Monday, January 27, 2014

Dissecting Raw Ratings by Person

We'll take a break from the Meltzer Star Analysis (parts 1/2/3) and take a quick diversion to the world of Raw Ratings. There's a been a lot of talk about whether WWE has a defense on why they haven't strapped a rocket on Daniel Bryan in light of the outspoken crowd reactions at the Royal Rumble. I decided to resurrect a project that I had started months ago, namely - dissecting the raw ratings quarter hour viewership gains & losses by person.

Methodology:
  1. Take all of the detailed Raw quarter-hour reports from Dave Meltzer’s Wrestling Observer newsletter (subscription; available online at wrestlingobserver.com) 
  2. For each “segment”, tally which people were involved. (More difficult than it sounds.) 
  3. By person, take the “average” of all their segments (noting whether we’re including the notoriously hot overrun segment) 

This data covers (though not every week has every quarter-hour, sometimes only over-run available):
2011: 52 weeks
2012: 53 weeks
2013: 42 weeks (weekly ratings only through 10/7)

This is clearly an imperfect science. It covers weeks with strong competition on television to weeks with little competition on television. It covers time-periods where you may be pushing a wrestlers as a main-event title contenders during one month and pushing them a comedy tag team goofball during a different time period. Why people do and do not tune in is not purely driven simply by who is on the screen. However, the hope is that by looking at the information over long periods of time we would be able to draw some conclusions about the trends we’re seeing. Lastly, it gets spotty since early October — this isn’t some fiendish plot by me to thwart people from evaluating Cena/Orton/Punk/Bryan in the post-Battleground world. Instead, Dave just hasn’t been providing the weekly Raw segment data in the newsletter. That’s all.

“The rule of thumb is not to overreact to one week’s rating” – Dave Meltzer
I decided to run the data essentially four different ways: with and without the overrun segment (the final period where Raw spills over the 11 PM EST hour and viewers for next USA programming tune in as well as fans that have been programmed to check out what’s going on to close out Raw) and using show averages and using show maximums. Since I was trying to tallying all the people involved in a segment, oftentimes a wrestler can appear more than once (interview, match, video package, etc.). To give the benefit of the doubt, I ran the numbers using both the “average viewership change” by wrestler in that show as well as a “maximum viewership change” by wrestler in that show.

You did have to appear in a minimum of 5 shows to be included in the calculation. This was to prevent any freak show circumstances from being overly influential and leave off guest hosts (though I think we can all agree that Wayne Brady is the key to a WWE renaissance).2011 – 2013 Viewership Gains/Losses for Raw

Biggest Viewership Gainers 2011

1 Jim Ross 344,893
2 John Cena 331,171
3 HHH 246,231
4 The Miz 241,065
5 CM Punk 165,536
6 Michael Cole 164,008
7 John Laurinaitis 104,608
8 R-Truth 101,238
9 Christian 95,217
10 Alex Riley 93,311
11 The Big Show 86,191
12 Jerry Lawler 83,522
13 Rey Mysterio 23,013

Biggest Viewership Losers 2011

1 Mike McGillicutty (336,938)
2 Zack Ryder (323,981)
3 Santino Marella (308,125)
4 David Otunga (245,324)
5 Kofi Kingston (233,407)
6 Evan Bourne (227,944)
7 Beth Phoenix (222,444)
8 Mason Ryan (218,875)
9 John Morrison (209,629)
10 Nikki Bella (196,341)
11 Kelly Kelly (194,762)
12 Cody Rhodes (191,909)
13 Natalya (184,154)
14 Jack Swagger (169,717)
15 Sheamus (150,990)
16 Dolph Ziggler (136,509)
17 Eve Torres (108,318)
18 Brie Bella (97,567)
19 Vickie Guerrero (76,179)
20 Daniel Bryan (69,000)

