Editorials and Articles Archive
Not Enough Cooks Spoil the Broth
In the wake of Pia Toscano's elimination, an intriguing correlation emerges
10 April 2011
We never saw it coming. And, we're sort of embarrassed by that. After ten years of watching, analyzing, and chronicling this daffy show in painstaking detail, we'd like to believe that no American Idol elimination could truly surprise us. But every year, sooner or later, our pride goeth before a mind-blowing, gut-wrenching, "They-voted-WHO-off?!" fall.
In our lame defense, our team was fully aware that a shock elimination was possible this week. Throughout Season Ten, the voters had shown absolutely no mercy towards an inferior performance, even ones by otherwise strong singers. Fact: there have been five "voting" episodes since the Finals began, and the contestants with the two lowest WNTS approval ratings of the night have hit the Bottom Three 10(!) times out of 10. Even the Semifinals had delivered an unprecedentedly strong correlation between performance quality and safety (though we'd certainly understand if Lauren Turner and Kendra Chantelle begged to disagree.)
However, with just two of nine performances falling below par this week, at least one person whom America liked on Wednesday had to be B3-bound on Thursday. The potential for a Sesame Street Effect was clear. That the third person was Pia Toscano was a mild surprise. That she was in fact the lowest vote-getter of all, however, was stupefying.
We'll let others across the Idolsphere try to explain how and what the voters are thinking - most of the top AI blogs do a good job with that sort of stuff, and besides, we don't have the foggiest clue anyway. Instead, we'll try to put a little bit of perspective on the situation, plus throw out a few statistics that might surprise you...including one, at the end, which might serve as a recipe to prevent future American Idol contestants from getting Pia'ed, so to speak.
Statistically, Toscano should have cruised into next week. Her 84 approval rating on River Deep - Mountain High was fifth-highest among elimination performances in the show's history. But even that doesn't tell the full story. Of the four ahead of her, two came in Finales, one was a duet, and one was the hapless Kimberley Locke facing Idol's all-time Dead Woman Walking situation.
Among weeks in which each Idol sang one song, and exactly one contestant was eliminated (there have been 58 so far), Toscano is the first contestant ever to exit on a 5-star performance. Just two others – Jennifer Hudson and Carly Smithson – even left on a 4-star number. The nearby table lists the exclusive club of six who went home with a 50+ rating on a 'one-and-one' week. All of them might mutter with some justification, "I wuz robbed!"
On top of that, Toscano is the first American Idol finalist to retire with a perfect game. We realized that on Thursday, but what totally escaped our attention until countless readers emailed us: Toscano is, at least for now, the highest-rated contestant ever. At a 78.8 average, she's 1/10th of a point ahead of the legendary Mindy Doo. (Be advised that this might change if and when we apply the Great Ratings Recalculation at the end of this season, though at the moment it appears that Toscano would only widen her lead.)
You know a "but" is coming, and here it is.
BUT...keep in mind that Season Ten is unlike any that preceded it. With Simon Cowell, the globetrotting Dramatic Fanatic, out of the Idol picture, the producers have focused on talent this season more than ever. Yeah, they promise that every January, but this time they actually delivered. AI10 not only has the highest average performance rating of any season, but it's by a ridiculous margin. At 53.4, it's almost two full points clear of AI8, it's nearest competitor. There have been three 1-star performances all season; just one (Naima Adedapo's What's Love Got To Do With It) since the finals began.
|Lowest rated F9 Contestant|
|AI7||Kristy Lee Cook||27.5|
The lowest-rated contestant still alive this spring is love-him-or-hate-him Paul McDonald. Coming into last week's episode, he'd sung to a 46.6 average, which he subsequently raised to 48.5. Thus far, a McDonald performance has fallen below 3-stars precisely...once. Note the chart at right, which shows where McDonald stands compared to the the hindmosts in the Final Nine of previous seasons (or the Final Eight, if there was no F9 that year.) If a chain is only as strong as its weakest link, you could lock up your house with Season 10's steel and sleep mighty soundly.
Every AI season begins with America weeding out the marginal Finalists; the contestants who churn out 2-star snoozers reliably. In some years (like, ahem, the last one), this chore seems to go on forever. Next, America turns its attention to what The Idol Guy calls "the midcard" – those contestants who offer solid potential plus occasional strong performances, but whose weaknesses are ultimately exposed by the exhausting Idol grind. This usually comprises the middle third of a season's lineup...but not this year. After just two weeks on the big stage, all eleven contestants remaining had an average rating at or above 3-stars. The midcard had arrived far earlier than anyone would have anticipated.
There's more. Due to the compressed, one-week semifinal format introduced in 2011, Season Ten reached its midcard phase in just three weeks...faster than any previous season had reached even its finals!
