Who is the Biggest UFC PPV Draw? – Part 2

August 18, 2010

Welcome to the second part of our series on which fighters are the biggest UFC PPV draws.  In part one, we mentioned that using aggregate information can be an advantage or a disadvantage.  In this installment, we’ll break down the overall and fighter-specific average PPV buyrates by year from 2006 to 2010.

Overview: One of the common debates among MMA fans concerns which fighter is the biggest draw.  From the business side, a fighter’s draw shapes PPV expectations (and, subsequently, PPV revenues) and should play a major role in sponsorships for both the given fighter and every other fighter on the PPV.  In this series of articles, we’ll examine several intuitive ways that one can estimate a fighter’s draw and examine the wide variation in these estimates.  We’ll be focusing on the ten fighters that Derek Jenkins identified as the biggest draws in a recent article at Yahoo! Sports.

Today’s Comparison: Average PPV buys versus average PPV buys for cards with a fighter broken down by year

The Number

Average PPV buys are the most basic numbers we can use to determine a fighter’s draw.  Today, we’ll take the average number of buys for all UFC PPVs (including events with all ten fighters) and compare it with the average number of buys for UFC PPVs featuring each fighter.  We used the same approach in part one, but we’ll break down those numbers by year and look for trends that were hidden in the aggregate numbers we looked at before.

The Fighters

As mentioned in the overview, we’ll be using the ten fighters that Derek Jenkins identified as the biggest draws: Brock Lesnar, Georges St. Pierre, Chuck Liddell, Rashad Evans, Quinton Jackson, Forrest Griffin, Lyoto Machida, BJ Penn, Anderson Silva, and Randy Couture.

The Data

We’ll still be using the PPV buys for all UFC pay-per-view events from UFC 57 (Couture-Liddell III) to UFC 116 (Lesnar-Carwin), and we will continue to do so throughout the series.  You can view the PPV buys for events up to UFC 107 in our MMAPayout.com Blue Book.

2006 –  Average number of PPV buys:  527,000

Fighter Average PPV buys for cards featuring fighter Difference from overall average
Brock Lesnar
Georges St. Pierre 400,000 -127,000
Chuck Liddell 650,000 123,000
Rashad Evans 400,000 -127,000
Quinton Jackson
Forrest Griffin 658,000 131,000
Lyoto Machida
BJ Penn 350,000 -177,000
Anderson Silva 300,000 -227,000
Randy Couture 400,000 -127,000

2007 –  Average number of PPV buys:  449,000

Fighter Average PPV buys for cards featuring fighter Difference from overall average
Brock Lesnar
Georges St. Pierre 523,000 74,000
Chuck Liddell 600,000 151,000
Rashad Evans 375,000 -74,000
Quinton Jackson 538,000 89,000
Forrest Griffin 338,000 -111,000
Lyoto Machida 508,000 59,000
BJ Penn
Anderson Silva 383,000 -66,000
Randy Couture 530,000 81,000

2008 –  Average number of PPV buys:  527,000

Fighter Average PPV buys for cards featuring fighter Difference from overall average
Brock Lesnar 745,000 218,000
Georges St. Pierre 578,000 51,000
Chuck Liddell 480,000 -47,000
Rashad Evans 740,000 213,000
Quinton Jackson 770,000 243,000
Forrest Griffin 770,000 243,000
Lyoto Machida 475,000 -52,000
BJ Penn 350,000 -177,000
Anderson Silva 313,000 -214,000
Randy Couture 1,010,000 483,000

2009 –  Average number of PPV buys:  617,000

Fighter Average PPV buys for cards featuring fighter Difference from overall average
Brock Lesnar 1,600,000 983,000
Georges St. Pierre 1,260,000 643,000
Chuck Liddell 650,000 33,000
Rashad Evans 635,000 18,000
Quinton Jackson 350,000 -267,000
Forrest Griffin 613,000 -4,000
Lyoto Machida 685,000 68,000
BJ Penn 797,000 180,000
Anderson Silva 750,000 133,000
Randy Couture 435,000 -182,000

2010 –  Average number of PPV buys:  600,000

Fighter Average PPV buys for cards featuring fighter Difference from overall average
Brock Lesnar 1,200,000 600,000
Georges St. Pierre 770,000 170,000
Chuck Liddell 520,000 -80,000
Rashad Evans 675,000 75,000
Quinton Jackson 1,050,000 450,000
Forrest Griffin
Lyoto Machida 520,000 -80,000
BJ Penn 525,000 -75,000
Anderson Silva 525,000 -75,000
Randy Couture 275,000 -325,000

Payout Perspective

In part one, we addressed the strengths and weaknesses of the numbers we’re using today.  Breaking down average buyrates allows us to address the issues raised by aggregating all the data into a single average for each fighter and overall average.  Since the general issues still remain in the numbers we’re using today, we’ll ignore those that don’t deal with changes over time.

