The Empire season...

Episode 1: 3.8 (9.9 million)

Episode 2: 4.0 (10.3 million)

Episode 3: 4.4 (11.1 million)

Episode 4: 4.3 (11.3 million)

Episode 5: 4.6 (11.5 million)

Episode 6: 4.8 (12.0 million)

Episode 7: 5.2 (13.0 million)

Episode 8: 5.4 (13.9 million)

Episode 9: 5.8 (14.3 million)

Episode 10: 5.8 (14.9 million)

Episode 11: 6.1 (16.7 million) (first hour)

Episode 12: 6.9 (17.6 million) (second hour)

Y’all. Y’ALL.

Heads up, Supernatural Showrunners: Season 10 is losing viewers faster than ever before in the show’s history

So, Dark Dynasty was the worst-rated episode ever in Supernatural’s history. I was adding that data point in to my little graph of Live+SameDay ratings that I post regularly to fandomnatural, because I am a geeky little PhD nerd like that, and I noticed a few things. First off Supernatural’s seasons always tend to lose some viewers across the season - that’s normal - but usually at a gentle rate. For example look at Gamble’s seasons in green, below. Ratings for her seasons weren’t that great but at least they were somewhat consistent. There was a nice gentle downward slope…  like a condor sailing down on a gentle breeze… 

As opposed to Carver’s three seasons, which bounce up and down like a pigeon caught in a hurricane. There’s first a huge bounce upward in week 2 of season 8. This happens to be exactly the week when the first crop of Netflix viewers would have caught up to the live show. Netflix has been feeding viewers steadily to the live show ever since, but there’s still a lot of bounces. Most of these are attributable to certain events like: Jared Padalecki’s Bieber twitter fiasco, Misha Collins’ directorial debut, Fan Fiction and the bounce it caused the following week, and of course the “Castiel and Dean actually have a conversation” lunch date promo (which resulted in the best ratings in nearly a year, and the best of all of season 10.)  The overall trend was: fans might check out for a while but they were willing to check back in when something caught their attention.

But the Carver ratings pigeon appears to have died in the middle of S10. Shortly after that Castiel/Dean lunch date promo, the ratings switch from erratic bouncing to a sudden steady downward slide that is unusually steep. This has given S10 dramatically increased variation in average ratings:

This seems to be because there was a sudden switch in ratings in the middle of the season. (the moment the Carver ratings pigeon died and began plummeting like a rock)  Check out S10′s massive drop in average # viewers per episode before vs. after the Christmas hiatus. Something changed as the show came out of that hiatus:

These numbers are still preliminary because S10 is not quite done, but this a huge drop. (note I have excluded the last two episodes from every season, so as to enable fair comparison to S10, which still hasn’t aired its last two episodes.) But it’s clear that right now, Supernatural is losing viewers more rapidly than at almost any other time in the show’s history. (exceeded, very slightly, only by season 1 when the show was still finding its legs, and finding its fandom.) 

So, what’s so different about Season 10? Several things:

- No antagonist / no greater good to fight for; “personal journey” approach; less emphasis on Sam/Dean as heroes

- Extremely slow plot progression and lack of new plot ideas: “accordioning out” of S9′s Mark of Cain plot to stretch over 1.5 years, rather than a new plot

- Extreme reduction in interactions, conversation and friendship between Dean and Castiel. (Whatever you’re trying to do here, CW, you’re doing it wrong)

- Too many “family/friends” lost w/o compelling replacements; most recently, death of the character who was supposed to represent the fandom (I know this only just happened, but there was such an instant outcry that it may have affected west coast viewership and same-day ratings). 

Message to the showrunners and to CW execs: Look at your numbers. Something is going very wrong with Supernatural and you need to fix it.

Last night’s game was the most-watched soccer match in U.S. history.

At least 17 million people tuned in to watch the Women’s World Cup match on television at 7:00 p.m. ET. Those numbers rose to an average of 25.4 million and peaked at 30.9 million between 8:30 and 8:45 p.m., once the United States victory became evident. Not only was it the most watched soccer game in U.S. history — it blew past other recent huge sports events.

Arrow Ratings Math - Show Your Work!

“Everyone is entitled to his own opinion, but not his own facts.” – Daniel Patrick Moynihan

I serve no purpose in the Arrow/Olicity fandom other than making snarky cracks & swooning over Heart-Eyes Oliver Queen on Twitter. This is my first (likely last) Tumblr post. I’m just an anal retentive data analyst who tears up over Olicity gifs (GIFS!) when she should be writing her dissertation.

Recently a “prominent” blog for Arrow’s other ship posted an Viewer Trend Analysis.  A screencap is below, but I’m not going to link to it. I see no reason to promote bullshit. There are only 6 of them-you can find it if you need to.

