Wednesday, February 12, 2014

Bury Points and Match Lengths

During my conversation with Matthew Timmons about his Kayfabermetrics website, one interesting idea involved the bury points metric.  Essentially, the idea is create a differentiation between quick losses and significant battles where the defeated wrestler may actually come out looking better than when they started.

The timeframes selected for points were:
  • Under Three minutes
  • Three minutes to Eight minutes
  • Eight minutes to Fifteen minutes
  • Fifteen minutes to Twenty minutes
  • Above Twenty minutes
For instance, losing in two minutes (a real squash) would hurts your reputation (gaining negative points) while the break-even point was hanging for fifteen minutes with people lasting twenty minutes or even gaining positive points.

While intuitively these lengths make sense, I wanted to comb through some data, and see what the wrestlenomics might suggest.

Here are the Average, First Quartile, Median and Third Quartiles for two types of matches:
SINGLES MATCHES - which were strictly one wrestler versus one wrestler (no handicap, no multi-man)
TAG MATCHES - which were strictly two wrestlers versus two wrestlers (no multi-teams, no six-man specials)

AVERAGE/MEDIAN/QUARTILE MATCH LENGTH FOR PPV/RAW/SMACKDOWN 1993-2013


Data Source: CageMatch data from 1993-2013.  Shows were classified based on their description and matches were classified based on number of teams/participants.  Only matches that had match time listed were included.  Matches that aired before the PPV (when part of Free for All, Kickoff or Pre-Show) were counted as PPV though matches that were part of Sunday Night Heat did not.

PPV

PPVSINGLES
MATCHES		SINGLES
AVERAGE		25thMedian75thTAG
MATCHES		TAG
AVERAGE		25thMedian75th
19932610:236:479:3714:01513:2010:4413:4114:22
19942711:415:538:5714:3059:206:457:419:16
19955111:257:1710:0214:571111:128:0810:5414:09
19966612:586:3011:5016:60119:337:4210:1011:06
19975712:548:4612:1116:501211:179:0410:3212:36
19987110:095:458:3013:51229:387:299:0010:58
1999839:445:398:3712:011610:307:0810:5114:10
20006110:434:598:0213:01188:457:068:2310:46
20017012:126:3011:1216:14149:446:498:5710:15
20028011:566:3110:4715:36168:485:187:139:42
20035812:247:0612:0816:46159:206:587:5511:07
20047212:236:2210:2516:12159:597:059:3513:47
20057812:306:2511:3717:16139:497:378:3312:22
20068011:077:249:4414:281411:428:1411:0413:31
20077911:127:1410:3615:021412:328:2911:2015:30
20087212:248:4211:1216:4568:136:168:009:57
20096713:027:2011:5716:04815:5911:3815:5220:26
20106411:576:3412:1615:5286:234:486:097:11
20116812:087:1510:5314:601010:307:139:3412:52
20126311:505:4610:2016:51109:097:128:2711:38
20136912:137:029:5916:57138:596:177:2412:14

PPV Singles matches a little under twelve minutes and PPV tag matches average right around ten minutes. I'd consider a "short" PPV match anything less then seven minutes (singles) / eight minutes (tag). I'd consider a "long" PPV match to be anything more than fifteen minutes (singles) / twelve minutes (tag).

 You'll notice the uptick in average tag match in 2009.  This was completely driven by DX PPV matches which went 18-23 minutes while PPV tag matches without HHH/HBK were back at the normal 10-14 minutes.  You'll also notice that our dataset for PPV two on two tag matches is pretty scarce.

