One ticket costs 2$.
Each ticket's odds of winning are 1 in 175 million.
So the tickets might actually be a favorable gamble if the payout meaningfully exceeded $350 million after taxes as a lump sum. However, if you crunch the numbers, you'll see that's not the case.
The headline is that you'll win $550 million, but that's a payout over a long period of time. If you take the cash as a lump sum, that drops to $360 million.
If you assume a 35% tax rate, the payout drops to $234 million.
You may also need to consider state taxes which could reduce this total by more than $30 million. Of course local taxes could further reduce this, depending on where you live.
There are a number of other smaller prizes, and if you analyze the math on that, they can increase the odds adjusted payout by about 17.5%, which is nice, but not enough. All non-jackpot prizes are fixed amounts. Most of the value of those lesser prizes comes from the $1 million dollar prizes.
If you opt for a power play it costs 1$ extra per ticket, and the cost-adjusted, odds-adjusted payout on lesser prizes increases by 59%. However, because the jackpot size does not increase, cost adjusted payout for the jackpot decreases by 33%, so your odds adjusted total payout per dollar drops by 19.5%. In short power play is not a good deal here.
I created a spreadsheet on all that :
Then you need to account for the fact that more than 175 million tickets may be sold. Some estimate that 189 million tickets will be sold. It's worth noting that ticket sales have more than doubled when compared to Saturday's $325 jackpot that nobody won. The idea that ticket sales might double when the jackpot increases by 70% is interesting, because it shows that ticket sales can increase faster than the size of the jackpot itself, which helps make sure that the odds favor the lottery and not ticket buyers.
There is always a chance the prize may be split, even if a small number of tickets are sold. However, if more tickets are sold, those odds increase.
Further, the estimated jackpot size increases as more tickets are sold.
So when you're estimating your odds of payout, you're doing so based on estimates of how many other tickets may be sold, and that estimate may prove to be incorrect.
The math behind reducing the jackpot for the possibility of a split ticket is interesting. The Poisson distribution is relevant there. The procedure is to calculate lambda which in this case would be equal to the number of tickets sold divided by the odds of winning. Most spreadsheet programs allow you to use Poisson formulas to calculate the odds of a given outcome depending on lambda. You then have to calculate the odds of 1 winner, two winners, etc., and the payout for each potential outcome and then sum that up. I did that in a spreadsheet. It's worth noting that if you have a winning ticket, that does not change the odds that others have won or lost, and this is really a flaw of some of the previous attempts by some other websites to calculate these values, which fortunately have since been corrected. Thank you so much to Mark Eichenlaub
for pointing that out. The result is that a 360 million dollar payout becomes a 220 million dollar payout, a reduction of 39%. Here's the spreadsheet on that :
If you plug those numbers into the original spreadsheet and you can get a better estimate of the odds-adjusted payout.
If you do that you can see that for a $2 ticket the total odds-adjusted lump sum payout may be approximately $1.62. If you factor in 35% taxes, this drops to $1.05. Of course this would be lower still with state and local taxes.
You might wonder just how big the jackpot would have to be in order for the odds to be in your favor, but the reality is that if the payout goes up, the number of tickets sold will likely go up as well. If ticket sales doubled and the jackpot went up 70%, the payout after taxes would be $1.16. Of course it's difficult to know what ticket sales might be, especially since the record for jackpots is not that much larger than the recent total and jack pots 70% larger than the current total may well attract additional attention due to their sheer size.
Update : On November 28th, 2012 the jackpot totaled $579,000,000 and there were two winning tickets. Officials had expected to sell 105,000 tickets per minute, but actually sold 130,000 a minute, raising payout likelihood to almost 75%.
 How Much Tax Would You Owe On A $550 Million Powerball Jackpot? - Forbes
 Are they ever in your favor?
 Numbers drawn for record Powerball jackpot
 Winning Powerball tickets sold in Arizona, Missouri