What I learned from Meltzer's 280+ WWF PPVs Ratings (Part One)

What I learned from Metlzer's 280+ WWF PPVs Ratings

by Chris Harrington (chris.harrington@gmail.com)

PREAMBLE AND RAMBLE
Last week, I posted a Five Year summary of Wrestling Observer Star Ratings from 2008-2013 covering WWE, TNA and other federations (Dragon Gate/Dragon Gate USA, New Japan Pro Wrestling, Ring of Honor, CMLL and so forth).  I was pretty proud of myself for scraping all of that data out of WO issues available online and extracting the interesting bits.  However, I was quickly humbly (thankfully not old-country way) when I received an email from a mysterious (and benevolent) benefactor with all Wrestling Observer Star ratings from 1985-2013 for WWF/WWE PPVs.

This isn't the first time I dove into the wonderful world of snowflakes aka WO Star Ratings.  A few months ago (right before I got caught up in WWE Network hysteria), I published a few pieces looking at Star Ratings including:

• Systematic Analysis of Meltzer's ***+ Ratings for WWE (1986-2013)
• Wrestlers Age vs Star Ratings: Who has the brightest future?

These two pieces were built on the backs of two sites (WWE Observer Star List 1986-Present, compiled by Michael Stamcoff and The History of WWE, administrated by Graham Cawthon).  By contrast, the latest work is the result of information supplied by Alex Sarti and match data extracted from Cagematch. (Since many of these Observers are not available online, using so many different points should ideally give us an opportunity to triangulate and validate the data.)

DATASET AND QUESTIONS

Wrestling Observer Star ratings for WWF/WWE PPVs from 1985 to 2013

Covers more than 2,000 matches and more than 280 PPV events

I leave the match/PPV counts somewhat vague because there's a few events that have incomplete coverage (NYR 05, IYH 18), and there are some matches which are only rated as either 0 (DUD) or N/A which I'd like to verify that there was no rating given.

Regardless of any small blemishes, this information provides a very rich and complete dataset covering nearly three decades of pay-per-view.  (Supplementing the early years with SNME ratings will nicely fill out the early years for a more complete picture.)

With the debut of the WWE Network looming with all past WWF/WWE, WCW and ECW PPVs immediately available for on-demand viewing, this archive of results provides an important gateway to tackle questions such as:

1. What were some of the best PPVs and best years in WWF/WWE history?  How should we weight the "importance" of various matches on the card against their performance?
2. Who are some of the best PPV performers in WWF/WWE history?  How did they evolve over time by age and by opponent?
3. What are some of the most predictable trends for PPV quality in WWF/WWE history?  

And lastly...

     4. What are the characteristics that define the match ratings?  (For example, what characteristics seem to be common to a **** match - competitor age/history/card placement/length/etc.?)

ANALYSIS

PART ONE: Simply the Best

We'll start with one of those questions that seems deviously simply: "What year had the best for WWF/WWE PPVs?"

Clearly, as with any query involving subjective ratings, we're not going to be able to "prove" anything (yet, I'm still going to try.) Please note that for the purposes of this analysis, I'm going to use Dave's WO ratings as a starting point for defining the "best".  (This is certainly not without controversy, but one of the primary reasons that I use Meltzer's snowflakes is the consistency it delivers: the same person watching the events as they happened and rating them in the context of that time.  Inevitably, as a critic he's prone to biases which are both fair and unfairly driven.  But as the leading Pro-wrestling journalist, his voice is a powerful and resounding one worth considering.)

As you start poking around, you realize that there's more than one way that you can synthesize the individual match ratings into event ratings into annual ratings.

The simplest is the unweighted (straight-line) average.  Every match that has a WO rating is assigned a numerical value.  We add up all of the values and divide by number of matches.

Figure 1: Unweighted Average Star Rating by Year for WWF/WWE PPVs

year	unwgt avg	matches		year	unwgt avg	matches		year	unwgt avg	matches
1985	1.63	23		1995	1.98	65		2005	2.21	104
1986	1.49	23		1996	2.12	67		2006	2.08	115
1987	2.00	16		1997	1.75	75		2007	2.11	113
1988	1.33	34		1998	1.61	100		2008	2.43	95
1989	1.23	32		1999	1.44	96		2009	2.59	94
1990	1.49	34		2000	2.05	100		2010	2.36	91
1991	2.01	38		2001	2.50	99		2011	2.60	89
1992	1.43	30		2002	2.30	97		2012	2.34	90
1993	1.95	37		2003	2.21	91		2013	2.53	99
1994	1.90	36		2004	1.97	107

Figure  2: Star Rating Distribution by Year for WWF/WWE PPVs

Initial Observations:

• The move to monthly PPVs (May 1995) greatly inflated the number of PPV matches per year.  (This is another reason why it would make sense to add in SNME results to the 80's numbers.)
• The number of matches expanded again in the attitude Era (circa '98-'02) as PPVs were longer and some had single-night multi-match tournaments.
• The percentage of negative stars as a proportion of total matches has generally been falling each decade.  Meanwhile, the percentage of matches rated above 2.5 stars has grown in that timeframe.