RAW

 RAW
SINGLES
MATCHES
SINGLES
AVERAGE
25th
Median
75th
TAG MATCHES
TAG
AVERAGE
25th
Median
75th
1993
83
5:52
2:39
3:51
7:04

17
5:57
3:32
4:14
8:00
1994
55
6:23
2:53
4:37
8:32

8
8:49
4:41
9:04
12:10
1995
53
6:15
2:46
5:46
7:39

10
5:35
3:02
4:46
8:09
1996
66
7:09
3:43
6:11
9:49

12
6:45
4:17
5:57
9:19
1997
144
5:56
3:07
4:23
8:31

45
6:09
3:39
4:55
7:54
1998
186
4:08
2:30
3:27
5:00

68
4:34
2:59
3:56
5:44
1999
239
3:39
1:48
3:00
4:50

82
4:12
2:26
3:46
5:11
2000
210
3:43
2:19
3:14
4:49

123
4:10
2:54
3:50
5:01
2001
221
4:04
2:26
3:23
4:46

75
5:18
3:23
4:35
7:02
2002
213
4:20
2:20
3:21
5:33

82
5:30
3:33
4:43
6:44
2003
172
4:49
2:30
4:04
5:56

85
5:39
3:10
4:35
7:09
2004
169
6:45
2:47
4:05
7:55

50
8:36
4:21
5:41
13:35
2005
192
6:17
2:37
4:00
9:18

38
7:39
4:11
6:11
10:00
2006
204
4:54
2:00
3:31
5:57

41
5:04
3:00
3:52
5:00
2007
187
5:08
2:00
3:35
5:57

58
5:32
2:53
3:52
6:31
2008
196
4:52
2:04
3:20
5:50

69
5:17
2:53
3:48
8:15
2009
194
4:55
2:13
3:25
6:10

53
5:42
2:40
3:40
7:17
2010
202
4:51
1:56
3:18
6:59

60
4:48
2:31
3:19
5:47
2011
195
5:24
2:14
3:48
8:23

61
5:58
2:40
3:53
9:39
2012
254
5:07
2:00
3:10
7:21

68
6:53
3:11
6:45
10:36
2013
287
6:44
2:45
4:35
10:42

66
5:43
3:26
4:28
6:10

Your average singles match on Raw is a little over five minutes long.  Your average tag match on Raw is a little under six minutes long (it's above half-a-minute longer than a singles match).
I'd consider a "short" Raw match to be under three minutes and a "long" Raw match to be over seven minutes (singles) or eight minutes (tag).

 I think it's remarkable to see how in the Attitude era (1998-2001) how short the average matches were.  In many ways that statistic reflects on how the era earned the name "crash TV".  It will be interesting to see how 2014 average singles matches compare to 2013; will three-hour Raws result in more long matches or just more matches overall?

SMACKDOWN

SM! 
SINGLES
MATCHES
SINGLES
AVERAGE
25th
Median
75th
TAG MATCHES
TAG 
AVERAGE
25th
Median
75th
1999
104
3:12
1:47
2:38
3:57

33
4:34
2:55
4:15
5:37
2000
199
3:42
2:30
3:24
4:37

108
4:14
3:03
4:02
5:09
2001
194
4:20
2:35
3:45
5:23

74
5:17
3:39
4:30
6:03
2002
202
4:41
2:40
3:35
5:44

76
6:16
3:60
5:04
7:53
2003
205
5:36
3:08
4:24
6:03

65
6:50
4:16
5:10
7:06
2004
148
6:17
2:50
4:19
8:00

50
8:31
4:35
6:29
11:55
2005
170
6:25
2:20
4:00
7:24

41
8:22
3:20
7:46
12:53
2006
203
7:02
3:06
4:58
9:11

48
6:56
3:27
5:45
8:23
2007
203
7:01
3:16
5:07
9:42

52
8:27
3:29
6:44
13:49
2008
206
5:56
2:30
4:24
8:19

51
5:31
3:22
4:44
6:32
2009
182
7:04
2:53
4:52
11:23

49
7:38
3:30
6:03
10:43
2010
188
7:10
2:29
6:03
11:30

49
5:55
2:36
3:45
9:48
2011
232
6:11
2:22
4:04
9:52

50
6:09
2:32
3:59
9:13
2012
222
5:18
1:53
3:14
8:47

51
5:02
2:41
3:34
5:56
2013
188
5:41
2:29
3:53
9:12

46
5:44
2:35
3:47
8:02

The average singles match on Smackdown is under 6 minutes and the average tag match  on Smackdown is a little over 6 minutes. Again, a "short" match on Smackdown would be under three minutes for both tag and singles varieties. A "long" match on Smackdown would be over eight minutes (singles) or nine minutes (tag).

