Quote: Originally posted by 6of6 on February 02, 2004
Thanks for the stats prob, never looked at the odds for the jackpot before, very interesting, But I had my tickets when mm got to that 350 mil mark, that was when it was split 3 or 4 ways , the only thing I remember about that jackpot was ( besides that I didn't win of course) was that one winner was a 18 or 19 year old girl who bought like 5 quick pics at a gas station, never played before and won.
Well, as I see it, there are circumstances under which winning a share of a large jackpot is less desirable than winning the entire earlier jackpot. It's happened several times that one would have done better by winning on a previous drawing than one has done by winning a share of a huge jackpot. There is actually a way to measure this; it's called the expectation value, which is a measure of the total possible return divided by the odds. If the expectation value is greater than 1, the bet is called by statisticians a "good bet." The total possible return is in turn, a function of the likely number of winners. This is why these kinds of statistics are useful, and why I watch them carefully.
Because of the long odds against winning the Powerball and MegaMillions the expectation value generally doesn't exceed 1 until around the $200 million mark, although I myself play well before that. I know that it's mathematically unwise, but the fantasy alone is worth the price for me.
The extremely long odds in these jackpots have another effect: They give single winner outcomes more probable than any other outcome until the jackpots approach or even exceed $200 million dollars. When Whittaker won, a single winner was the most probable outcome at at 33%, followed by two winners at 26% followed by zero winners at 21%.
There's no practical reason why a quick pick should have any less likelihood of winning than any other ticket. The difficulty that most people face is whether or not buying more than one ticket actually increases the probability of winning by a factor equal to the number of tickets bought. In my earlier post above I didn't phrase very well what happens because of wheeling and people playing their birthdates. These factors increase the likelihood of rollovers and decrease the liklihood of large numbers of winners. If you play birthdates, you probably increase the probability of a shared prize and reduce your expectation value. (Most people would be happy to share a jackpot though, rather than curse themselves for not playing their birthdate on the day it actually comes up in the lottery.) Wheeling, if I understand it correctly, means that you cut your odds per ticket played by slightly less than you would in playing quick-picks. In spite of this I don't play quick-picks myself though. I have a program I wrote years ago that selects my numbers by a different mechanism, which may slightly increase my expectation value and assures that I actually reduce my odds by an amount equivalent to the number of tickets played. I still haven't won the lottery however, since there is no realistic way of increasing the odds of winning the lottery. There are only ways to decrease your odds to less than you'd expect when you buy a lot of tickets.