NHL Fan Analysis Part 5: Defining Fan-Team Relationships with Social Media

Note: This is Part V of our study of NHL Fan Quality.  This week we will be ranking NHL teams/fans on the following dimensions: Fan Equity, Social Media Equity, Fan Equity Growth, Price Elasticity, Win Elasticity, and Social Media based Personality.  For more details on our measures of quality, please click here.  For Part I, click here.  For Part II click here.  For Part III click here.  For Part IV click here.

Social media is increasingly being used as a market research tool, and we believe that it provides opportunities to develop some richer descriptions of NHL fan bases.  The foundation for today’s analysis is something known as social media sentiment.  The idea behind sentiment is that we look at the “tone” of tweets surrounding each team.  In this study, we are examining the distribution of positive versus negative tweets for each team over the past three years.

Our actual approach uses a variety of statistics used to characterize distributions (e.g. mean, variance, skewness, kurtosis, etc.…), and then we employ a technique known as cluster analysis.  We will avoid the details (feel free to contact us) but the general idea is to find teams that have similar distributions of social media sentiment.  We use cluster analysis on team social media sentiment on Twitter over the past three seasons to dynamically segment fan bases (we allow fan bases to move across clusters over time).  Perhaps, it is more accurate to describe what we are doing as segmenting the types of relationships fans have with their teams.  Do fans have unconditional love for their team?  Do they have violent mood swings?*

Based on our dynamic cluster analysis of Twitter sentiment, we are able to describe each NHL fan base.  The chart below summarizes the social media “personality” of most NHL fan bases over the past three seasons.

Twitter Based NHL Personalities

*One caveat to this study is that since this is all based on Twitter data, the results reflect the opinions of fans on SOCIAL MEDIA only.  Also, please note that unlike our previous study of NHL social media equity that was based on the size of each team’s following, this analysis is based on sentiment or tone.

Mike Lewis & Manish Tripathi, Emory University 2014.

Chicago Business Journal: Why Chicago Blackhawks fans aren’t the best in the NHL

Chicago Business Journal: Why Chicago Blackhawks fans aren’t the best in the NHL

This is a statistics thing — at least as the numbers were crunched by the Emory (University) Sports Marketing Analytics team. Not sure if Atlanta, GA.-based Emory has the market cornered on expertise in hockey analytics. But they’re doing it.

And Chicago wound up in sixth place on Emory’s list of cities where hockey fans are most invested (think ticket sales) in their teams — right behind the aforementioned five Canadian cities.

NHL Fan Analysis Part 4: Social Media Equity

Note: This is Part IV of our study of NHL Fan Quality.  This week we will be ranking NHL teams/fans on the following dimensions: Fan Equity, Social Media Equity, Fan Equity Growth, Price Elasticity, Win Elasticity, and Social Media based Personality.  For more details on our measures of quality, please click here.  For Part I, click here.  For Part II click here.  For Part III click here.

NHL 2014 Social EquityToday we continue our analyses of NHL fan bases with something thoroughly modern: Social Media Equity.  In this analysis, we look at how teams’ combined social media following on Facebook and Twitter compares to teams that have similar records and populations.  Social Media Equity has some significant pluses in that it is not constrained by stadium capacity, and allows for including non-local fan support.  Social Media Equity may also be a forward-looking metric since social media is more prevalent among younger consumers.

The social media rankings are dominated by the traditional NHL powers.  Detroit is first followed by Boston, New York (Rangers), Pittsburgh and Chicago.  A significant difference between the revenue premium based brand equity ranking and the social media based rankings is the relative position of US and Canadian teams.  The US teams dominate the social media rankings while the Canadian teams dominate the Fan Equity rankings.  At the bottom of the rankings we have Anaheim, Columbus, Tampa Bay, Phoenix and St. Louis.  These tend to be the teams that struggle on many of our fan metrics.

Mike Lewis & Manish Tripathi, Emory University 2014.

NHL Fan Analysis Part 3: Sensitivity of Attendance to Price

Note: This is Part III of our study of NHL Fan Quality.  This week we will be ranking NHL teams/fans on the following dimensions: Fan Equity, Social Media Equity, Fan Equity Growth, Price Elasticity, Win Elasticity, and Social Media based Personality.  For more details on our measures of quality, please click here.  For Part I, click here.  For Part II click here.

The next study in our NHL fan analysis is a quick look at price sensitivity.  The central idea here is to understand (statistically) which teams’ fans are the most value-oriented and which are willing to spend.  The results are based on a regression model that predicts home attendance based on winning rates, prices, team fixed effects and interactions between the winning rates, prices and team dummies.

The most “value-oriented” fans live in Columbus, Ohio.  The Blue Jackets are in a tough competitive position.  While the Blue Jackets have fewer resources than most of the other NHL teams, their local competition is one of the biggest spenders and most successful college programs.  Following Columbus, other frugal fans reside in Colorado, St. Louis, and Minnesota.