All this makes us believe that while Toscano "wuz" indeed "robbed", the degree of larceny wasn't quite as high as we first imagined. AI10 provided America with more talent than ever before, and less time than ever to evaluate the roster before the difficult cuts had to be made. Also, do recall that Toscano was not the first seeming frontrunner to finish last in a weekly voting tally. The first Top 11 show produced what would have been a very surprising eliminee, Casey Abrams, had the judges not chosen to use their Save...perhaps prematurely, as it turns out.
Please don't get us wrong. We at WNTS are grateful for the stronger degree of talent this season (even if we happen to believe privately that the web reviewers have been a bit kinder to many of the performances than we'd have been.) We're also on board with what we've dubbed the "drive-by semifinals" experiment, at least for one more season. (It felt rushed to be sure, but it unquestionably cut the season to the chase quickly, which is something the franchise desperately needed.) But, it's important to remember that, in nature, there are no punishments or rewards to any action. There are only consequences. It appears that one consequence of American Idol's long-overdue overhaul of its product is to put the voters in a situation where they have to make tough decisions sooner than ever before. Expect more bumps and thumps on the road to the Nokia Theater.
Our inbox the past couple of days has been swamped from correspondents asking us to address a few of the more common theories circulating around the Idolsphere. We're running out of time and room in this week's editorial to do anything formal, but we'll throw our mostly-uninformed $0.02 in. That's more fun anyway.
First: the first five eliminees this year have all been female. Does that mean the Idol electorate is overwhelmingly comprised of "Teens, Tweens and Cougars" as one outlet put it? Are they voting with their hormones instead of their ears? Is it now all but impossible for a woman to win American Idol?
Our answers are, in order: "Probably, but that's nothing new", "No more so than before", and "Of course not!"
True, we showed in a pre-AI8 Idolmetrics study that girls are at a definite statistical disadvantage to guys when it comes to getting whacked by the voters. We estimated the gender gap at 4.2% on a week-by-week basis, which certainly adds up over the course of a season. Many readers have asked us to re-run the numbers up to and including this week, which we did. The gender gap has indeed increased, skyrocketing all the way to...uh, 4.4%. Two-tenths of a point. That's it.
What everyone seems to be overlooking is that, by and large, the girls of the past three seasons haven't exactly set the world on fire. The average approval rating for female contestants has risen just 0.1 point, to 50.9. Meanwhile, the average rating of the guys has jumped one full point, from 47.6 to 48.6! Taking just Seasons Eight through Ten into account, both sexes have averaged 51, with the ladies ahead by decimal points. American Idol has successfully closed the talent gap between men and women in recent years, which we believe has made the longstanding "guy bonus" that much more noticeable to the public.
None of this is to imply that 19E and AI (and, perhaps, the nation in general) have no gender-bias/glass-ceiling problems to fret about. It merely means that the problems haven't gotten appreciably worse lately. It's just that the men have gotten better.
(And from our "Oh, By The Way" Dept.:...it's true that the first four eliminees this year were women, but it's equally true that they were the four lowest-rated contestants of the finals thus far. Toscano, the fifth eliminee, was the first to break the trend.)
Second: many fans are blaming the judges for Toscano's elimination, asserting that they have failed to provide clear enough direction to America as to who deserves to stay or go.
Our own opinon is that the judges have indeed gone a bit too easy on the contestants. Randy Jackson started the season strong but seems to be slipping back into his old "dawg, that was hot!" habits. Jennifer Lopez showed early on that she could deliver constructive criticism in a sober and evenhanded manner (and personally, we'll take that every day and twice on Sundays over Simon's caustic approach), but she too has backslid recently. Steven Tyler...well, to borrow a classic Dilbert joke, it appears the amiable Tyler's primary job description is to generate carbon dioxide, because otherwise the potted plants in the studio would die.
Still, those pointed criticisms of the judges are a canyon away from blaming them for Thursday's outcome. Until that point, every Finalist voted off the show had one of the three lowest approval ratings of the week (and that includes the Abrams Save.) If the judges' rah-rah approach is to blame for Toscano, then surely they also deserve full credit for arguably the longest stretch of justifiable eliminations in American Idol history, no?
Always remember: Correlation does not prove causality. That's a fundamental tenet of science, and it should also be a cornerstone of Idolmetrics.
Which brings us to our last topic of the weekend, and let us reiterate right from the get-go: correlation does not prove causality. Ready?
USA Today's Idol Chatter board, in which he or she opined that one reason for Toscano's early ouster is that she'd "been singing songs that have been done before on Idol, in the same style that they've been sung before." Indeed, five of the six songs Toscano sang were repeats, and most had been performed several times prior. The one song she did introduce to the series was Where Do Broken Hearts Go, and even that one was by American Idol's third-most covered artist, Whitney Houston.
"Doe," who we immodestly assume must be a regular WNTS reader, went on to document how Toscano's penchant for rehashed songs contrasted starkly to the strategy of recent Finale participants. The results were striking. Whereas the likes of Kris Allen and Lee DeWyze routinely christened new songs to Idol, Toscano and fellow Tenner Jacob Lusk have, uh...not.