This blogger was attempting to use ratings data to argue that;

  • Arrow has lost 500K viewers from S1 to S3
  • Arrow viewership has trended down over the course of 3 season
  • Arrow has not experienced a viewership increase in Season 3

The implication being this downward trend;

  • Is (or should be) a big problem for the CW
  • Is the result of the Olicity becoming the main couple on Arrow

The social scientist in me thought the numbers looked wonky. It took about 3 minutes of internet research to see this blogger had used incorrect ratings data to overinflate S1 & S2 ratings.

Everyone is entitled to their own fandom & their own ships. I’m obsessed with Olicity & reserve the right to side eye Lauriver into perpetuity. If Lauriver is your bag? Do you. HOWEVER, if you are going attempt to argue in favor of your ship using “data” then you need to come correct! You are not entitled to misrepresent data & make up your own fact. The basic rules of 5th grade math still apply – Show your work. Here is mine.

Three Seasons of Arrow Data Analysis

  • If you overestimate your baseline (S1 Ratings), you can’t do any type of accurate comparison across groups. You are also overestimated between group and within group averages.
  • If you are going to do a trend analysis based ENTIRELY on averages then you have to acknowledge/consider removing statistical outlier. This blogger left significant outliers in their trend analysis.

 STATS RANT 1: If one number in a group is so much bigger or so much smaller than the others it will exert undue influence on the average. For example, say you pick 100 random Americans and average their annual income to estimate average U.S. income. If one of 100 random people you picked was Bill Gates his annual income would completely distort the average. Any reputable analysis would remove his data as a statistical outlier.

STATS RANT 2: The statistical norm for averages says that 95% of all data points in a group will fall within 2 standard deviations (the amount of total variation) above or below the average. Outliers fall in the 5% of the data beyond 2 standard deviations from the average.

  • Ratings alone are not the only numbers that matter. To get a full statistical picture one should also incorporate demo numbers into the trend analysis.

Bullshit Number #1: The claim that the S1 Ratings AVG was 3.68 & the S2 Ratings AVG was 3.28. This appears to be based on the average season ratings listed on Arrow’s Wiki page, but independently calculating the season averages finds that this overestimates the S1 AVG by .47M & the S2 AVG by .66M.

I pulled full ratings & demo data for every Arrow episode from http://tvseriesfinale.com/tv-show/ and then calculated the season average, adjusted averages, & standard deviations in Excel

The CW will always use the highest numbers possible to promote their show. Advertisers don’t care about the big highs & lows of crossovers & season premieres. They want the predictability of knowing “on the real average,” how many eyeballs are seeing their product advertised on a weekly basis. Raw ratings totals have little utility to an advertiser. They want to know the “adjusted average” of viewer (excluding the statistical outliers) to ensure they are getting “bang for their buck.” 

Over 3 seasons (so far), Arrow has had 2 episodes that are ratings outliers. The highly promoted series premiere in Season One (4.14, +.15 2SD S1 AVG) and the much hyped Flarrow crossover episode in Season Three (3.92, +.46 2SD S3 AVG) were well beyond 2 standard deviations from their respective season averages. These are great number of Arrow & the CW, but they are bad numbers for the law of statistical averages. They need to be removed the analysis. The adjusted season ratings averages decline only slightly & the standard deviation decline shows that each season’s adjusted average now has less statistical variation & better predictability. 

Did Arrow lose 500K Viewers from S1 to S3?

Bullshit (Sorta) Number #2: This dude uses the unadjusted Nielsen Ratings that show from S1 to S3 Arrow lost .44M viewers. That would get rounded to 400K not 500K viewers. But we’ve already established math ain’t his strong suit. Hilariously, the adjusted ratings DO show a decline of .45M viewers, which could be rounded up to 500K. No idea how he got this almost right since he was using the wrong data! Partial credit.

Arrow only has one demo number that is a statistical outlier, S3 Flarrow crossover (1.41, +.11 2SD S3 AVG) is again beyond 2 standard deviations from the season average & should be removed from the analysis. As with the ratings, the demo average only slightly declines, but the decline in standard deviation allows less variation & better predictability of the data.

Has Arrow viewership trended down over the course of 3 season?

Yes, there has been a decline in from adjusted average viewership from Season 1 (3.16M; demo: 1.05) to Season 3 (2.71M, demo: 1.00). That is 14.2% decline in adjusted ratings & 4.7 decline in adjusted demo. But on the average, Arrow’s ratings and demo numbers have been fairly consistent.   

Plotting the adjusted (no outlier) ratings across all three seasons, we can see that Arrow did some early huge numbers, but by halfway through S1 the peaks level out and Arrow performs mostly in the 2.5M to 3M viewers band.  Through 3x21, the adjusted ratings average across all three seasons is 2.83.

Plotting the adjusted (no outlier) demo, we can see near perfect data variation (“normal distribution”) around Arrow’s 1.00 adjusted average demo across all three season.

Has Arrow really experienced a viewership increase in Season 3?