On the other extreme, we have teams whose fans seem to be insensitive to the prices charged.  At the top of this list are the Buffalo Sabres.  The Sabres are followed by the Red Wings, Blackhawks, Bruins and Dallas Stars.  And interesting aside in this analysis is that several of these markets like Detroit and Chicago also made our list of teams with high sensitivity of attendance to winsWhile our conclusions are limited by our use of aggregate data this observation does suggest that these teams would be better served by increasing prices and using the incremental revenues to enhance team quality (assuming they have room under the salary cap). 

Mike Lewis & Manish Tripathi, Emory University 2014.

NHL Fan Analysis Part 2: Sensitivity of Attendance to Wins

Note: This is Part II of our study of NHL Fan Quality.  This week we will be ranking NHL teams/fans on the following dimensions: Fan Equity, Social Media Equity, Fan Equity Growth, Price Elasticity, Win Elasticity, and Social Media based Personality.  For more details on our measures of quality, please click here.  For Part I, click here.  For Part III, click here.

Today, we start to dive a bit deeper into our analyses of NHL fan bases.  Specifically, we look at the top teams in terms of sensitivity of attendance to wins.  These rankings are based on a regression model that includes team fixed effects, average prices, winning percentage and interactions between the team dummies and the winning and pricing variables.  We use data from 2000 to 2014 for the analysis.

These analyses are instructive as they provide a look at the consumer behaviors that generate our fan equity rankings.  But these analyses also raise questions.   The best way to consider the results is that they are driven purely by the data.  What we don’t have is the underlying theory.  The best example of this occurs with our win sensitivity rankings.  What does it mean if attendance is very sensitive to winning rates?  Are the fans discerning and demanding of quality or are they bandwagon fans that abandon the team when things go bad.  We leave the interpretation to the reader.

Top Win Sensitivity NHL

The top five teams in terms of sensitivity to winning rates are the Detroit Red Wings, Chicago Blackhawks, St. Louis Blues, Buffalo Sabres, and Dallas Stars.  We suspect that this list will raise eyebrows.  The Red Wings are one of the most dominant franchises in the NHL, and are a great box office draw.  But, a closer look at the data reveals that this is a fan base that has perhaps become a bit spoiled.  When the Red Wings went from being a 70% plus winning percentage team in the early part of the decade to a 60 something winning percentage team around 2007 to 2010, Red Wing attendance dropped below 20,000 for a few years.  While this drop was small, the rate of attendance change versus winning percentage was large compared to other NHL teams.

The Blackhawks might be the best example of a fickle fan base.  When the Blackhawks are winning 30% of their games they draw around 13,000 fans per home game.  When they win 70%, they draw 22,000 fans.  The figure below shows a scatter plot of winning percentage and home attendance for the Blackhawks.

ChicagoScatterplot On the flip side, we also have a list of the five fan bases that react the least to changes in winning rates.  Here, again we have multiple interpretations.  Are these fans completely loyal, and therefore they show up regardless of if the team is winning or losing?  Or are these fans there for the fireworks and ancillary entertainment, and thus barely know that a game is going on?

Bottom 5 Sensitivity NHL

The teams with the least sensitive fans bases are Arizona (Phoenix), New Jersey, Columbus, San Jose and Florida.  Arizona, in fact, has a negative correlation between winning rates and attendance (see figure below).  Yes, we know that there are some factors that are not in the model such as time in market, but this pattern is striking.  Basically, the take away is that for these teams performance on the ice is secondary.  This actually could be the right strategy in small markets, or markets where hockey doesn’t have a deep tradition.  It makes sense to play up the entertainment aspects in a place like Columbus where a team may never be as consistent a winner as the across town competition.


Mike Lewis & Manish Tripathi, Emory University 2014.

NHL Pricing: A Social Media Based Approach to Assessing Ticket Pricing “Fairness”

Of late we have been looking at value provided by sports franchises in different leagues.  For most of these analyses, we have basically focused on how much fans are asked to pay for each win.  We also make adjustments for factors related to market size, median income and capacity.  Today’s analysis looks at pricing in the NHL.

Of all the pricing analyses we have done, the NHL is the strangest.  The most surprising result is a lack of a positive correlation between winning rates and ticket prices.  Our standard procedure is to develop a model that predicts ticket prices as a function of winning percentage, payroll, market size, median income and other factors that we would expect to be related to demand for tickets.  We do a lot of testing in these models in terms of evaluating different specifications (interactions, nonlinear effects, etc…). In none of these specifications did we find a significant positive relationship between winning rates and prices.  The most powerful predictor was median income.

The other thing that we have been experimenting with in these models is using social media data as an explanatory variable.  The logic is that social media metrics (follows and likes) provide an unconstrained measure of fan support.  This provides a means to assess the relative aggressiveness of how team’s price.

