Gold Cup Analysis Part 2/2: Understanding and Predicting 2015 Attendance

In the second of this two part series, Business of Soccer examines CONCACAF’s motives, preferences, and risk considerations when selecting host venues for its regional showpiece event.  Specifically, we evaluate whether attendance data and key drivers identified in our Historical Trend Analysis might be used to forecast attendance for the 2015 CONCACAF Gold Cup, applying a simplified regression model to the schedule and announced venues to predict individual group stage match turnouts, while evaluating “best case” and “worst case” outcomes for the knockout stage

The Field is Now Finalized, and Anticipation for the 2015 Gold Cup is Justifiably High.

The 2015 CONCACAF Gold Cup field is now officially set, with Honduras narrowly avoiding a shock defeat to French Guyana to claim the last remaining berth, rounding out the field of 12 teams who will contest the region’s bi-annual championship this summer.    The result, along with the March 12 tournament draw, clarifies the initial groupings and match schedule that participant teams will face when the tournament gets underway in July.

As things stand, many anticipate that the 2015 edition of the Gold Cup will be among the most closely contested in the tournament’s history, with World Cup quarterfinalists Costa Rica and 6-times winners Mexico both expected to provide a strong challenge to defending champion USA, who themselves have the added incentive of securing a spot in the 2017 FIFA Confederations Cup with a win.   Meanwhile, traditional “also-ran” countries such as Panama (2013 GC Runners-Up) and Honduras (2013 GC Semi Finalists and World Cup qualifiers in 2010 and 2014) have begun achieving competitive results that suggest they might legitimately consider themselves “dark horses” for the regional crown.   Given the historical absence of competitive depth in CONCACAF, there is reason to believe that the tournament will be among the most compelling to date.

Certainly, CONCACAF’s leadership will be hoping that the spectacle justifies the anticipation, in no small part because of the potential revenue that a successful tournament can generate for the confederation.    Unlike its more established international predecessors – namely the European Championship and the Copa America – the Gold Cup does not command massive global brand recognition, and television revenues remain modest, at best.    Even within the region, the tournament faces stiff competition for the attention of casual sports fans and “neutrals”.  At least for the moment, gate receipts and direct fan engagement remain  critical to the tournament’s financial viability and growth potential.

Recent editions of the competition have given organizers good reason to feel optimistic in this regard.  As previously detailed in Part 1 of our analysis, the tournament has been increasing significantly in popularity since its inception in 1991, with the record-setting 2011 edition topping 1 million in total attendance, bringing the tournament on par with its global peers.

Exhibit 1

So… 2015 Gold Cup Attendance Could Set an All-Time Record, Right?

Perhaps… but it’s not so simple.  The Gold Cup continues to be a “work in progress” in many senses, and organizers have experimented considerably over the past 20+ years in attempting to develop a blueprint that suits the unique challenges a region long characterized by both economic and sporting disparity.

Questions facing organizers are numerous:  When should the tournament be played?  What does an ideal “host city” look like (given the US is likely to remain the primary host for the foreseeable future)?  What is the target stadium size?  Does replacing regional minnows like Belize or Guadeloupe with higher-profile “guest teams” like Brazil or Colombia add value to the competition?  How can risks to attendance be mitigated?  Do the changes to the tournament come at the expense of “competitive fairness” of a tournament that already appears to favor the established competitive and economic regional juggernauts?

CONCACAF deserves credit for retaining an open-minded approach in their efforts to address these questions, taking numerous risks with the tournament format and showing an admirable awareness of what has “worked” and what has not.  The field has now featured 24 different countries (including 6 non-CONCACAF guests), and has been hosted in 34 different stadiums throughout the US and Mexico (with Toronto set to host the first Canadian-based matches this summer), encompassing both summer and winter months.  For the most part, these innovations have paid dividends, with attendance boasting a nearly threefold increase since the Gold Cup’s inception in 1991.

Exhibit 2


However, organizers also face challenges that are beyond their control, but which nonetheless have a considerable bearing on the tournament’s financial performance and brand.   Namely, how does one account for and/or mange the fact that overall attendance and interest have been undeniably linked with the performance of a small number of dominant teams (most notably Mexico)?