 Again, you can see how a show born during the attitude era started off with plenty of star power but quite short matches.


Overall, you can see the curve of both singles and tag matches and how they both peak early (three minutes for singles matches, a little under four minutes for tag matches).  More than half of all televised matches across Raw, Smackdown and PPV were less than five minutes long.  (You can almost hear the road agent telling the guys to get out there quick and they've got three minutes to do their stuff.)

Getting back to the original question, what might be five time lengths I'd use for singles bury points (if I was applying them uniformly to Raw/Smackdown/PPV matches?)

Reviewing Original Propositions for Singles Matches matchlength for Bury Points

A. "Under Three minutes" (covers 28.5% of singles matches)
B. "Three minutes to Eight minutes" (covers 42.0% of singles matches) 
C. "Eight minutes to Fifteen minutes" (covers 20.6% of singles matches)
D. "Fifteen minutes to Twenty minutes" (covers 5.83% of singles matches)
E. "Above Twenty minutes" (covers 3.0% of singles matches)

 Right now 91% of singles matches land in first three buckets.  That may be a little skewed for me.  If I was recalibrating, here's my take:

I would probably keep bucket A (under three minutes) the same since that's near the first quartile. I might scale back Bucket B a little so it only covered about a third of matches (3 minutes to 6 minutes). Then I'd adjust Bucket C to be (6 minutes to 10 minutes) which is about 16% of the singles matches. Bucket D would be (10 minutes to 15 minutes) about 13% of singles matches and bucket E would just be "above 15 minutes" which is 8% of singles matches.  If you really want to subdivide bucket E, I'd probably put the marker at 18 minutes which not only surpasses a quarter hour (their general unit for marking of live TV shows) but leaves an elite bucket for the less than 5% of singles matches that last 18+ minutes.  It's possible that what works for TV Bury Points may need to be adjusted for PPV Bury Points because average matches are so much longer on PPV.

That's just my thoughts.  I think Matthew did a great job with designing the Kayfabermetrics and until he has several year's completed, it will be difficult to judge how well it works or what false positives may arise. Hopefully throwing down all of this data helps paint a broader picture for everyone to start considering what it really means to wrestle a match for so long that even when you lose, you looked competitive throughout.

-Chris Harrington (@mookieghana)

If you haven't already, please do check out Matthew's website: Kayfabermetrics

ADDENDUM:

I realized that I had combined both men's, women's and mixed tag matches in my metrics.  Here's the data broken out by gender.  (There's very few man vs woman singles match - most are either Chyna or strange gimmicks like Kane vs Maria.)


Posted by at No comments:

Wrestlenomics Radio: Episode One with Kayfabermetrics


http://www.voicesofwrestling.com/2014/02/12/wrestlenomics-radio-kayfabermetrics/
mp3: http://traffic.libsyn.com/voicesofwrestling/001.mp3

Host Chris Harrington (@mookieghana) is joined by Matthew Timmons of Kayfabermetrics for 90 minutes of discussion on the art and artifice of creating pro-wrestling statistics. We cover inspirations, dataset challenges, inventing custom metrics, key learnings, great project names, overcoming hurdles and much more. Matthew explains his novel approach for looking at “velocity” going into a PPV, what it means to be pushed (or buried) based on match length, how to rate opponents’ strength and going far beyond the dull win-loss metrics that don’t really illustrate the full story. Also, we look at the weirdest keywords to gain readers, what tools we use, the mistakes that keep us up at night, the lonely life of making arguments using numbers and future goals.
* You can see Matthew’s work at www.kayfabermetrics.com (@kayfabermetrics) and Juice Make Sugar.
* Mookie’s work is up on his website is at indeedwrestling.com and indeedwrestling.blogspot.com (@mookieghana).

