Written by Ion Saliu on January 19, 2002.
"All's well that ends well."
"MDIEditor and Lotto WE" is finally here!
• I had announced earlier that 1_WE was ready, except for the tutorial. Nick Barker wrote it. I made a contribution as well. Then, I announced a delay as I came across a better randomizing function. That function was implemented in version 5 of the online generator (IonSaliuGen). I refined the randomizing function further. It is now not only time specific, but user specific as well. The random seed varies widely, between millions and billions. Chances are no two users will ever use the same random seed. Some users have found a problem with my software. They complain “It's no good to use winning software if it has many users. The prizes will decrease!” Well, I see it as a positive. One should take advantage of a situation with few users in each lottery jurisdiction. That's the reality now.
On the other hand, I thought it was a good idea to spread around the winning occurrences in some situations. One such situation would be running MDIEditor and Lotto WE with relaxed (loose) filter-settings. The new randomization scheme makes virtually impossible for two users to win in the same lottery jurisdiction, in the same drawing. Except, perhaps, for a very large number of combinations.
If tight filters are used, however, chances are several users can win the same game, the same drawing.
I expressed in my previous post the importance of randomness. Many are asking the question “What is randomness, after all?” Some reduce it to “unpredictability”. “It is random, therefore it is unpredictable.” Some come closer to the real nature of it and consider randomness to be a process without order.
My philosophy is that randomness is the very essence of the universe. The forces in the universe interact randomly. It is the fundamental logic of the Totality. If the elements would not interact randomly, an outside force would have to enter the picture. There is no “outside” as far as the Totality is concerned. The illogical counter-argument would be “an intervening force from the outside that imposes a rule to the interaction.” That force would necessarily require another outside force which, in turn, would require another “superior force”, and so on ad absurdum. The universe has, however, lower logical levels. There are systems and subsystems. The level of organization can be higher.
Therefore randomness appears to be...inexistent. In fact, we are dealing here with various degrees of randomness. A degree of randomness is defined by a level of the probability of appearance of the effect of interaction. I'll exemplify by the lottery. There is a number of lotto balls in a drawing machine. The degree of randomness is high if no outside force imposes any rule. That is, the “drawing is fair”, as they say. The “outside ruler” would be members of the drawing team (including the auditors!). Let's consider the opposite level of randomness: very low (or a strongly ordered system). The “outside force” (members of the drawing team) checked all the tickets played. Any player, but themselves did not play this combination: 1,2,3,4,5,6. They add a special material to six of the lotto balls (maybe flubber?): 1,2,3,4,5,6.
The six balls are so light that the probability is very high that all six will rocket out the machine. In this case, the probability of appearance is very high; therefore the degree of randomness is very low. The drawing team has a very good chance to share a big jackpot (and run away from future investigations!). But it is not certain, for there is no logical certainty. An earthquake can strike the area seconds before the drawing! Or a tornado! How about vagabond asteroids (random wonderers)? Or a huge black hole was born in our galactic neighborhood. Potentially, the entire solar system could be sucked in… So much for “unpredictability”! It's all “predictable”. The genius is in the details. The probability is extremely detailed. I might say, nothing is unpredictable, nothing is certain. Almighty Number takes infinite forms, between 0 and 1.
•• Then I discovered an astonishing behavior of one MDI-specific filter: Ion5. It could reach very high levels, although it was not immediately evident! That filter depends on the biggest number in a lotto game. The user might be deceived by the real setting. I needed to increase the minimum number of draws in a data files. The new minimum setting closely resembles the requirement in LotWon32. The lotto games require at least 10,000 combinations in their data files. Obviously, only a few hundred draws may be real. The rest are added by number generators in the application. I passed the new “trick” to Nik (Barker, that is). He put it to a test in his 6/49 lotto game. He discovered the same “abnormally” high levels of Ion5: over 400 with a 10,000+-combination file. The frequency is 2-3 times in 100 draws.
Combinations such as:
1 3 11 12 20 24 71
26 29 31 32 39 46 203
1 2 3 8 30 37 81
30 35 41 44 46 49 245
5 6 16 17 19 22 85
26 35 38 39 44 46 228
4 8 10 12 15 30 79
29 32 39 46 47 49 242
28 36 37 38 42 47 228
3 5 7 10 14 20 59
28 34 38 39 46 47 232
20 31 32 35 43 45 206
24 33 42 43 44 47 233
23 38 40 44 47 49 241
28 34 41 45 46 49 243
The 7th numbers represent the sum-totals.
Those are the kind of combinations that some lotto players “die” for. You can notice 4 numbers from the last “decade” (40-49). Or 3 numbers from the 30's combined with numbers from the 40's…
Nik tried to see how many combinations the program would generate in a “hit” situation. He let the application run for hours, I think 14 hours in one case. No combinations generated. I checked the draws for the cases when Ion5 was 400+. I discovered that some draws were just outside the ranges of the normal probability rule. I relaxed the inner (innate) filters to accommodate the type of combinations generated by a tight Ion5. The new inner filter settings cover better the extremes of the bell curve. This new strategy, consisting of just one filter setting has a huge potential. It can hit the jackpot with only a few tickets! The users need patience more that money to play!
They need also a fast PC! "MDIEditor and Lotto WE" has very demanding filters. Meanwhile, as I said, Windows is like the person who was hired as a rocket scientist based primarily on looks! I made nevertheless efforts to improve the algorithms. The increase in speed is dizzying, as far as unfiltered number generation is concerned. The new “Odds and Random” routines generate well over 100,000 combinations per minute! In the case of my 6/69 lotto game, the “Odds and Random” module generated 150,000 combinations in one minute. I used next the “Optimized” function with a 20,000-combination data file. I only used the inner filters (all external filters disabled). The program generated 500 combinations in one minute. Only the inner filters have an efficiency ratio of 300. The application has also a new function that strips off duplicate combinations. Many know by now, that when combinations start to repeat in a large volume, chances are good the winner is among them. Provided that, of course, the filters were set correctly…
This software is no longer licensed.
This message is still kept online as it contains useful information on using my freeware.
Resources in Lottery Software, Systems, Lotto Wheeling