Historical data leaves no doubt that Mexico is the tournament’s most popular draw, with average attendance of 49,100 dwarfing that of its closest rival USA (34,600).  Among the 23 matches in Gold Cup history that have sold out “full-scale” stadiums (capacity greater than 30,000), none have come without Mexico’s matchday participation, with primary host USA providing the only statistically meaningful “premium” above average/median attendance in this regard.

By contrast, the 2000 Gold Cup, in which neither Mexico nor the USA advanced past the quarterfinal, saw disastrously anemic attendances of approximately 3,000 for the semifinals and 7,000 for the Canada-Colombia final.  In fact, the “worst attended” matchdays in Gold Cup history share an unmistakable genetic marker—“No Mexico” seems to be the one true recipe for poor turnouts.

For organizers who must concern themselves with revenue and brand value, this “top heavy” dependence represents a serious risk.  Indeed, it is a challenge that is particularly unique to CONCACAF–Do organizers in UEFA worry whether the final is contested by Spain and the Netherlands?  England and Germany?  Italy and France?  Doubtful—the strength and diversity of competition nearly ensures that the latter stages will be contested by a small segment of established global powers.  Even the presence of the odd “underdog” is just as likely to enhance the appeal to neutral observers.

At this Stage, Simply Maximizing “Overall” Attendance May No Longer Be the Primary Goal  

“Call me crazy, but it is my philosophy that in order to be successful, one must project an image of success, at all times”

Some may recall the above quote from Buddy Kane, the smarmy local real-estate mogul portrayed by Peter Gallagher in 1999’s Oscar-winning American Beauty.   While the “Real Estate King” may not have been talking about soccer when he delivered this memorable line, his pearl of wisdom may nonetheless be effectively applied to CONCACAF and its recent approach toward the Gold Cup.

Given the incredibly rapid growth in Gold Cup attendance over the past two decades, one might assume that tournament organizers have focused on securing the largest and most spectactular venues available within the continent’s unrivaled collection of  soccer-ready professional sports venues, with an eye toward eventually becoming best-attended regional tournament on the globe .

In practice, CONCACAF has taken a considerably different approach. The current philosophy appears to be focused far more on the prudent management of capacity factor (i.e. documented attendance divided by stadium capacity) than the outright maximation of total attendance.    Taking a page from Major League Soccer’s recent success in boosting in-stadium viewership and engagement, it would appear that CONCACAF would prefer to “fill the house” for as many games as possible, rather than risk a reputation of being a second-rate competition , attended sparsely by “soccer hardcores” in familiar stadiums regularly filled to capacity by more popular sporting events.

As shown in Exhibit 3 below, CONCACAF organizers have actually been reducing the size of host stadiums considerably over the years, with large NFL-style stadiums being replaced by  smaller “soccer specific”venues , particularly for group-stage matches.    The average capacity of stadiums used for Gold Cup matches has now declined by nearly 44,000 (or about 48%) since the inaugural edition in 1991.

Exhibit 3

The difference becomes even more pronounced when we evaluate the selected venues for Group Stage matchdays featuring Mexico games, and those that do not.  As shown in Exhibit 4 below,  CONCACAF has been drastically reducing the venue size for Group Stage matchdays that do not feature Mexico since adoption of the current format in 2005.  While Mexico wil continue to play its scheduled initial Group C matches in current NFL stadiums with an average venue size in excess of 66,000, the remaining teams in Groups A & B will play in venues with an average capacity of just under 30,000, or less than half the size of the average Group C venue.

 Exhibit 4

So just how does this strategic decision to reduce venue size based on participants figure to affect the attendance for the 2015 Gold Cup?  In order to properly assess this question, it is clear that we need some means of predicting match-by-match turnouts so that we can evaluate and quantify the potential impact of this shifting philosophy.

