To Johnny5:
Yours is probably the most intelligent response I've read here. This isn't meant to insult anyone else who posted to this thread, but Johnny5's response seems to be very well thought-out.
Most of you are missing an important element in your number selection: the consideration of bias. A coin toss can't be fairly compared to any lottery drawing, regardless of the number of balls used. In a coin toss, you're using one coin whose weight remains constant. It would be difficult to find two lottery balls that weigh exactly the same. Some factors affecting the outcome of the coin toss are 1: The beginning position of the coin before it is launched (heads up or tails up); 2: The rate at which the coin spins while it's in the air; 3: The altitude the coin ultimately reaches on each spin, and 4: The time the coin spends in the air. Let's say you have a half-dollar which weighs 1/4 ounce (yes, I know, but this is just an example), and you're going to toss this half-dollar 100 times. Each toss begins with the coin in a static, heads-up position. Once tossed, the coin reaches a height of exactly three feet above the launch point and, for the four seconds it's in the air, it makes thirty-one complete revolutions. There can be only one outcome: The odds are 1 in 1 that heads will come up 100 times in a row. If any one of these constants are allowed to change, by the smallest degree, then that change will have a direct effect on the overall results.
With lottery balls, the bias is determined by the weight of each individual ball (they're manufactured to be within strict specifications measured in milligrams), which can differ by micro-grams, the amount of paint on each ball ("8" will weigh a little more than "1"), the aerodynamic properties of each painted ball, the force exerted on the entire ballset by the blast of air used to mix them up and, finally, the amount of time each ball spends in the air before the column of air forces one of them up the tube.
For daily games using the ping-pong ball method, I believe it's easier to predict which numbers will probably not appear in the next drawing, thereby narrowing the field considerably. Play the left-over digits in box form.
BobP brings up an excellent point: Frequency analysis is a very valuable tool when used properly, especially in the daily games. The Law of Averages states that, over time, all components of a repetitious random dvent must achieve equilibrium. Put another way, this means that, given enough time, every ball in every ballset must come up an equal number of times, regardless of bias or other mitigating (non-constant) factors. Using 100 draws in a Pick-3 game as an example, let's say that the digits 0, 1, 2, 3 and 4 came up 50 times each; 5, 6 and 7 came up 40 times each and 8 and 9 each came up 30 times. The Law of Averages says that 8 and 9 are most likely to be drawn next, while 5, 6 and 7 are all a close second. Logic tells us that we can pretty much ignore 0 through 4, so all we have to do is wheel 5 through 9 to have a good shot at winning.
Not so fast. What if 8 and 9 both came up in the last twenty draws, while the digits 0 and 3 haven't been heard from in 40 days? This changes everything in the example above. What to do? The answer is simple: Just break your frequency analysis down by position. This is the secret to making frequency work for you. By tracking your numbers by position, you're actually following three separate, and smaller, random dvents instead of just one.
I've posted an example of this somewhere on this site, but you'll have to look it up if you're interested, since I've forgotten which thread I posted it to (I'm supposed to take ginko-biloba for this, but I can't remember where I put it).
I'm sorry to hear of your misfortune, Johnny5. My father had a similar experience several years ago, so I know how you must feel. He played the same three lotto combinations on and off for years. I told him once, "If you're going to play the same numbers all the time, you should play in every drawing." He said, "If I'm meant to win, I'll win." Several months later, I dropped in to see my folks after work. Mom was doing the dishes when I walked in, and Dad was sitting in his easy-chair in the living room. He didn't look well; his skin looked like gray plastic. I went to the kitchen and asked my mother what was wrong with Dad. She started to laugh and said, "Go ask him." He reached for a slip of paper and handed it to me. On the paper, which was terribly worn from being carried in his wallet for years, he had written each of the three lotto numbers he occasionally played, one of which was circled in pencil. "Yeah, so...," I said, "these are your lottery numbers." "They came up last night," he said. " And you didn't buy your tickets," I asked. "And I didn't buy my tickets." That was a $9M jackpot, and there were no winners. He never played again.
Hope this helps, and good luck to all of you.