It's my return to podcasting after a long, long break. I thought it would be fun to talk about wrestling statistics with other people who are also working on similar projects. Hopefully it's something that will entertain and enlighten listeners. The audio quality was so-so (my fault) but Rich at Voices of Wrestling did a wonderful job cleaning up my mess and turning this into something presentable.

Push Index Number (which we discuss in depth) is at: http://www.kayfabermetrics.com/835-2/

I didn't tell anyone about this project until I already recorded the first podcast. (I had to approach the fellas at VOW with hat-in-hand this morning basically like, "I recorded a podcast but I don't really have any sort of platform to do anything with it. Please sir, can you help?")

Honestly, I don't know whether it will be interesting to anyone else, but talking with Timmons was both therapeutic (like we say on the podcast, you can feel like you spend a lot of time in a vacuum) and educational (for me). I had seen the Kayfabermetrics Push Index Number (great name) before but I honestly didn't understand it so I really appreciate that he took the time to walk through what it all meant, and what was his basis of thinking while crafting some custom wrestling metrics. If nothing else, I hope others listen and can provide him with feedback on how to improve, use, abuse what he's done. I know I came out of the conversation with some ideas and anxious to apply the methods to some of the datasets that I've already assembled!

In the much greater context, my hope is that if I start these conversations with other people who are working on wrestling stats, perhaps we'll be able to figure out a way to come together and collaborate, or at the very least interact more, on some forum or place of choice.
Posted by at No comments:

Tuesday, February 11, 2014

Deciphering WCW Buyrates

One of the great advantages of WWE being a public company is that it's possible to decode their financial filings into concrete numbers regarding attendance, TV Rights fees and PPV buys. Alas, things are never quite as clear for poor WCW (World Championship Wrestling)


That isn't to say there isn't any information out there -- it's just scattered and sometimes contradictory.

Attendance
Dave Meltzer collects monthly stats on major wrestling federations which supplies us with WCW House Show Attendance 1991-2001:
TV Rights
Television Rights Fees were much less of "a thing" for pro wrestling back in the 1990s as compared to today.  Still, WCW was owned by a media network (Turner Broadcasting System, Inc.) and they were trying to grow the product base and make money through additional broadcasting.  They added a third hour to Nitro in 1999, launched Thunder, held internet-only audio broadcasts and created other specialized broadcasts such as WCW Live all which no doubt were designed to gain money through advertising and promoting the PPVs.  

Sadly, information from public filings only scantly mention World Championship Wrestling; there are random notes on ventures such as the Nitro Grill in Las Vegas, delicious Trimfast WCW Bars, nWo t-shirts, and WCW Action Figures.  WCW was rarely mentioned in Turner Public Filings -- usually just listed as one of the many, many assets of a major media and entertainment corporation.

PPV
Scattered across the web, there's a few places that have some WCW PPV numbers. But since it's about the mid-1990s, you're stuck in the world between buyrates (percentage of capable homes that purchased an event) and buys (the number of homes that bought the event).

Some of the data sources I have used:
The final column is CDH Estimate which is basically my formulaic guess at what might be plausible.


WCW EventDatePWH Buyrate GER Buyrate PWO buys CDH Estimate
Starrcade 1987
11/26/1987
3.30
20,000
20,000
Bunkhouse Stampede
1/24/1988
3.50
200,000
200,000
Great American Bash 1988
7/10/1988
2.20
190,000
190,000
Starrcade 1988
12/26/1988
1.80
100,000 to 150,000
150,000
Chi-Town Rumble
2/20/1989
1.50