Up until now, we've always looked at repeat factors on a season-by-season and episode-by-episode basis. Strangely enough, we hadn't really examined it on a contestant-by-contestant basis. The table at left is our attempt to right that wrong in one fell swoop.
The table lists every American Idol finalist from Season Three to the present, ordered by the number of times on average that each of their songs had been performed previously – their 'Average Previous Renditions'. (For example, if a singer performed ten times, with 8 first-time songs, 1 song performed once before, and 1 performed twice before, his APR would be 3/10 = 0.30.) We omitted AI1 and AI2 contestants because the show was too young to recycle material significantly often (though they certainly tried), and the tally also ignores reprise performances and those ever-adored Original Winners Songs®. Hearing them one time is usually once too often, anyway.
On your long downward journey to find Toscano, note that almost all Idol champions (denoted by gold crowns) and most of the runners-up (blue diamonds) appear in the top half of this list. In fact, most contestants who reached the Top 4 – that is, the ones who emerged from the midcard – have relatively low repeat factors. Correlation, causality, yeah, we know, but still.... Could it be that the contestants who fare the best with the voters are in fact the ones who come off as fresh and trailblazing each week even if they're not necessarily the strongest performer, ratings-wise?
Consider: Allen is regarded by most Idol fans as the biggest underdog winner ever. In college hoops terms, he's the Butler/Utah/George Mason mid-major who actually closed the deal and won the tournament. His song choices and presentations were among the boldest imaginable, to the point where even his middle-of-the-pack performances often stood out. (Case in point: Falling Slowly, which scored just 64 but is seared in most fans' memories.) That he is the champion with the highest repeat factor underscores how deft he was at making old songs sound new.
Or take DeWyze, the reigning champ. Say what you will about his pitch control (and, ahem, we have), but his song choices were usually impeccable and he rarely took the easy way out. Look at his first six performances alone, beginning with Chasing Cars and culminating in his lone 5-star offering, Treat Her Like A Lady. All were of pretty good to excellent pop songs, all were fresh to Idol, and all allowed him to show off a different facet of his personality. True, his first five ratings all fell between 46 to 69, which normally screams "Midcard!", but his intrepidness clearly made an impression on America. AI has brought us plenty of divas and Broadway wannabes, but very few singers twisted enough to try to turn Fireflies into an alt-rocker.
Presumably you've located Toscano by now, and if her position on the list surprises you, it shouldn't. Again borrowing a term from TIG, Toscano is the quintessential Epoch One contestant. That she was vocally outstanding is not in dispute. That she gave many people the faint impression of "been there, done that" shouldn't be, either.
Epoch Four is still a work in progress, taking shape slowly on our TV screens each week, and Season Ten is its transition year. How it will all turn out is anybody's guess, though it appears pretty clear now that Idol fans will not accept a return to a prior era. To quote folkie Anna Nalick, "Life's like an hourglass glued to the table / And no one can find the rewind button now / So cradle your head in your hands / And breathe...."
Or if breathing's not your thing, perhaps try...cooking? Undoubtedly you noticed who fronts the Freshness List, and if you snickered at Kristy Lee Cook's name, well, so did we...at first. But then we saw David Cook not far behind, and we marveled at the coincidence. Two (unrelated) Cooks who, in the previous transition season, introduced one new song after another to American Idol. One won. The other lasted far longer than her approval ratings warranted (but which, as one blogger at the time put it, allowed her to make "an oddly compelling journey from wretchedness to mediocrity.") Hmm....
KLC went a perfect nine-for-nine in freshness. The rate that contestants have recycled songs lately, that looks to be a batting average that no one will ever touch. Meanwhile, DC remains the only human being on earth who declined the chance to reprise an earlier performance in an American Idol Finale, preferring instead to sing three new songs. (Note to AI2 fans: Ruben Studdard and Clay Aiken weren't given the option, so save your letters.) In fact, the only two songs that D. Cook rehashed were The First Time Ever I Saw Your Face and I Don't Want To Miss A Thing...and neither was his idea. Both songs were assigned to him in the Final 3 episode by the producers and judges. Left to his own devices, the AI7 champion never, not once, chose a song that had been performed before on AI!
In short, every time either Cook took the stage in 2008 with freedom to choose, there was a 100% chance that viewers would be hearing a brand-new song. When Pia Toscano took the stage, the probability was just 17%. We suspect that John Doe is on to something. Perhaps the lesson to be learned this week is that Idol contestants are better off foreswearing a diet of pre-processed foods. Nothing, it seems, beats home cooking.
(Note to readers: Unless something outrageous happens on this weeks' shows, we will almost certainly not have a new editorial next weekend, April 16th and 17th. We have some work to do on the site, and this week's lengthy article pretty much chewed up all of our free time. We plan, however, to create a new Idol-themed Sporcle quiz to keep you sharp; check back next Sunday for the link.)
- The WNTS.com Team