Average buyrate by year fluctuates, but PPV buys are not cyclical. The average buyrate in 2006 was 527,000, which dropped to 449,000 in 2007.  We then see 2008 reach the 2006 levels, with additional growth in 2009, followed by a dip so far in 2010.  It’s easy to assume from these numbers that MMA follows a cycle where buyrates are up one year, then down the next, and then pick back up in the third year.  However, there’s not really a cycle if you look at the event buys in our MMAPayout.com Blue Book.

In part one, we discussed how a single fighter’s numbers can drive the overall average (eg. Brock Lesnar).  The same thing is happening in the annual numbers, where the buyrates for individual events generally hover around 300,000-500,000.  In the boom years, however, the annual average is being driven by a small number of events that get significantly higher buyrates.  The difference between 2006 and 2007 is mostly due to UFC 61 (Ortiz-Shamrock II with 775,000 buys) and UFC 66 (Liddell-Ortiz II with 1,050,000 buys).  Aside from those two events, the 2006 and 2007 event buyrates look similar.

We then see the jump in 2008 due to back-to-back events hitting the million buyrate mark in UFC 91 (Lesnar-Couture) and UFC 92 (Griffin-Evans).

The numbers are down for 2010 thus far but may look much better at the end of the year.  The year started off with three events in a row that maxed out at 300,000 buys with UFC 108 (Evans-T. Silva), with 109 (Couture-Coleman) and 110 (Nogueira-Velasquez) doing 275K and 240K, respectively.  Since then, two events have broken one million buys, UFC 114 (1,050,000 buys headlined by Evans-Jackson) and UFC 116 (1,200,000 buys headlined by Lesnar-Carwin).  One or two more big events in 2010 will likely push the number up to or above the 2009 mark, and GSP-Koscheck at UFC 124 should be one of these events.[i] Online rumors are placing UFC 117 (Silva-Sonnen) at around one million buys, which would push the 2010 average above the current average of 600,000 (since the 2010 numbers above only include PPVs through UFC 116).

Fighters look like terrible draws in 2006. Only Chuck Liddell and Forrest Griffin perform above average among the seven fighters who were active in the UFC in 2006; Lesnar, Evans, and Machida did not fight for the UFC during the year.  The other five active UFC fighters – St. Pierre, Evans, Penn, Silva, and Couture – did at least 100,000 buys below average.  As mentioned previously, it’s important to realize that the 2006 average was strongly influenced by two cards, UFC 61 (Ortiz-Shamrock II) which featured none of these fighters and UFC 66 (Liddell-Ortiz II, with Griffin-Jardine also on the card).  The massive buyrate for UFC 66 drives up the overall average and the average for both Liddell and Griffin.  The other five active fighters all drew 300,000-400,000 buys, which was a normal draw for the UFC with the few notable exceptions we’ve mentioned along with Hughes-Gracie at UFC 60, which featured none of the ten fighters we’re looking at.

You can look and see a similar trend in the 2010 numbers, as most fighters are performing below average, even those in the 500,000-600,000 range.  As was the case in 2006, this is primarily due to two events, with UFC 114 and UFC 116 breaking one million buys.

Single years can give very different impressions about the same fighter. Consider the case of Randy Couture.  He has the fifth highest average among all years with an average of 1,010,000 in 2008, behind only Lesnar twice (2009 and 2010), St. Pierre (2009), and Jackson (2010).  However, he also has the lowest average of all, with 275,000 in 2010.  Couture also only fought once in each year.  His sole fight in 2008 was against Lesnar, and his only fight in 2010 was against Mark Coleman.  In both cases, the choice of opponent likely played a big role in the buyrate.  Also, as mentioned in part one, having only one fight in a year means your average annual buyrate is equal to the buyrate for the one card you fought on.

Another good example of this phenomenon is Quinton “Rampage” Jackson.  He went from an average of 350,000 in 2009 to 1,050,000 in 2010, meaning that his average buyrate tripled in a single year.  We can state this change in a different way – Jackson’s buyrate grew 200% in one year, which is amazing by any standard.[ii] Like Couture, Jackson only fought once in each year, against Jardine in 2009 and Evans in 2010.  The build-up of the Evans-Jackson fight via a combination of a season of The Ultimate Fighter and Jackson’s mainstream appearance in a lead role of The A-Team contributed significantly to this growth, and Jackson may not hit his 2010 number again as a result.