130,000
WrestleWar 1989
5/7/1989
1.30

120,000
Great American Bash 1989
7/23/1989
1.50

140,000
Halloween Havoc 1989
10/28/1989
1.77

175,000
Starrcade 1989
12/13/1989
1.30

130,000
WrestleWar 1990
2/25/1990
1.60

175,000
Capital Combat
5/19/1990
1.40

160,000
Great American Bash 1990
7/7/1990
1.70

200,000
Halloween Havoc 1990
10/27/1990
1.30

160,000
Starrcade 1990
12/16/1990
1.30

165,000
WrestleWar 1991
2/24/1991
1.20
1.20

160,000
Superbrawl I
5/19/1991
1.04
1.04

150,000
Great American Bash 1991
7/14/1991
1.00

145,000
Halloween Havoc 1991
10/27/1991
0.80
0.80

120,000
Starrcade 1991
12/29/1991
1.00
0.70

155,000
Superbrawl II
2/29/1992
0.96
0.70

160,000
WrestleWar 1992
5/17/1992
0.61
0.60

105,000
Beach Blast 1992
6/20/1992
0.40
0.40

70,000
Great American Bash 1992
7/12/1992
0.40
0.40

70,000
Halloween Havoc 1992
10/25/1992
0.90

165,000
Starrcade 1992
12/28/1992
0.50
0.50
95,000
95,000
Superbrawl III
2/21/1993
0.50
0.50
85,000
95,000
Slamboree 1993
5/23/1993
0.50
110,000
100,000
Beach Blast 1993
7/18/1993
0.50
0.50
100,000
100,000
Fall Brawl 1993
9/19/1993
0.46
0.46
95,000
95,000
Halloween Havoc 1993
10/24/1993
0.50
0.50
100,000
100,000
BattleBowl
11/20/1993
0.27
0.27
60,000
55,000
Starrcade 1993
12/27/1993
0.55
0.55
120,000
115,000
Superbrawl IV
2/20/1994
0.50
0.50
110,000
110,000
Spring Stampede 1994
4/17/1994
0.53
0.50
125,000
115,000
Slamboree 1994
5/22/1994
0.48
100,000
105,000
Bash at the Beach 1994
7/17/1994
1.02
1.02
225,000
225,000
Fall Brawl 1994
9/18/1994
0.53
0.53
110,000
115,000
Halloween Havoc 1994
10/23/1994
0.97
0.97
220,000
210,000
Starrcade 1994
12/27/1994
0.60
0.60

130,000
Superbrawl V
2/19/1995
0.95
0.95
215,000
180,000
Uncensored 1995
3/19/1995
0.96
0.96
200,000
180,000
Slamboree 1995
5/21/1995
0.57
0.57
100,000
110,000
Great American Bash 1995
6/18/1995
0.51
0.51
90,000
100,000
Bash at the Beach 1995
7/16/1995
0.82
180,000
160,000
Fall Brawl 1995
9/17/1995
0.48
0.48
85,000
95,000
Halloween Havoc 1995
10/29/1995
0.60
0.60
120,000
120,000
World War 3 1995
11/26/1995
0.43
0.43

90,000
Starrcade 1995
12/27/1995
0.36
0.36

75,000
Superbrawl VI
2/11/1996
0.60
0.60

210,000
Uncensored 1996
3/24/1996
0.70
0.70

250,000
Slamboree 1996
5/19/1996
0.44
0.44

155,000
Great American Bash 1996
6/16/1996
0.48
0.48

170,000
Bash at the Beach 1996
7/7/1996
0.71

250,000
Hog Wild 1996
8/10/1996
0.62
0.62

220,000
Fall Brawl 1996
9/15/1996
0.65
0.65

230,000
Halloween Havoc 1996
10/27/1996
0.70
0.70

250,000
World War 3 1996
11/24/1996
0.55
0.55

200,000
Starrcade 1996
12/29/1996
0.95
0.95

345,000
Souled Out 1997
1/25/1997
0.47
0.47

170,000
Superbrawl VII
2/23/1997
0.75
0.75

275,000
Uncensored 1997
3/16/1997
0.89
0.89

325,000
Spring Stampede 1997
4/6/1997
0.58
0.58

210,000
Slamboree 1997
5/18/1997
0.60
0.60

220,000
Great American Bash 1997
6/15/1997
0.60
0.60

220,000
Bash at the Beach 1997
7/13/1997
0.89
0.89
500,000
325,000
Road Wild 1997
8/8/1997
0.65