This last point is particularly important when you're evaluating the unweighted annual PPV averages.

Q: How should we deal with negative star ratings? 
A: Let's consider when the Mega-Powers exploded..

April 2, 1989: WRESTLEMANIA V (results courtesy of Cagematch)

• Hercules defeats King Haku (w/Bobby Heenan) (6:57) = 1/2*
• The Twin Towers (Akeem & The Big Boss Man) (w/Slick) defeat The Rockers (Marty Jannetty & Shawn Michaels) (8:02) = *3/4
• Brutus Beefcake vs. Ted DiBiase (w/Virgil) - Double Count Out (10:01) = *3/4
• The Bushwhackers (Butch & Luke) defeat The Fabulous Rougeaus (Jacques Rougeau & Raymond Rougeau) (9:10) = -****
• Mr. Perfect defeats The Blue Blazer (5:38) = **1/4
• WWF World Tag Team Title Three On Two Handicap: Demolition (Ax & Smash) (c) defeat Mr. Fuji & The Powers Of Pain (The Barbarian & The Warlord) (8:20) = DUD
• Dino Bravo (w/Frenchy Martin) defeats Ronnie Garvin (3:06) = DUD
• The Brain Busters (Arn Anderson & Tully Blanchard) (w/Bobby Heenan) defeat Strike Force (Rick Martel & Tito Santana) (9:17) = **1/2
• Singles Match (Special Referee: Big John Studd): Jake Roberts defeats Andre The Giant (w/Bobby Heenan) by DQ (9:44) = -***
• The Hart Foundation (Bret Hart & Jim Neidhart) defeat Greg Valentine & The Honky Tonk Man (w/Jimmy Hart) (7:40) = **1/4
• WWF Intercontinental Title: Rick Rude (w/Bobby Heenan) defeats The Ultimate Warrior (c) (9:36) = **1/2
• Bad News Brown vs. Jim Duggan - Double DQ (3:49) = DUD
• The Red Rooster defeats Bobby Heenan (w/The Brooklyn Brawler) (0:32) = DUD
• WWF World Heavyweight Title: Hulk Hogan defeats Randy Savage (c) (17:54) = **3/4

14 matches. Ratings were all over the map: two received negative stars, four were DUDs (zero stars), five were positive but below two-and-half stars and three were at at/above two-half stars.  

The straight-line average for WM5 would be 0.66. That's pretty abysmal for a PPV, let alone the major pay-per-view event of the year! 

However, does it really make sense that a four star classic can be completely cancelled by a negative four star travesty?  Furthermore, when your negative matches are on the undercard and your positive matches are the main events, shouldn't that matter?  How should we weight factors such as card importance, length of match, placement of match/star power in match, etc?  How should we deal with shows that do not have ratings for all matches?  Should a "DUD" rating count as zero stars, or just be ignored?  Likewise, how should we deal with the negative ratings (which were more prevalent as a total percentage of matches in WWF 1980s)?

Alternatively, we could look at just matches that at least had a positive star rating.  Now, eight matches at Trump Plaza qualify and our unweighted positive star average for WM5 is 2.03 which may be low, but is at least a passable score for major PPV.

Figure 3: Unweighted Positive Star Ratings (removes DUD and below)

year	pos avg	matches		year	pos avg	matches		year	pos avg	matches
1985	2.47	18		1995	2.22	60		2005	2.55	91
1986	2.45	16		1996	2.49	59		2006	2.50	99
1987	1.81	15		1997	2.33	59		2007	2.46	100
1988	1.74	27		1998	1.97	87		2008	2.56	90
1989	2.02	24		1999	1.91	81		2009	2.83	87
1990	2.02	28		2000	2.34	90		2010	2.45	87
1991	2.35	33		2001	2.72	93		2011	2.65	87
1992	2.06	24		2002	2.51	90		2012	2.51	85
1993	2.32	32		2003	2.34	86		2013	2.56	97
1994	2.24	32		2004	2.28	96

Illustrated, when we plot Table 1 (unweight star ratings by year) versus Table 3 (positive star ratings by year), we see that while both methods yield generally similar resultst, what tends to happen is that delta between the two lines has continued to diminish over time (as would be suggested by table 2).

Figure 4: Unweighted Star Ratings (all) vs Unweighted Star Ratings (positive)

The three years that are most severely adversely affected by negative star ratings are 1985 (-0.84), 1986 (-0.96) and 1989 (-0.79). Both 1985 and 1986 move from bottom of the pack (22nd and 25th respectively) to middle of the pack (12th and 14th respectively) when you compare their rankings under the unweighted all matches methodology (red) and unweighted positive star ratings only methodology (green).  (On the flipside, 2003 takes a tumble from 9th place to 18th place when you exclude negative ratings; this suggests that there's less great matches that year at the very top as compared to other years.)

