Post a Comment On: Indeed Wrestling

When we looked at Raw Viewership changes by quarter-hour cross-referenced with WWE Superstar, it was garnered a lot of discussion about whether it "proved" Daniel Bryan could be a draw.  As Dave Meltzer has noted, the major elements that WWE is looking when they're judging whether they believe someone can make money is their effect on television ratings, their ability to move merchandise and their ability to draw (PPV buys, live event houses, etc).

I was challenged a few weeks ago to look at how do Daniel Bryan house shows draw compared to John Cena house shows.  I took a stab at it and found the domestic, non-televised Cena tour outdrew the Daniel Bryan tour by an average of about 1,000 over the few months both Cena & Bryan were active.  However, it wasn't a satisfying analysis. Not only was there only a very small dataset to compare, we weren't comparing the same cities. It seemed like we could produce some better #wrestlenomics.

So, we started with a much larger dataset:

WWE Event Results from Wrestling Observer Newsletters (6/11/08 to 2/10/14): 1,686 events.
Isolate WWE Results with Attendance Numbers Available: 1,263 events.
Isolate WWE Results w/ Attendance Numbers that were not TV Tapings: 976 events.
Isolate non-TV Taping WWE Results w/ Attendance Numbers held in USA/Canada: 666 events.

Jun 2008-Dec 2008: 43 live events (baseline) Jan 2009-Dec 2009: 86 live events (baseline) Jan 2010-Dec 2010: 112 live events (baseline) Jan 2011-Dec 2011: 127 live events (baseline) Jan 2012-Dec 2012: 136 live events (baseline) Jan 2013-Dec 2013: 143 live events (compare) Jan 2014-Feb 2014: 19 live events (compare)
Let's take Jun 2008-Dec 2012 and establish that as our baseline. This gives us 504 shows and a baseline attendance average for 234 cities (ranging from one visit to five/six visits).  Trivia: the eight cities with the most complete data were for Youngstown Ohio, Evansville Indiana, Salt Lake City Utah, Asheville North Carolina, Hidalgo Texas, White Plains New York, Augusta Georgia and Peoria Illinois.

Meanwhile, in our Jan 2013-Feb 2014 dataset we have 162 shows covering 144 cities (of which 120 cities are in our baseline.)  We'll look at these 136 non-televised live WWE events and compare which how they drew compared to the 56 month baseline.

Furthermore, events were split into four categories:

Shows with both John Cena and Daniel Bryan wrestling (BOTH) = 26 shows Shows with John Cena but not Daniel Bryan                  (CENA) = 31 shows Shows with Daniel Bryan but no John Cena                  (BRYAN) = 39 shows Shows with neither John Cena or Daniel Bryan             (NEITHER) = 40 shows
Let's look at an interesting (and extreme) example: Syracuse, New York
Baseline (2008-2012) August 9, 2008: 3,000 people December 30, 2009: 13,000 people June 10, 2011: 5,400 people May 12, 2012: 5,500 people Average: 6,725 people Median: 5,450 people
Comparison (2013-2014) March 3, 2013: 6,100 people (CENA) = -625 versus average; +650 versus median October 4, 2013: 3,700 people (BRYAN) = -3,025 versus average; -1,750 versus median

Going through all 120 cities/136 events and adding up the results, here's what we find:
RESULTS table.tableizer-table { border: 1px solid #CCC; font-family: Arial, Helvetica, sans-serif font-size: 12px; } .tableizer-table td { padding: 4px; margin: 3px; border: 1px solid #ccc; } .tableizer-table th { background-color: #104E8B; color: #FFF; font-weight: bold; }
TypeSHOWSDifference
vs Baseline Avg Difference vs
Baseline MedianAvg 1. CENA AND BRYAN 26 800 746 5,458 2. CENA ONLY 31 (177) (140) 5,916 3. BRYAN ONLY 39 (791) (671) 5,138 4. NO BRYAN/NO CENA 40 (899) (769) 4,080
Commentary Essentially, if the expected house was about 6,000 people, not having John Cena could mean things would be about 10% lower. There's not a huge difference between using median and using averages.  That implies that while some of the baseline datapoints may be inflated (such as tours during popularity swells such as Christmas-New Years), overall that isn't what's driving the results. While I compared almost 14 months of data, the major discussion of Daniel Bryan as a drawing card has been since his ascension in Summer of 2013.  Let's look at the data on a timeline basis: [Image]
You'll observe that Cena really stopped working non-televised house shows for several months from the Summer 2013 through end of the year as he recovered from his injury.  Bryan's city-by-city results were a mixed bag with some notable wins in major cities (Boston, Dallas, Vancouver) and some under achieving in other cities (Washington DC, Denver, Indianapolis).  The "Bryan & Cena" dates were all beginning of 2013 before started treating Punk & Bryan as the headliners for the B-tour.
Additional Considerations
A. Change the Time Frame to look at Bryan's Ascension If you limit the time-frame for comparison to just June 2013-Feb 2014, we lose a lot of the "Cena Only" show data but interestingly that pretty much erases the deficit between the two wrestlers.  In that scenario, the 31 Bryan only shows are down -1,180  versus the average and the 17 Cena only shows are down -1,023 versus the average.  They'd be essentially tied, but neither are providing WWE with upward momentum.
B. What about Int'l Results? This is only looking at WWE "Domestic" (US/Canada/Puerto Rico) tours.  WWE historical did much stronger internationally and tour lineups on international shows would seem to be more heavily scrutinized by fans for their favorite stars. Average European house show was 7,800 from 2008-2010 but things really cooled off in the last three years down to 6,200.  Meanwhile, "domestic" live events have actually warmed up a bit from 4,800 to 5,00 in that period.  I would attribute some of that rise to the changes we've seen over the last five years.  For instance, WWE used to tour as Smackdown and Raw brands which had a notable difference in strength: Domestic Raw drew 5,400 average while Domestic Smackdown drew 4,200 average hence the decision to move to an all "Supershow" routine.(There was a much smaller difference internationally since they often beefed up talent on the cards and only sent over touring crews periodically.)
C. Can we "fix" the baseline? There's more we could do to control for the baseline.  We could look at which touring entity came first, and establish a formula for what a "joint" show ought to draw.  We could look at the time between when WWE returned to cities and what day of the week they were touring.  We could look at who was headlining and what time of year it was.  All of these factors are no doubt influential in the attendance for the show. 
There's a lot more to learn but at least we have an interesting framework to start the comparisons.
-Chris Harrington (@mookieghana) chris.harrington@gmail.com