240,000
Fall Brawl 1997
9/14/1997
0.53
0.53

195,000
Halloween Havoc 1997
10/26/1997
1.10
1.10

405,000
World War 3 1997
11/23/1997
0.56
0.56

205,000
Starrcade 1997
12/28/1997
1.90
1.80
625,000 to 650,000
700,000
Souled Out 1998
1/24/1998
1.02

380,000
Superbrawl VIII
2/22/1998
1.10
1.10

415,000
Uncensored 1998
3/15/1998
1.10
1.10

415,000
Spring Stampede 1998
4/19/1998
0.72
0.72

275,000
Slamboree 1998
5/17/1998
0.72
0.72

275,000
Great American Bash 1998
6/14/1998
0.75
0.80

290,000
Bash at the Beach 1998
7/12/1998
1.50
600,000
580,000
Road Wild 1998
8/8/1998
0.93
0.95

365,000
Fall Brawl 1998
9/13/1998
0.70
0.70

275,000
Halloween Havoc 1998
10/25/1998
0.78
0.78
500,000
310,000
World War 3 1998
11/22/1998
0.63
0.70

250,000
Starrcade 1998
12/27/1998
1.15
1.15

460,000
Souled Out 1999
1/17/1999
0.78
0.78

330,000
Superbrawl IX
2/21/1999
1.15
1.10

485,000
Uncensored 1999
3/14/1999
0.77
0.73

325,000
Spring Stampede 1999
4/11/1999
0.60
0.60

255,000
Slamboree 1999
5/9/1999
0.45
0.48

195,000
Great American Bash 1999
6/13/1999
0.43
0.43

185,000
Bash at the Beach 1999
7/11/1999
0.40
0.39

175,000
Road Wild 1999
8/14/1999
0.54
0.55

235,000
Fall Brawl 1999
9/12/1999
0.29
0.35

130,000
Halloween Havoc 1999
10/24/1999
0.52
0.52

230,000
Mayhem 1999
11/21/1999
0.45
0.45

200,000
Starrcade 1999
12/19/1999
0.32
0.23

145,000
Souled Out 2000
1/16/2000
0.26
0.25

115,000
Superbrawl X
2/20/2000
0.15
0.15

70,000
Uncensored 2000
3/19/2000
0.13
0.13

60,000
Spring Stampede 2000
4/16/2000
0.25
0.27

115,000
Slamboree 2000
5/7/2000
0.14
0.14

65,000
Great American Bash 2000
6/11/2000
0.19
0.20

85,000
Bash at the Beach 2000
7/9/2000
0.22
0.25

100,000
New Blood Rising
8/13/2000
0.18
0.18

85,000
Fall Brawl 2000
9/17/2000
0.16
0.16

75,000
Halloween Havoc 2000
10/29/2000
0.15
0.15

70,000
Mayhem 2000
11/26/2000
0.12

55,000
Starrcade 2000
12/17/2000
0.11

50,000
Sin
1/14/2001
0.17
0.10

80,000
Superbrawl Revenge
2/19/2001
0.15

70,000
Greed
3/18/2001
0.10

50,000

One thing that has been exceptionally hard to track down is "total number of PPV capable homes" so there is some basis to try and convert a 1.0 buyrate into something meaningful. You can read some of my notes on what I learned last time I tried to do that.

