An inevitable march to the final two?

[Analyst]

Here's a simplistic analysis based on some comments that Stellar Wind (I think) made. I am assuming that all of the contestants pick good songs and sing well - this is not predicated on any particular singer stinking up the stage, which would change everything.

The first assumption is that Melinda and LaKisha share to a very large extent the same demographic. In other words, if LaKisha gets voted out, a large portion of the her votes will shift to Melinda and vice versa.

In a similar vein, Blake and Jordin's demographics overlap to a reasonable extent. Again, this is very simplistic - there are 12 year olds voting for LaKisha and 90 year olds voting for Blake - but can serve for a first-level analysis. However, I don't think the voters will switch quite as much in a block as above - in particular, if Jordin goes then I think Melinda/LaKisha will pick up some votes even though Blake will probably get the lion's share.

What this means is that whoever gets voted out this week will propel their "partner" to the finals. If LaKisha/Melinda goes, then Melinda/LaKisha should pick up enough votes from their new constituency to guarantee that she makes it the finale. Same with Jordin/Blake. Therefore, there is virtually no chance of either a Melinda/LaKisha or Jordin/Blake finale.[1]

Based on all available evidence (DI, B3, performances, judges' comments) I don't think that LaKisha beats Melinda in any reasonable scenario. Therefore, LaKisha is either out this week or next.

Blake likely picked up a sizable amount of votes from the Chris R. block; given that he was quite possibly already ahead of Jordin (if not, he was almost certainly close), I think this gives him a significant advantage over her.

Bottom line: I think it's highly, highly likely that LaKisha goes out this week and Jordin the next. Whether Blake or Melinda wins will depend on how Jordin's votes split.

[1]Not all voters switch to another singer, of course - some just stop voting. If Melinda were to go out this week and a lot of her voters give up, I could see LaKisha remaining in third place, leading to a Blake/Jordin final. I think this is a very unlikley scenario if only because I don't see LaKisha outlasting Melinda.

[thing85]

Good analysis, Analyst (ouch, redundancies). I agree with all of that and based on the assumptions you made, those conclusions seems likely.

To maybe take this analysis further, here's a question I'd like to pose: Let's assume ALL the contestants have bad performances - obviously this is purely hypothetical and WON'T happen. Who is the most volatile? Who is affected most by a poor performance and who is affected the least? Assume that the general public assumes everyone to be equally bad.

[GoRedBirds]

To me it seems that most of the people I talk to are EITHER Melinda or Lakisha. It's almost like a RedSox/Yanks thing. I can't see many of the people I know who like Melinda ever voting for Lakisha, and likewise Lakisha's voting for Melinda. They're somewhat polarizing. Either you like Melinda and Lakisha is an also-ran or you like Lakisha and Melinda has no neck or personality. At least that seems to be the case on these boards.

[thing85]

Yeah, I think that's the case. I wonder, how strong are these fans? Are Lakisha fans anti-Melinda fans? In other words, if /when Lakisha goes, will these fans vote for Blake and Jordin because they view Melinda as the "rival?" Maybe I'm taking the baseball analogy too far. It's possible though.

[revots33]

Someone recently made a very good post about the object of idol is to get VOTES, not VOTERS. I agree completely, which is why these scenarios are hard to determine in advance. The quality of a specific performance influences how hard their fans vote (or not).

So let's say for argument's sake Blake has more fans than Jordin. But let's say Jordin does awesome and Blake does just so-so. Jordin's fans will be motivated to vote more, determined not to see their fave go home after such a good performance. Blake's fans will still vote but maybe for an hour instead of 2. A weak performance is very de-motivating to the fans.

I really think this is why Lakisha didn't go home last week. She may have a smaller fanbase but they voted like hell for her after she did so well last week.

[thing85]

What about a contestant's group voting harder after a poor performance and/or poor judge comments? They know their favorite is in danger and feel more compelled to keep him/her around.

I can honestly see it either way.

[revots33]

[ QUOTE ]
What about a contestant's group voting harder after a poor performance and/or poor judge comments? They know their favorite is in danger and feel more compelled to keep him/her around.

I can honestly see it either way.

[/ QUOTE ]

HMM maybe but I still think good performances are rewarded with more average "votes per voter" than bad/mediocre ones. I could be wrong though.

[thing85]

I'd say you're right with the average, casual voter. Instead of voting once, he or she may vote 3 or 4 times for a good performance. I think the hardcore fans who have been loyal throughout the season are more inclined to power vote their favorite who does not appear safe. It's hard to really tell which amounts to more votes.

I think the most dangerous recipe here would be a contestant with a loyal fan base (although now that we're in the top 4, I would assume everyone has some sort of loyal fan base) who puts on a mediocre performance that makes him or her seem safe but doesn't wow anyone. This would keep the casual voters from voting more than usual and the hardcore voters from fearing that their favorite is in danger.

[GrannyMae]

[ QUOTE ]
[1]Not all voters switch to another singer, of course - some just stop voting.

[/ QUOTE ]

i think it has been shown that most do stop voting.

i know this is anctedotal, but everyone i know that votes, stops completely when their horse is gone.

[StellarWind]

[ QUOTE ]
i know this is anctedotal, but everyone i know that votes, stops completely when their horse is gone.

[/ QUOTE ]
Can this idea be reconciled with the weekly increases in the total number of votes?