James, Casey surge in Idol rankings

Two days before this week's performance show James Durbin has a slight edge on Pia Toscano as the favorite to win Season 10 of American Idol. Both carry the same odds, but James technically has a mathematical edge. Scott McCreery is still hanging on in third place, and Lauren Alaina completes the top four.

Meanwhile, Casey Abrams has made a huge leap, though it might not look that way to readers. Last week Casey was rated at 12/1; this week he is at 25/2, which is slightly lower. However, we strongly believe that after the performance night those numbers will change dramatically and that he will surge into the group of leaders. One indication of this is our new "Odds to Go." In the "Odds to Win," Casey is sixth behind Lauren and Jacob Lusk, but in the "Odds to Go," both of those performers are now more likely to go home than Casey, though this week, all three should be safe.

Naima Adedapo's generally well-received Motown performance hasn't improved her chances of winning, but it has increased the chances that she'll survive this week's Elton John-themed night. While she's still a huge longshot to win the season, she has gained some ground this week and is now nudging Thia Megia for the 9th spot. Pending this week's performances, we expect a return to the all-girl Bottom Three.

Here are this week's odds to win Season 10 (calculated not for wagering but for entertainment):

Season 10: Odds to Win

1) James Durbin (17/4)

2) Pia Toscano (17/4)

3) Scott McCreery (21/4)

4) Lauren Alaina (8/1)

5) Jacob Lusk (12/1)

6) Casey Abrams (25/2)

7) Paul McDonald (15/1)

8) Stefano Langone (37/2)

9) Thia Megia (20/1)

10) Naima Adedapo (31/1)

11) Haley Reinhart (39/1)

Season 10: Odds to Go

1) Haley Reinhart (17/4)

2) Naima Adedapo (11/2)

3) Thia Megia (7/1)

4) Stefano Langone (9/1)

5) Paul McDonald (19/2)

6) Jacob Lusk (11/1)

7) Lauren Alaina (25/2)

8) Casey Abrams (13/1)

9) Scott McCreery (31/1)

10) James Durbin (31/1)

11) Pia Toscano (36/1)

Methodology. This would never be considered scientific, but it's based on the theory that the more buzz people are generating, the more people are likely to vote for them. With that in mind, we track a wealth of sources ranging from fan forums to Vegas bookmakers to see what people are saying and predicting about the Idols. We then crunch that information into a formula that spits out the numbers you see above. Again, it's not exactly scientific and these numbers could never be presented for wagering. But it seems to work. Each year for the past three seasons, this has helped us predict the winner. In Season 7, we could see David Cook's popularity building very early in the season. In Season 8, we could see mid-season that a Kris Allen-Adam Lambert finale was in the works, and we predicted Kris' win at the end. And in Season 9, this formula sadly predicted what we didn't want to predict: a win by Lee DeWyze.

Why odds? Because we understand them, and so do other people. It's easy to look at odds quickly catch what they're pointing to. Even though these aren't true betting odds, they are a good representation of who has the best chance to win and who has the best chance to go home. At the very least, they let you know exactly who is getting the most buzz online, and that usually translates into who has the most loyal fan base.