For years, I’ve harnessed the “Observer Star Ratings” blog to look at Star Accumulations by Star or Who has the Brightest Future (Age vs Star Ratings).  While I had the information of who was in the matches, I lacked other important data points such as length of the match, type of show or finish of match.

Through sheer will of force and extensively cross-referencing with The History of WWE I’ve mapped each of the matches to collect these data points.

I’ve been interested in whether or not it’s possible to qualify the qualities of what distinguishes a good *** match from a great **** match.  By examining the dataset in full, can we draw any conclusions about which factors seem to influence how Dave Meltzer rated the matches?

Initial Findings

The first thing that immediately jumps out is the relationship between star ratings and average match length.

Expressed graphically, the trend is even clearer:

It’s important to pause here and note some things:

Long does not necessarily equal good.

The initial dataset was composed only of matches that were rated three stars or higher. Any findings here only apply to comparing similarly situated (three stars or higher) WWE matches.  That is to say, there are plenty of long WWE matches which were not rated even three stars.  Simply being long doesn’t equate with being good.  However, among good matches, there is certainly a relationship between being longer (or at least of a certain length) and being rated highly.

Generalizations Versus Individual Examples

When you look at a scatter plot of match length versus Star Ratings (WWE), you quickly get the sense that while the general average is a relationship between longer matches and higher star ratings but you’ll note that there are plenty of exceptions on either side of the trendline.

For instance, a fifteen minute match could finish almost anywhere:

  • 1/23/95 RAW: Smoking Gunns vs. 123 Kid/Bob Holly [15:32] = 3.00 stars
  • 5/18/03 PPV: Brock Lesnar vs. Big Show (Stretcher) [15:28] = 3.25 stars
  • 7/28/09 ECW: Christian vs. Zach Ryder [15:28] = 3.50 stars
  • 7/25/99 PPV: Steve Austin vs. Undertaker (First Blood) [15:31] = 3.75 stars
  • 7/27/03 PPV: Shelton Benjamin/Charlie Haas vs. Rey Mysterio/Billy Kidman [15:01] = 4.00 stars
  • 6/10/13 RAW: Daniel Bryan vs. Seth Rollins [15:39] = 4.25 stars
  • 5/19/02 PPV: Edge vs. Kurt Angle (Hair vs. Hair) [15:28] = 4.50 stars
  • 4/1/01 PPV: Edge/Christian vs. Matt Hardy/Jeff Hardy vs. Bubba Ray Dudley/D-Von Dudley (Tables, Ladders & Chairs) [15:41] = 4.75 stars

Length alone is clearly not the only factor.  Simply knowing the length of the match does not give us enough information to accurately predict whether it’s a good, great or spectacular match.

98% of Observations are below 4.5 stars

Even with over 25 years of history and a thousand datapoints, there is only a tiny portion of 4.75 and 5 stars matches from WWE to look at.  I feel that speaks volumes to the intangibles (such as crowd reaction, storyline build, performer charisma, match execution, match timing and psychology) which just aren’t going to be captured in the analysis that I’m performing.

Other Dataset Statistics


  • PPV: 695 matches (all five ***** matches were on PPV)
  • RAW: 137 matches
  • SMACKDOWN: 124 matches
  • ECW: 26 matches
  • REMAINING: 20 matches (includes televised MSG shows Action Zone, Velocity, Wrestling Challenge, Prime Time Wrestling, PPV pre-show, WWE Main Event, and USA specials such as “Road to Wrestlemania”)

I think it’s especially interesting to see that your average PPV match is at 17:13, while your network television special is only at 9:46. RAW/Smackdown bridging the gap (15:10 to 16:25 minutes).


