6, Max_Val6. I corrected the code. I included the file ST3.x in the new package. It has all the filters at the tightest possible level.
The numbers in the first element of the pairs are the exact values reported in the W3 files. For example, 568 is the value of Tot1 for that particular draw. It is the minimum level of the filter. The second number in the pair (569) represents the maximum value of the filter. It must be larger than the minimum level by at least 1 unit.
The numbers above represent the tightest possible state of the filters for the drawing I analyzed (6,4,5). It was the first line in the sample data file D3. In order to check the St3.x strategy, I must delete the draw 6,4,5 in the D3 file. It is logical. We never know what the future draw will be. Therefore the data file does not contain a “future” drawing. I must also save the D3 file under a different name (I used D3.x). Running POWER-3 with a strategy such as ST3.x and the D3.x data file always generates one and only one combination. In the example, it was 6,4,5, the draw for which I created the strategy file.
Actually, running POWER-3 with min_Tot1=568 and Max_Tot1=569 always generates one combination. In other cases, just two or three filters at the maximum admissible tightness generate on combination only. LotWon uses so many filters in order to increase the probability of selecting winning levels. It is very hard, next to impossible to select the exact value of a filter. The less tight the filter, the higher the probability of selecting a winning value. In the example above, the probability increases dramatically if selecting min-Tot1=100 and Max_Tot1=1000. Of course, the amount of combinations to play will be much larger as well. Tha’s why the large number of filters: Many more options to select safe values. By safe values I mean filters with a higher probability to occur.
I made also changes to Layer 7 of the W3 reports. The 3rd condition group shows now only 1 or 2. 1 represents now “less than OR equal to” the digit in the previous draw. 2 represents an increase. I eliminated 0, because it has the special meaning of disabling a filter. Now, the filters in the 3rd condition group are disabled with 0, 0, 0, instead of 3, 3, 3.
The “tutorial” PICK3.TUT was updated for these fixes and changes.
This "job" is no joke. I am overwhelmed. I seriously consider withdrawing my freeware and the gambling matters off the Internet. I know, both friend and foe would be glad if that happened. But, realistically, it is near impossible for one individual to handle all these matters. Your best bet is to get as much as you can, especially the updates and fixes. Hold on to it, because the extravaganza might not last long.
Again, I apologize for my reaction when faced with error reporting.
Wish all of you best of luck!
Ion Saliu
* DEZ has a pick-3 strategy that can be confirmed by FFG. He is welcome to post it here.
Resources in Lottery Software, Systems, Lotto Wheeling
Resources in Philosophy of Human-Computing-Beast
Comments:
Home | Search | New Writings | Software | Odds, Generator | Contents | Forums | Sitemap