"As a player should the focus be on the odds of winning or how often people are winning regardless?"
How often (other) people win matters in two ways. If somebody else wins the same drawing you do you only get your share of the jackpot. If anybody won recently the jackpot is less than it is when nobody has won for a longer period of time. I can hope to reduce my chances of sharing a jackpot by choosing a combination I think most players wouldn't choose (and I'd prefer that there was no QP option and every ticket was based on the player choosing which numbers to play), and I can choose how big a jackpot I'm playing for by choosing when to play.
"I've looked for ticket sales data for Millionaire for Life and can't find it. "
I've ignored the game almost completely, but I presume there are smaller prizes and there's information about how many of those prizes are won. It's not perfect, but the odds for the lowest prizes and the number of winners are a decent proxy for ticket sales.
"We are yet to have a jackpot or even a 2nd prize winner with MFL."
Probability is about the chance that something might happen, not a hard rule describing what will happen. The easiest way to see that in action is to flip two different coins (meaning a penny and a nickel, dime and a quarter, etc) 100 times. Probability suggests that each coin will come up head once for every two sets of flips but because each set of two flips has four possible outcomes you'll probably only get 1 H,T in about 25 of the 100 sets. You'll probably also get about 25 with T,H, 25 with no heads and 25 with two heads.
Much like all those sets with two flips but no heads, a lottery can sell twice as many tickets as there are possible combinations and get no winners, 1 winner, 2 winners, or more than 2 winners. The odds for flipping a coin are low enough that flipping a coin 100 times will usually give you a result that's fairly close to 50/50, but with the very high odds for lottery games with really big prizes you need far more tries before it's reasonable to expect the results to be close to what probability suggests. To get results comparable to flipping a coin 100 times you'd need MFL to sell about 1.145 billion tickets (22.9 million * 50).
FWIW, here are the probabilities for a any number of winners for tickets sales for MM in multiples of the 292 million possible combinations. The probabilities for MFL will be the same (for multiples of about 10% as many tickets).
- (290.47 million tickets): 63.21% *
- (580.94 tickets): 86.47%
- 95.02%
- 98.17%
- 99.33%
* Anytime there are as many tries as there are possible outcomes than chance of a hit are 63.21%. Chance of getting a 1 when rolling a die 6 times: 63.21%. Chance of drawing the ace of spades when choosing 1 card from a freshly shuffled deck 52 times: 63.21%. Chance of winning P3 if you play it 1000 individual times: 63.21%.
The bottom line is that even if MFL has sold 70 million tickets since it was launched we should expect about a 1 in 20 chance that nobody would have won yet.