  • New York City, NY: 52 matches (3.67 star avg with 17 ****+ matches)
  • Chicago, IL: 46 matches (3.66 star avg with 15 ****+ matches)
  • Los Angeles, CA: 35 matches (3.62 star avg with 9 ****+ matches)
  • Baltimore, MD: 29 matches (3.53 star avg with 4 ****+ matches)
  • Pittsburgh, PA: 26 matches (3.43 star avg with 5 ****+ matches)
  • Philadelphia, PA: 25 matches (3.49 star avg with 4 ****+ matches)
  • Houston, TX: 25 matches (3.64 star avg with 6 ****+ matches)
  • Boston, MA: 24 matches (3.47 star avg with 4 ****+ matches)
  • Detroit, MI: 23 matches (3.47 star avg with 5 ****+ matches)
  • East Rutherford, NJ: 21 matches (3.38 star avg with 4 ****+ matches)
  • Phoenix, AZ: 20 matches (3.64 star avg with 7 ****+ matches)
  • Long Island, NY: 19 matches (3.46 star avg with 3 ****+ matches)
  • St. Louis, MO: 19 matches (3.57 star avg with 4 ****+ matches)
  • Washington, DC: 16 matches (3.55 star avg with 3 ****+ matches)
  • Providence, RI: 16 matches (3.55 star avg with 3 ****+ matches)
  • Indianapolis, IN: 16 matches (3.48 star avg with 3 ****+ matches)
  • Atlanta, GA: 16 matches (3.53 star avg with 4 ****+ matches)
  • San Jose, CA: 15 matches (3.72 star avg with 5 ****+ matches)
  • Cleveland, OH: 15 matches (3.65 star avg with 6 ****+ matches)
  • Charlotte, NC: 14 matches (3.34 star avg with no ****+ matches)
  • London, England: 14 matches (3.41 star avg with 3 ****+ matches)
  • San Antonio, TX: 13 matches (3.6 star avg with 5 ****+ matches)
  • Dallas, TX: 13 matches (3.88 star avg with 6 ****+ matches)
  • Seattle, WA: 13 matches (3.69 star avg with 6 ****+ matches)
  • Miami, FL: 13 matches (3.65 star avg with 3 ****+ matches)
  • New Orleans, LA: 13 matches (3.81 star avg with 5 ****+ matches)
  • Raleigh, NC: 12 matches (3.71 star avg with 4 ****+ matches)
  • Kansas City, MO: 12 matches (3.54 star avg with 4 ****+ matches)
  • Milwaukee, WI: 12 matches (3.54 star avg with 2 ****+ matches)
  • Toronto, Ontario: 12 matches (3.46 star avg with 3 ****+ matches)
  • San Diego, CA: 11 matches (3.66 star avg with 4 ****+ matches)
  • Tampa, FL: 11 matches (3.52 star avg with 3 ****+ matches)
  • Buffalo, NY: 10 matches (3.48 star avg with 1 ****+ matches)
  • Las Vegas, NV: 10 matches (4 star avg with 6 ****+ matches)
  • Hartford, CT: 10 matches (3.55 star avg with 3 ****+ matches)
  • Anaheim, CA: 10 matches (3.68 star avg with 3 ****+ matches)

I think it’s interesting to see the top three cities (essentially Wrestlemania II cities – though I have broke out Long Island from NYC in this example) all hovering at the same 3.6 average.

Regression Analysis

Lastly, having a deeper understanding of the dataset gives us the opportunity to quantify a bunch of variables and see how those factors seem to influence the star rating.

Relevant Variables

  • Length of Match in minutes
  • Was it on broadcast on Monday Night Raw/Smackdown?
  • Is it not a singles match?
  • Was a ladder used during the match? (i.e. TLC, Ladder match, Extreme Rules/No Holds Barred match)

FORMULA = 3.09 (INTERCEPT) + (2.77% of Match Time) + (0.11 x RAW/SM) + (-0.11 x NOT_SINGLES) + (0.36 x LADDER)  
(R-squared for this formula is 0.507 but adjusted R-squared is down 0.25 which isn’t great.)


It’s interesting to see that Singles matches (multi-man singles matches were considered singles, handicap matches were considered tag) are more highly related than tag matches and that your best bet for getting a highly rated match (in general) was on RAW and SMACKDOWN (individually they had almost the same intercept so I combined them) rather than on PPV. The bump for LADDER match is pretty hefty – no doubt owing to the fact the 74 matches that I noted in the dataset had 4.0 average rating and MITB mutli-man scrambles have been pretty impressive spectacles.

Other Variables Examined But Not Used

Interestingly, there was not a conclusive relationship with TITLE CHANGES (higher p-value over 0.12); intercept on TITLES CHANGE was 0.05 so it had a very nominal positive effect. Cage matches (HITC, Elimination Chamber, Steel Cage) was also at about 0.05 intercept but an incredibly high p-value. Being on PPV was actually associated with a negative tenth of a point but a higher p-value.

Isolating Wresters

I experimented with isolating wrestlers that had strong performances in the tallies – namely Flair, HBK, Angle, Foley, Punk, Edge, HHH, Jericho, Rey, Eddie and Benoit.

Ric Flair, CM Punk, Rey Mysterio Jr and Eddie Guerrero all had high p-values (far above 0.25) and low intercepts which suggested there wasn’t an effect that could be isolated in this regression.

Of the remaining crew, the impacts were: Foley (+0.19), Angle (+0.19), Benoit (+0.16), HBK (+0.16), Edge (+0.09), Jericho (+0.07) and HHH (-0.06, high p-value at 0.11).

Isolating wrestlers does diminish the effect of RAW/SM from 0.11 to 0.06, so it suggests some of it may be just a function of which wrestlers often appeared (such as the prime period of Smackdown).

See the current North American Star Tally and the full WWF dataset (order by match length) on my statistics website

You can read all of my statistical work on and follow me on Twitter @mookieghana. Lastly, don’t forget to purchase my upcoming book(s) “Wrestlenomics” by emailing me at for details on pricing and delivery options.