Something to consider in these pricing analysis is the question of how prices are set.  At one extreme, we might suppose that prices are set in order to maximize revenues.  This is a reasonable starting assumption but the implication is that teams are extracting every dollar possible.  On the other hand, teams may price below fan’s reservation prices if the team is trying to build brand loyalty.  The key point is that while consumers might be willing to pay very high prices, if they don’t view the prices as “fair” then loyalty can be adversely affected.  Perhaps the best way to look at our list is that the teams at one extreme price the least aggressively (most benevolently?) while the teams at the other extreme are trying to extract every dollar they can from their fans.

At the top of the list we have Ottawa, Dallas, Boston, San Jose and Chicago.  After adjusting for market sizes, income levels and social media presences we find that these teams underprice. This is an interesting list as it contains both high brand equity teams like the Blackhawks and the Bruins as well as less prominent teams like Dallas and San Jose.  It is also notable that the Blackhawks and Bruins price above the league average while Dallas and Ottawa price near the bottom.  Interestingly, over the past 3 years Ottawa has basically sold out its arena.  The implication is that Ottawa (and the other teams on the list) could likely impose a price increase without too much loss of demand).

At the other extreme we have Philadelphia, Florida, Winnipeg, Toronto and Edmonton.  Again, this list contains both high (Toronto, Philly) and low profile teams (Florida).  Toronto is especially notable as they charge by far the highest prices in the league.  Winnipeg’s price are also extreme as they price higher (according to Team Marketing Report) than teams in New York, Chicago or Los Angeles.

Mike Lewis & Manish Tripathi, Emory University 2013.

Red Wings & Bruins Top NHL Social Media Equity Rankings

Last week, we published our ranking of NHL Team Fan Equity.  We have coined the term “Fan Equity” as a sports specific version of customer equity.  This metric is driven by “economic” measures of loyalty.  But, we do realize that a fan base also includes factors such as passion and engagement that may be (to some extent) overlooked in an economically driven ranking.

As a second look at fan base quality in the NHL, we use an approach that removes factors such as non-revenue maximizing pricing policies and capacity constraints that limit our ability to measure the customer equity of a hockey team.  The ranking we present today is based on what we call “Social Media Equity”.  The ranking is developed as follows:  First we collect information on each team’s social media presence such as the number of Twitter followers and Facebook likes.  We then develop a statistical model that quantifies the relationship between these social media metrics and measures of performance such as the team’s winning percentage for the last three years, and market factors including median income and metro area population.  We also include each team’s number of tweets in the model.  We then look at the difference between predicted social media followers and actual social media followers.  This delta between predicted and actual followers is reflective of Social Media Equity.

This social media based rankings has both pluses and minuses.  On the plus side, fan interest is not constraint by either high prices or stadium capacity.  On the negative side, while liking a team on Facebook or following them on Twitter shows fan interest, we can’t economically quantify this interest (these teams are businesses after all).

The number one team in our social media ranking is the Detroit Red Wings.  This is not a surprise, as the common wisdom is that the Red Wings are the number one team in Detroit (at least according to Professor Lewis’ sister in law).  The Red Wings have great social media presence on both Facebook and Twitter.  In positions 2 through 5 we have the Bruins, Devils, Flyers and Avalanche.  These are all big time fan bases with the exception of the Avalanche (actually we are not sure about the Devils but there was an episode of Seinfeld involving face painters so we assume Devils’ fans are indeed very passionate).

The Avalanche is where the story gets interesting.  While the Avalanche rank only 21st in the league in Twitter followers and 12th in terms of Facebook likes, they have achieved these results in a relatively small market, while often struggling on the ice (16-25 last year).  Our model suggests that Denver is potentially a strong hockey market.

The Social Media Equity results are a bit different than the economically driven results of last week.  Notably, the Canadian teams drop quite a bit.  In the social media rankings, Montreal finishes 6th, Vancouver 8th and Toronto 10th.    We can only speculate as to why the results differ.  Perhaps the previous results overrate the fan bases of the Canadian teams because Canadians are too nice to balk at high prices.

At the bottom of our list we have Tampa Bay in 5th from the bottom, Ottawa in 4th, the Kings in 3rd, Columbus in 2nd and Anaheim in last place.  Later in the year we will combine our various rankings to come up with a list of the best and worst sports cities.  We expect that Tampa Bay will be a front runner for one of those lists.  Ottawa performed similarly poorly in the economics based list so the verdict is in, Ottawa is the worst hockey city in Canada.  LA has a decent social media presence, but when we adjust for team performance and market size the results are not pretty.  Finally, perhaps Columbus should call themselves the “Buckeyes” rather than Blue Jackets.

The Winnipeg Jets are excluded from the rankings because the team moved from Atlanta during the period of the study.