Predicting 2015 Gold Cup Attendance – Regression Methodology & Assumptions

Understanding exactly what drives attendance for Gold Cup matches is obviously not a perfect science—intuition plays a significant role in interpreting what ultimately amounts to a limited sample encompassing 12 tournaments and fewer than 250 total matches, only half of which have been played under the “current format” (12 teams / 3 groups / no guest teams) adopted in 2005.    Nonetheless, for those willing to accept a certain degree of conjecture, the data does provide a serviceable starting point for interpretation, extrapolation, and ultimately prediction.

Regression analysis is a tool that involves using historical input data – such as match timing, location, stadium capacity, tournament stage, and participant teams – to explain and/or predict an “output” (in this case, match attendance).  In so doing, we can create a model that forecasts match-by-match attendance for the 2015 Gold Cup that is informed by what we have learned from past tournaments,( at least in theory).

In the Business of Soccer case, the regression formula used is summarized in Exhibit 5 below:

Exhibit 5

Certainly, it would be possible to construct a more detailed (and theoretically more accurate) case using additional variables and/or more sophisticated and complete datasets.   For example, one might also choose to include venue-related data from other related competitions (World Cup qualifiers, club competitions, friendlies, etc), as well as general market factors that measure the growing regional “soccer interest” (TV ratings, league/club level attendances, participation rates in recreational leagues, etc).   The possibilities for fine-tuning the analysis are endless, and organizers from CONCACAF and US Soccer doubtless have greater resources than the limited publicly available data used herein.

Nonetheless, the formula above provides us with a serviceable starting point for forecasting what this summer’s attendance might look like given the dates, venues, participant teams, and history. Applying these previous Gold Cup attendances for the past decade, we can see in Exhibit 6 below that it actually tracks reasonably well to actual historical results:

Exhibit 6

Regression Results and Implications: Group Stage

Using the above formula as a starting point, we can now make some initial predictions about potential turnout for the 2015 Gold Cup, first beginning with the Group Stage:

Exhibit 7

A simple glance at the above table demonstrates just how carefully – perhaps even conservatively – CONCACAF has planned in attempting to maximize match capacity factors while limiting sub-standard (less than 50% capacity) turnouts that plagued the competition in its early iterations.   Our regression suggests that 10 of the 18 scheduled Group Stage matches will be sellouts, with a very  impressive overall group stage capacity factor of 87.6%.  This expected strong capacity is no “random accident”.  Clearly, organizers have attempted to place “less popular” matches in more intimate venues where they  feel confident that teams can produce sellouts or near-sellouts, regardless of the venue’s size.

For example, Groups A and B (seeded by USA and Costa Rica, respectively) will play their initial matches  primarily in smaller (less than 27,000 seat) soccer-specific stadiums, and appear likely to sell out these matches based on historical data.  Though not shown in the data above, the BOS forecast actually predicts that demand for these games will outstrip seating availability by as much as 160,000. The lone exception involves the USA’s Group A match with Haiti at the 68,000-seat Gillette Stadium in Foxboro, MA, a doubleheader which also features a showdown between Honduras and Panama that may be critical to the hopes of both teams in competitive terms.   Organizers will be hoping that the venue’s historically strong performance as a host of USA matches, as well as the match’s timing (Fridays have been favorable days for summer Gold Cup matches) will enhance the forecasted attendance of approximately 46,000.

By contrast Mexico (the Group C seed) will play its group-stage matches in NFL stadiums with an average capacity of approximately 66,000. Given Mexico’s overwhelming history as the “big draw” in driving tournament attendance, CONCACAF is betting that EL Tri will be able to attract crowds that offset the smaller turnouts from the Group A & B games.

All in all, total aggregate Group Stage attendance of 602,753 (out of a maximum 756,362) remains a very attractive proposition for CONCACAF organizers, and provides a solid “going-in” foundation entering the less-certain set of fixtures that comprise the Knockout Stage.