Interestingly, we have the same top five highest rated years under either methodology:

Top Five Years for WWF/WWE PPVs by Unweighted Star Ratings

1. 2009: 2.59 unweighted (all matches), 2.84 unweighted (positive star ratings)
2. 2011: 2.60 unweighted (all matches), 2.66 unweighted (positive star ratings)
3. 2001: 2.50 unweighted (all matches), 2.72 unweighted (positive star ratings)
4. 2013: 2.53 unweighted (all matches), 2.58 unweighted (positive star ratings)
5. 2008: 2.43 unweighted (all matches), 2.58 unweighted (positive star ratings)

There has been a remarkable cluster of years (2008-2013) with excellent per capita match quality on WWE PPVs as compared to the vast history of the product.  (2010 and 2012 are at 6th/7th with 2.34/2.36 unweighted all match averages).  This result is even more amazing when you consider that during this era there's been an explosion of TV Rights Fees for WWE while PPV revenues have stagnant and dropping. (With the recent huffing and puffing about the death of PPV quality due to the WWE Network, three-hour Raws and meteoric rise of future TV Rights, might this represent a counterargument that "PPV" quality isn't going to necessarily plummet?)

Bottom Five Years for WWF/WWE PPVs by Unweighted Star Ratings

1. 1988: 1.33 unweighted (all matches), 1.81 unweighted (positive star ratings)
2. 1999: 1.44 unweighted (all matches), 1.91 unweighted (positive star ratings)
3. 1989: 1.23 unweighted (all matches), 2.02 unweighted (positive star ratings)
4. 1992: 1.43 unweighted (all mathces), 2.06 unweighted (positive star ratings)
5. 1998: 1.61 unweighted (all matches), 1.97 unweighted (positive star ratings)

(Honorable mention 1990 with 1.49/2.02 scores.)

While it's not surprising to see late-1980s WWF poorly received by Dave Meltzer (especially in light of the stellar work in WCW at that time).  Similarly, 1990 and 1992 remind me of the malaise that emerged in the early 1990s in American professional wrestling (as been noted in other places, such as House Show attendance.)

But listing 1999 on a list of worst years of WWF/WWE quality may surprise some especially since that year was in the heart of boom period for WWF and the middle of the Monday Night War(s).  This datapoint brings us to the next question, should other factors - such as card position - play into how we evaluate the totality of a PPV?

For instance, the positive star rating average for 1999 PPVs (which excludes the tragic Over the Edge 1999 event) is a less than impressive 1.91.  However, if you split the matches into those on the first half of the card (36 of 81 matches with positive ratings, 1.49 positive star average) and the second half of the card (45 of 81 matches with positive ratings, 2.24 positive star average), there is a distinct quality difference.  (For this purpose I am using the actual card order as it aired on the PPV. This isn't factoring in the "strength" of the matches in terms of importance.)

Figure 5: 1985-2013 PPV Positive Star Ratings Unweighted Average (first half vs second half of Card)

What does this Radar chart show (besides that I am currently fooling around with Excel in a vain attempt to spice things up)?  The green line represents the second half of the card- where the "main event" who traditionally occur.  The maroon line represents the first half of the card (often including pre-show when that was rated) which would traditionally include the undercard.

Where the lines cross (1991, 1994) there was a distinct dynamic where the later matches had lower average ratings than the earlier matches.  When the lines meet (1993, 2010) there was essentially parity the matches scattered across the card.

In 1991, there was only three WWF PPV matches that were rated ****+:

• Rockers/Orient Express (Royal Rumble 91) ****
• Ultimate Warrior vs Randy Savage (Wrestlemania) ****1/4
• Bret Hart vs Mr Perfect (SummerSlam 91) ****

All three of these matches were considered "first half" matches.  (Even the Warrior/Savage match because there was so many matches on that card.)  Conversely, the three matches that cracked *** in the "top" half of 1991 PPV cards (Beverly/Nasties vs Bushwhackers/Rockers SS and Repo/DiBiase vs El Matador/Virgil Tues in TX) were 3.5.

1994 is more stunning when you consider there were TWO five-star matches (Bret/Owen Cage and HBK/Razor Ladder) which both were in the second half of the card.  And YET, dismal matches involving Mabel, Razor/IRS and Yokozuna vs Lex Luger were rough enough to be a giant albatross.  (This also says something about the quirks of looking at unweighted numbers.)

So, should we weight the main events over the undercard?  And if so, by what system?  Also, are there other weightings we could use - based on time, importance of championships, OCELOT values of wrestlers involved, number of wrestlers involved, etc?