Bullshit Number #3: Yes! As you can see in the ratings & demo averages table above & the rate of change table below, S3 Arrow is up 0.9M adjusted viewers for a modest 3.4% ratings increase & 0.9 in adjusted demo for robust increase of almost 10% over Season 2. 

Arrow’s decline in ratings from S1 to S3 should be concerning to the CW?

Bullshit Number 4: This is laughable. The CW can’t renew Arrow fast enough. It has only fallen *eye roll* to the second place show on the network because of the success of The Flash. And The Flash only exists because of Arrow. Let’s humor this shit and compare Arrow’s rating to CW’s prior flagship shows, Supernatural & The Vampire Diaries over their first three seasons. Demo numbers are unavailable for S1 of Supernatural, so we can only compare ratings. 

As with Arrow, I calculated averages & standard deviations for Supernatural & TVD S1-S3, removed all outliers, and calculated adjusted averages. Arrow has had a 14.2% adjusted decline in viewers over three seasons, far fewer than Supernatural (-32.7% adjusted) or TVD (-23.3% adjusted). Arrow, similar to Supernatural, lost the bulk of its audience from S1 & S2 (-16.8%). Arrow is the only show of the three to pick up viewers going into the third season!

Arrow modest rate of decline from S1 to S3 and picking up ratings from S2 to S3 has resulted in the CW brass popping champagne corks and giving Berlanti & Co. the keys to the car for 2 more network properties. The idea that Arrow ratings are a concern to the network defies logic. 

Is this decline result of Arrow embracing Olicity?

Bullshit Number 5:  On the average, CW TV shows lose viewers over their first three seasons. Other CW dramas have lost around 25-30% of their viewing audience over that time while Arrow has only lost 14%. In short, Stephen Amell could offer to make out with every viewer & the law of averages says the show will still decline in viewers over time. (Also, get on that CW Promo Department!) We can all hypothesize about why Arrow had its largest decline in viewer between S1 & S2, and then a slight increase in viewers for S3. Anyone’s opinion is just as valid as mine.

  1. If your theory is Arrow has had a significant drop in ratings over three season & ratings aren’t up for Season 3 you are mathematically incorrect.
  2. If you theory is that things that didn’t happen, did happen because “typical Arrow viewers” hate Olicity as much as you please adjust your tinfoil hat.
  3. Don’t make data nerds angry because you can talk all the shit you want but basic math don’t lie.

All Arrow ratings data was taken from http://tvseriesfinale.com/tv-show/. The CW comparison rating data for Supernatural and The Vampire Diaries came from http://www.supernaturalwiki.com/index.php?title=Ratings#2005-2006_.28Season_One.29;  http://vampirediaries.wikia.com/wiki/Vampire_Diaries_Ratings

Fear the Walking Dead Shatters Cable Records - Series premiere is the most viewed in cable history

Last night’s special 90-minute series premiere of “Fear the Walking Dead” made television history, delivering 10.1 million live/same day viewers and becoming the #1 series premiere in cable television history for total viewers and all key demos. (read more)

'Empire' On Track to Pass 'The Walking Dead' Ratings

‘Empire’ On Track to Pass ‘The Walking Dead’ Ratings

Fox’s newest hit musical drama series, Empire continues to break records week after week. It is officially the only TV series since 1992 to have its rating continue to rise for its first seven episodes. The last show to be as successful as Empire was Roseanne. Now, Empire is on track to beat AMC’s The Walking Dead, ratings.

According to Variety, last night Empire averaged a 5.7 rating in the key…

View On WordPress

simplyslc asked:

Absolutely loved the finale last night. The last scene between Bellamy and Clarke + the music was simply beautiful. Now, I know we are getting a season 3 (I couldn't survive if it ended like that!), but unfortunately despite the amazing writing and wonderful characters you guys give us, we need to get the ratings up! Any tips on how we can voice our love for The 100 and fight for a season 4?

I so appreciate that fans want to help. Really, you guys are amazing. 

It’s such a tough question. In terms of live ratings, it’s just a matter of getting people with Nielsen boxes to watch the show. I don’t know how you do that. Sadly, a lot of people don’t even realize our little show exists. We hear all the time, “Wow, I’ve never heard of it but this show is really good!” Which is great, we love it when people dig the show, but we seem to really fly under the radar. 

For those of you without Nielsen boxes, legitimate digital outlets do count every view. So watching the show on The CW’s site, on Hulu, on Netflix, or buying it on iTunes are all ways for you to make your support known. The CW is very wise in that digital viewership is taken into account.

International viewers: if you can watch it live when it airs in your home countries, that helps, too. International performance is also taken into account by The CW. Beauty and the Beast does very well outside of the US and that’s often touted as part of why it continues to be renewed (as it did just recently for season 4, congrats to them!).

At the end of the day, we’re just really pleased that we can make a show you all love enough to want to help. So thank you. We’ll keep making the best show we can and hope it continues for as long as people want to watch it.