Regression Results and Implications: Knockout Stage

As for the Knockout Stage, there remains the obvious matter of predicting match results, which is required in order to determine participants for latter stage games.  While this is clearly an arbitrary exercise and  subject to debate, we begin by constructing a simple “Base Case” using the monthly Coca-Cola FIFA ranking (obviously imperfect, but suitable for our purposes as a “starting point”).  From there, we sensitize the potential results under a number of “extreme” scenarios in order to determine an overall range of aggregate  potential attendance.  Assumptions and results for this Base Case are summarized below:

Exhibit 8

Exhibit 9



Overall Conclusions:  What Does This Data Imply:?

Ultimately, risk-adjusted attendance of 1.0-1.1 million – even under the highly unlikely “worst case” scenario described in Exhibit 9 above – represents exceptional progress for a tournament that lacks the global appeal, history, revenue streams and competitive depth of its more established rivals.

The emerging theme in CONCACAF – perhaps first identified by Major League Soccer, but certainly adopted enthusiastically by Gold Cup Organizers – involves a clear and demonstrated preference for optimizing the capacity factor over simply “maximizing attendance”.

In practice, this approach is informed by a number of critical realizations:

  • Recognizing and consciously avoiding the “old” reputation of CONCACAF soccer – particularly when played in the US – as that of “second tier” sport, marked by empty seats, a niche audience, and limited appeal compared to more established and popular sports such as American football, basketball, and baseball
  • Taking advantage of the growing number of new soccer-specific stadiums available in North America, which provide better in-stadium fan expiences and improved operating cost efficiencies compared to conventional large-scale venues, and which also project a more exciting atmosphere for viewers at home watching via television
  • Showing awareness of “what can be controlled” (i.e. scheduling and venue selection for Group Stage Matches), and minimizing downside risk associated with less predictable matches, such as those in the Knockout Stage
  • Concentrating potential attendance “upside” in the hands of the teams most likely to realize it. Mexico are the tournament’s strongest drawing team by a considerable margin, and CONCACAF has made certain that El Tri will play all of its matches in NFL stadiums with capacities in excess of 60,000.   Indeed, as the regression suggests, matchdays involving Mexico are likely to draw crowds in excess of 16,000 more than “normal” matchdays.
  • Limiting the downside impact of the region’s less-popular teams by assigning their matches to smaller, soccer-specific stadiums in regions where teams have historically drawn strongest atentandacnes. This might even involve leaving excess demand “on the table” in some cases.
  • Recognizing that diversification of host venues is a great way to ensure strong across-the-board attendances. It is worth noting that the 26 total matches will be contested in 13 different cities (almost entirely in same-day doubleheaders), meaning fans will not have much ability to “pick and choose” the games/teams they would like to see before purchasing tickets.  In previous tournaments, where multiple matchdays were hosted at venues like the Rose Bowl, Orange Bowl, Cotton Bowl, and Estadio Azteca, organizers witnessed severe dropoffs in attendanes for less-popular or less-convenient matches.  Without this optionality, fans interested in attending games will simply have to take what is given to them, reducing the risk of poorly-attended “duds” in the pool of potential matchups.
  • Understanding that continued growth may not happen as quickly in the coming decade as it has during the previous two decades, The “maximum possible” attendance for the 2015 Gold Cup – assuming every game is a sellout – is just under 1.30 million, only about 160,000 (14%) higher than the all-time record attendance of 1.14 million, set recently in 2011.   This does not leave a lot in the way of excess capacity in the event that the tournament proves to be wildly popular, showing that CONCACAF is more concerned with stabilizing attendance for an increasingly mature tournament, rather than adopting a “shoot for the moon” approach

A successful 2015 Gold Cup, followed by a similarly successful Copa America in 2016, appear to be the next steps for organizers concerned with “projecting an image of success”  to both casual domestic fans as well as the global soccer audience.  Continuing to stage compelling, competitive, and creative tournaments will strengthen the long-term brand value of soccer within the region, and may eventually help unlock the step-changing global TV audience that has thus far eluded regoinal competitions at both the club and international level.


What do you think about the potential for Gold Cup attendance and how the tournament has progressed over the years? Let us know in the comments section below, or via Facebook or Twitter.

Reporting on the business side of the world's game.