Did the 2006-2007 numbers penalize fighters from a less popular time? In part one, I made the following comment:

Using numbers from 2006 and 2007 may penalize fighters – like Liddell and Couture – who were headlining PPV events when the sport was less popular.

We can now look at the numbers year-by-year and see whether including the 2006-2007 numbers really did make these fighters look less popular.  Let’s put Liddell and Couture’s numbers beside each other year-by-year:

Chuck Liddell / Randy Couture

2006       650,000 / 400,000

2007       600,000 / 530,000

2008       480,000 / 1,010,000

2009       650,000 / 435,000

2010       520,000 / 275,000

Except for Couture’s single 2008 fight against Brock Lesnar and his single 2010 fight against Mark Coleman, the numbers look fairly consistent over time.  In both cases, keeping the 2006 and 2007 numbers actually help the fighter’s overall average buyrate.  While we can’t be certain why the 2010 numbers for both fighters are down, we can reasonably speculate that it has to do with both fighters being in the twilight of their careers and ending up in fights which are interesting to fans but are unlikely to have any relevance to the title.

QUICK SUMMARY – SO FAR

In the first two parts of this series, we’ve compared the average buyrate for a fighter’s pay-per-view appearances with the overall average for UFC PPVs.  We looked at the aggregate numbers in part one and broke those numbers down by year here in part two.  Breaking down the numbers by year showed us that the growth in buyrates has not been consistent year after year and, more importantly, the averages – annual or aggregate – tend to be driven by single events with buyrates substantially higher than the average.

The primary benefit to using the aggregate numbers from part one is that we don’t see the wild variation due to one great or lackluster card – as we saw here with Couture and Jackson.  The aggregate numbers still give a good indication of draw, even with increased popularity of the sport, since the fluctuation in numbers is relatively consistent over the years.  The annual numbers do provide additional information, but one has to break those down even more and look at the buyrates for each event to fully understand the sources of variation.

UP NEXT

Going back to part one, we mentioned the following weakness of comparing fighters to the overall average buyrate (which includes all ten of the fighters we’re looking at):

Makes good draws look worse and bad draws look better. Consider Brock Lesnar, whose worst PPV did 600,000 buys.  His five PPVs in the dataset sold 600K, 625K, 1.01 million, 1.6 million, and 1.2 million.  These PPVs drive up the overall UFC average, which is what Lesnar is being compared to here.  A better comparison for Lesnar is to compare his average PPV buyrate to the average buyrate for all UFC PPVs that do not feature Lesnar.

I added emphasis to the last sentence because that’s exactly what we’ll do in parts three and four.  In part three, we’ll again look at the aggregate numbers, as we did in part one.  We’ll then break those down by year in part four, as we did here in part two.


[i]It should be noted that UFC 120, headlined by Michael Bisping versus Yoshihiro Akiyama in London, will not be a pay-per-view event.  The card would likely draw low PPV numbers and will instead be broadcast via tape delay on Spike TV.

[ii] To clarify any potential confusion, Jackson’s 2010 buyrate is 300% of his 2009 buyrate, but it only represents a 200% growth rate.  Growth rate is equal to (2010 average – 2009 average)/(2009 average).

7 Responses to “Who is the Biggest UFC PPV Draw? – Part 2”

  1. Diego on August 19th, 2010 6:15 AM

    Nice analysis. It’s getting more and more interesting.

  2. shawn on August 19th, 2010 12:09 PM

    Yes I agree they sholud do part 5 in nov. And 6 at the end of the year

  3. Big Daddy on August 19th, 2010 11:12 PM

    How many people do you think will buy UFC 118?
    Thanks!

  4. Kelsey Philpott on August 20th, 2010 11:21 AM

    Shawn,

    Tony’s got part’s three and four nearly written and completed. Expect those in the coming weeks. He’ll likely be contributing heavily to our annual power rankings this year (Jan 1), so be on the look out for those as well!

    B.D.,

    My gut right now is that we’re looking at 700-800k for 118, but that number could increase based upon a strong fight week promotion on behalf of the UFC. If the company really gets behind this thing, it could sell 1 million.