Biggest Viewership Gainers 2012
1 Undertaker 572,657
2 HHH 380,624
3 Vince McMahon 342,781
4 The Rock 310,149
5 Shawn Michaels 303,315
6 Brock Lesnar 290,827
7 Vickie Guerrero 223,403
8 John Cena 218,700
9 Paul Heyman 192,610
10 CM Punk 179,546
11 John Laurinaitis 179,165
12 The Big Show 153,624
13 Chris Jericho 124,365
14 AJ Lee 121,523
15 Kane 68,418
16 The Great Khali 58,000
17 Mark Henry 56,738
18 Wade Barrett 50,308
19 Randy Orton 36,935
20 Ryback 30,146

Biggest Viewership Losers 2012
1 Kelly Kelly (251,571)
2 Jinder Mahal (245,750)
3 Santino (216,836)
4 Cesaro (202,467)
5 Primo (191,933)
6 Kofi Kingston (189,067)
7 Epico (187,592)
8 Christian (186,100)
9 Zack Ryder (156,023)
10 Layla (153,273)
11 Kaitlyn (150,700)
12 Titus O’Neil (138,000)
13 Damien Sandow (134,454)
14 Darren Young (131,929)
15 Rey Mysterio (131,580)
16 Cody Rhodes (121,731)
17 Tyson Kidd (121,571)
18 Justin Gabriel (121,167)
19 Jack Swagger (105,278)
20 The Miz (95,891)

Biggest Viewership Gainers Jan-Oct 2013
1 The Rock 335,698
2 Brock Lesnar 330,068
3 John Cena 328,401
4 Paul Heyman 317,531
5 HHH 259,420
6 CM Punk 246,534
7 Stephanie McMahon 242,841
8 Vickie Guerrero 234,359
9 Vince McMahon 203,475
10 Brad Maddox 168,819
11 Seth Rollins 167,388
12 Roman Reigns 166,292
13 The Big Show 161,143
14 Daniel Bryan 155,341
15 Dean Ambrose 135,652
16 Ryback 91,294
17 Curtis Axel 59,699
18 Kane 48,198
19 Sheamus 40,064
20 Randy Orton 36,306

Biggest Viewership Losers Jan-Oct 2013
1 Cameron (496,714)
2 Layla (486,500)
3 Aksana (397,000)
4 Naomi (389,417)
5 R-Truth (322,143)
6 Brie Bella (322,133)
7 Nikki Bella (311,375)
8 Natalya (299,300)
9 Alicia Fox (234,875)
10 Christian (208,000)
11 The Uso’s (205,417)
12 Kofi Kingston (202,158)
13 Great Khali (201,200)
14 Damien Sandow (184,140)
15 Fandango (179,640)
16 Santino Marella (179,438)
17 Bray Wyatt (171,225)
18 Zack Ryder (153,643)
19 AJ Lee (152,838)
20 Erick Rowan (142,250)