I am incredible grateful to the hard work of everyone who has labored to keep track of these archaic statistics for stat hounds like me. I don't believe that what I published here is correct, but I do think it's in the ballpark of accurate.  I look forward to continuing to learn more so I can refine techniques and learn more about what are the true WCW Buys vs Buyrates.
Posted by at 1 comment:

Sunday, February 09, 2014

A word about ROH and TNA Draws 2009-2013

I pulled Wrestling Observer stats June 2008-Jan 2014 from the results section so I could look at WWE Draws:

Switching over to ROH, the dataset is obviously far more limited (than WWE)

June 2008-Dec 2008: 18 shows (632 avg) and 3 other shows (no attendance listed)
Jan 2009-Dec 2009: 30 shows (740 avg) and 16 other shows (no attendance listed)
Jan 2010-Dec 2010: 20 shows (593 avg) and 15 other shows (no attendance listed)
Jan 2011-Dec 2011: 16 shows (528 avg) and 3 other shows (no attendance listed)
Jan 2012-Dec 2012: 16 shows (481 avg) and 2 other shows (no attendance listed)
Jan 2013-Dec 2013: 27 shows (497 avg) and 2 other shows (no attendance listed)
Jan 2014: 2 shows (900 avg)

So, from this perspective, they've been fairly flat since 2010, but you could argue by running more shows and actually increasing their average attendance over the prior year, they were actually up.

When you look at it in terms of total annual attendance (avg x # of shows)...

Total ROH Attendance
2009: 51 shows x 740 avg = 37,740
2010: 41 shows x 593 avg = 24,313
2011: 28 shows x 528 avg = 14,784
2012: 29 shows x 481 avg = 13,949
2013: 33 shows x 497 avg = 16,401

(To count ROH shows for this final calc - which you'll see is higher than the # of WO results -- I looked at the combination of The History of WWE and the WO Results. Note: some NJPW/PWG/1PW/AIW shows are in the ROH results, so I took those out of my count. I used the History of WWE count for 2009-2012 but their 2013 results are incomplete, so I used the cross-referenced the two sources and came up with 33 shows.)

For comparison, TNA average (using WO results) is a lot higher - more shows, and more people per show.

June 2008-Dec 2008: 36 shows (1188 average) and 35 other shows
Jan 2009-Dec 2009: 61 shows (1198 average) and 49 other shows
Jan 2010-Dec 2010: 67 shows (1284 average) and 51 other shows
Jan 2011-Dec 2011: 74 shows (1314 average) and 58 other shows
Jan 2012-Dec 2012: 58 shows (1064 average) and 69 other shows
Jan 2013-Dec 2013: 60 shows (1493 average) and 30 other shows
Jan 2014-Feb 2014: 13 shows (1723 average)

So, we're looking at at least twice as much per show, and twice as many shows.

Total TNA Attendance
2009: 131,769 (110 shows)
2010: 151,556 (118 shows)
2011: 173,416 (132 shows)
2012: 135,102 (127 shows)
2013: 134,325 (90 shows)

This was based completely on WO Results for TNA Shows and may be missing some PPVs.
Posted by at 2 comments:

Friday, February 07, 2014

House Show Main Eventers

As a follow-up to summarizing WWE House Show Tours (June 2008-Feb 2014), I was also looking at who wrestled in the main events on the WWE Live Events (f.k.a. House Shows).

Methodology:
  • Isolate the WWE Events that were not TV Tapings
  • Flag wrestlers that are in the last three matches per show at least 70% of their matches
  • Filter for wrestlers that average in the last quarter of the card for that month







Posted by at No comments:

WWE Live Events - Daniel Bryan versus John Cena

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:

  1. Shows with both John Cena and Daniel Bryan wrestling (BOTH) = 26 shows
  2. Shows with John Cena but not Daniel Bryan                  (CENA) = 31 shows
  3. Shows with Daniel Bryan but no John Cena                  (BRYAN) = 39 shows
  4. 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

TypeSHOWSDifference
vs Baseline Avg 		Difference vs
Baseline Median		Avg
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:

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
Posted by at 1 comment:
Subscribe to: Posts (Atom)