  5. Eric Nitsch on August 20th, 2010 7:30 PM

    You have provided us with some interesting data and are leading us to what I believe will be a thorough assessment of the topic. However, you present us with information like “Fighters look like terrible draws in 2006.” and “Single years can give a very different impression about the same fighter.” without ever explaining how these phenomena tie into the model that you are creating or how you plan to account for them. Perhaps I am just jumping the gun again, but I would like to know more about how you are going to resolve the “problems” you have encountered in your research. Keep up the thoughtful insight.

  6. Tony Williams on August 23rd, 2010 3:44 PM

    Hi guys,

    Sorry for the delay in responding. It’s been a hectic week.

    Diego – Thanks for the comment. Nice to know that someone out there is reading!

    Shawn – The first two parts ran one week apart, and you should see a new one this week. Work on my dissertation is my top priority, and this is the schedule that I thought I could commit to Adam and Kelsey when I came on board.

    Big Daddy – It’s hard to make a guess here, since we have no past information telling us what sort of interest there is in a James Toney MMA bout. You saw with the buyrate for the last PPV that Chael Sonnen wrecked the trending data that seems to be generally reliable by promoting the fight in a way we haven’t really seen before. I would guess 500,00-600,000, with 550K-575K being my best guess. However, it all really hinges on how many people want to see Toney fight.

    Eric – Great response, again! I’ll be long-winded in responding, but let me know if I don’t really answer your question. The two statements you quoted were meant to point out some of the dangers in simply taking these numbers as being “right.” For instance, I for some reason I wanted to write a piece about Randy Couture being overpaid and over-the-hill, I could just pull the numbers for 2008-2010 above and say that Couture went from pulling over a million buys to drawing a little more than a quarter of that just two years later. Even though the numbers are true, you’re still missing a lot of the story if you just look at the numbers and don’t think about where they’re coming from. This is one of the main points that I hope comes across in the series.

    As for fixing the problems… only some of them are problems, and only some of the problems can be fixed. The big problem that you’ll see possible solutions for has to do with finding an appropriate comparison. We can find alternative comparison groups, look at those numbers, and see if we learn something new.

    For problems that are difficult (maybe impossible) to “fix,” you’ll see a lot of that in the final piece. As a preview (since I haven’t started writing it yet), one issue is that solving one problem may require not solving another one. I’m sure that sounds odd, though, so let me try to come up with an example. Kelsey wrote a few weeks back on the title versus non-title gap. In the comments on my first article, Diego and BrainSmasher brought up the issue of adjusting for title fights in the number I’m presenting.

    The statistics nerd in me says you should toss these things into a regression with a bunch of dummy variables, but that may not even be possible with all the interaction terms you would want to include (since they end up being the numbers you’re really interested in) because of the relatively small number of UFC PPVs (which gives our sample size). You lose a degree of freedom for each additional thing you estimate. For instance, say you want to see the draw for Lesnar, the draw for St. Pierre, and then you want to see what they pull together (since they’ve been on two cards together). OK, run a regression with a dummy for Lesnar, a dummy for GSP, and a dummy for having both on the card. When you try to do that for all the possible combinations that we’ve seen (Penn & Silva, Couture & Liddell, etc), you start losing degrees of freedom rapidly. We can easily come up with a list of important factors driving PPV buys that’s longer than the number of UFC PPVs we have… but tossing all of those into a regression isn’t possible because you need more data points (i.e. UFC PPVs) than parameters that you are trying to estimate (eg. Lesnar, GSP, and Lesnar-GSP draws).

    To give a somewhat disappointing response, there’s never going to be a magic number that tells us everything we want to know, and I’m not going to lie to you and say that I’ve found one. Even with these very simple numbers, we find (and have been discussing) a lot of problems… and coming up with more complicated numbers brings up even more complicated problems. So, if we are curious about how strength of card affects PPV buys in the future, we may have to ignore other factors like specific fighters’ draws or controlling for international PPVs.

    See, I wasn’t kidding about being long-winded and not answering your question!

    The answer is that there’s not a single model I’m trying to build towards. To say that a bit differently, it’s more about finding an appropriate model to answer the question we’re asking. The “right” model to tell us how many PPV buys a fighter brings by being on the card isn’t necessarily the same as the “right” model to tell us how strength of card affects PPV buys.

  7. Eric Nitsch on August 23rd, 2010 6:03 PM

    Now that sir, is the kind of answer that I have been looking for. I will eagerly continue to follow your piece to its conclusion. Thank you for your responses.

Got something to say?

You must be logged in to post a comment.