The viewership number calculated here is an amalgamation of the four numbers I previously mentioned. It’s the average of w/ & w/o overrun #s split between where you’re using the average method (75% weighting) and the maximum method (25% weighting).
Again, this is quite imperfect but I must say the results do seem to align to general WWE-think. That is to say, when you look at who they push on television and which segments they put those people, there is an intent to promote certain people. While I find the results quite interesting, I do want to emphasize several points:
  • This looks at quarter-hour viewership changes. That’s how many people tuned in or tuned during the fifteen minutes measured. There’s a host of reasons that viewers tune-in and tune-out through a show. Some of it has to do with specific time periods (top of the hour, the end of show overrun). Some of it has to do with television competition – specifically major sports events like Football games. Some of it has to do with who is on the screen. Some of it has to do with who was on the screen (i.e. big-drop offs following major viewership gains). Some of it just appears to related to the unexplained fickle variations that you get from Nielsen household reporting. Also, people in the first segment can be short-changed. Essentially, there isn’t a “delta” to compare them against, so usually the participants for that entire segment don’t get credited with anything even though they were on Raw. (In fact, we know that the night after a PPV usually experiences a major 1st hour boost as people tune in to see what transpired last night.) A possible improvement would be to add a secondary variable looking at hourly Raw viewership so we could account for the people that appeared throughout all four quarters (and smooth it out a bit). 
  • This pretends everyone in each “segment” was equally responsible for driving the viewership change. If JTG is destroyed by Brock Lesnar and a half million people tune in, both JTG and Brock would get a +500,000 for that segment. Clearly, there’s room for improvement because a thoughtful analysis would consider what acts appear to be driving the quarter-hour rating and what acts just happen to have a little cameo during that time. My workaround was to try and focus on wrestlers that appeared on several shows (not just several segments, but many different episodes of Raw) as well as to look across large swaths of time for the average. 
  • This (mostly) ignores normal Raw ratings patterns. There are quarter-hours when Raw viewership normally picks up and there are quarter-hours where Raw viewership normally drops. After more than two decades, WWE has trained and re-trained their fans about when the important stuff normally happens. Interestingly, the dawn of the weekly 3-hour Raw has generated another set of viewership habits where Raw often loses viewers from the start of the show to the end of the show. WWE is hardly ignorant of the trends, and therefore it’s not surprising that they often program similar material and similar people (at least on a status basis) in the same slots week-over-week. In some ways this can become a self-fulfilling prophecy – treat someone like a goof in a blow-off timeslot and the audience will view them that way for a long time. That doesn’t “prove” they can’t draw- it just shows that the company doesn’t think enough of them to protect them. However, without a fully functioning model of who draws and repulses viewers, all we have is our scattered data points. The caveat to this was that I did throw in some safeguards around the “overrun” segment (the “big angle” before WWE goes off the air each week). The overrun segment can see a million people tune in — now, it says something about your placement in the company when you’re in the overrun segment, but on the other hand, it’s going to greatly boost your numbers the most often you’re slotted in there, and that doesn’t necessarily imply you’re the driving force behind why those viewership numbers explode. That’s why I find it necessary to look at people’s numbers with and without overrun included. 
A lot of people have asked to see the data since Daniel Bryan’s start-stop push in August hoping to prove/disprove that he is/is not a draw/failure. I’ve looked at the numbers and honestly feel that our dataset is too constrained to really pull meaningful conclusions.




In this comparison, Bryan has gone from losing viewers (-101,792) in the first quarter, to gaining viewers (+186,331) in the second quarter to strong gains (+491,967) in the third quarter and beyond. Sounds compelling right? But consider Big Show from September to November he was averaging viewership gain of +647,167. However, there hasn’t been a groundswell of Paul Wight supporters trying to prove he was cheated out of a championship run. That’s because simply using skewed viewership numbers without context is just yelling into a vacuum. You can prove or disprove whatever you want based on whatever narrative you have. If you don’t control for who is in the overrun segment (which is key because it’s such a disproportionate viewership swell) you are just going to prove that whomever was in the big angle that week, is the big draw. It’s a self-fulfilling prophesy. (And if people are wondering why authority figures draw well – that’s because when they’re on television, usually important people like the champion is also on TV.)

The reality is that WWE carefully crafts who they put in each segment. Consider how they doubled the length of the ADR/Kofi match this past week so it would fill a quarter hour originally set aside for a CM Punk interview. They didn’t want to throw off the rest of the schedule. Women’s wrestling doesn’t usually draw big numbers. Even the days of Sable driving viewership through the roof are gone. However, WWE routines books the women in the same slot and they routinely lose viewers. Hell, even great wrestling matches like the Shield versus Dolph Ziggler/Usos from 10/14/13 Raw lost over a million viewers because there was a Football competition. In 2011, Daniel Bryan wrestled Sheamus on the 3/14 Raw and lost 1.1 million viewers. There’s people like Rock & Brock who pop ratings and John Cena has been a proven ratings mover. Beyond that, people tune in the beginning of Raw after a major PPV to see what happened, and sometimes they tune in for big returns (like Big Dave Batista a few weeks ago).

- Chris Harrington (@mookieghana)

1 comment:

WWMoan said...

awesome analysis Chris, very interesting. Twitter: @WWMoan