Mike Lewis & Manish Tripathi, Emory University 2013.

NHL Fan Base Rankings: Americans may like Hockey, but Canadians Love it


For our NHL Social Media Equity Rankings, please click here.

A quick search of the Internet about who has the best fans in any sport will lead to multiple articles and rankings.  These rankings tend to rely a lot on personal opinion, and very little on any type of analysis.  The best of these studies tend to use a little bit of data concerning metrics like attendance, or maybe how many “likes” the team has on Facebook.   Occasionally, the ranking will be some type of weighted average of several pieces of data.  The vast majority of these approaches are badly flawed.  In the case of looking at raw numbers such as attendance, a frequent mistake is to ignore that attendance is driven by winning rates.  If this is the case, then such a study inflates winning teams’ fan bases by including bandwagon fans.  In the case of using a weighted average of multiple criteria, we still have the problem of not accounting for winning rates, but we also have the problem that the “weights” for each factor tend to be arbitrary.

What we do in our rankings is to use a wide variety of data and some statistical modeling to get around these issues.  We use something called a revenue premium approach to assess a team’s fan equity (value of the fan base).  The basic procedure begins with a statistical model that predicts a team’s box office revenues based on market potential (population and median income), team quality (winning rates) and other factors (such as team payroll).  We then compare the predictions from this model with each team’s approximate box office revenues to determine which teams over and under perform.  More details on the approach are available here.  In today’s post, we rank NHL fan bases using the above approach.  Later in the week, we will present results that rank teams based on social media equity (rather than the economic value of the fan base).

Using the past three years of results, we find that the best NHL fan bases live north of the border.  In first place, we have the Toronto Maple Leafs.  The Leafs pack the fans in despite charging the highest prices in the league.  The key point is that while the Leafs have been up and down the last few seasons, the fans continue to show up and pay premium prices.

In second and third place, we have Edmonton and Montreal.  The Oilers ranking second may be a bit of a surprise given some of their recent struggles on the ice.  But Edmonton continues to sell out their building on a regular basis, while charging fairly high prices and losing more than half their games in recent seasons.  Remember, Edmonton does this with a metro area population that barely exceeds one million.  The Canadiens are number three on the list.  A comparison between the Canadiens and the Chicago Blackhawks might be instructive.  These two clubs are fairly similar in box office performance. The Hawks sell a few more tickets but Montreal charges higher prices.  But, Montreal achieves their results in a metro area a third the size of Chicago’s, and without being one of the best teams in the league.

In positions 4 through 6 we finally see the Americans represented.  The Penguins come in 4th, the Rangers 5th and the Flyers 6th.  Our initial reaction to these results was that Pittsburgh is a heck of a professional sports city.  The Steelers were the leaders in our study of social media equity in the NFL.  The Rangers and the Flyers are both solid franchises across all dimensions.

One of our favorite parts of doing these rankings is determining the bottom five.  It’s fun because we typically get to be insulted by folks from all over (thankfully, the Trashers left Atlanta so we are spared the local abuse*).  San Jose and Anaheim are 5th and 4th from the bottom, respectively.  Californians seem to be the opposite of Canadians (take it as a compliment or insult).  Third from the bottom is the Phoenix franchise (We’re not even sure of their name). Second from the bottom we have the Ottawa Senators.  This is just embarrassing for a Canadian team.  Let us respond to the Ottawa fans right now.  We don’t care that you sell out – read the description of the method.  In last place, we have Dallas.  Why would anyone move a hockey team from Minnesota to Dallas?

*On a related note, the Winnipeg Jets are excluded from the rankings because the team moved from Atlanta during the period of the study.

Mike Lewis & Manish Tripathi, Emory University 2013.

Why Sports Fans Hate Dynamic Pricing

The dynamic pricing trend keeps accelerating with the recent announcement that the Toronto Maple Leafs will be pursuing a form of dynamic pricing.  The Leafs’ policy seems to involve more of quality based pricing than any true dynamics in that the plan involves charging different prices for different tiers of games (based on opponent and game time).  Some might quibble that this should be termed variable rather than dynamic pricing, but ultimately this is a form of price discrimination that can adversely affect how fans think about teams.

In our first two entries on dynamic pricing, we have discussed some basic principles of dynamic pricing (revenue management).  In this entry, we begin to consider the specific challenges of implementing these practices in sports contexts.  Industries that employ dynamic pricing will often make impassioned defenses of the practice.  Typically, the industry defense emphasizes that dynamic pricing is “fair” because market forces are dictating prices.  What could be more “fair”?

Obviously, many consumers don’t view dynamic pricing as a positive development.  Dynamic pricing is something that at best makes consumers nervous, and at worst angry.  This is not a surprising reaction, since dynamic pricing is at heart a system of price discrimination and inventory rationing.  If we think about the connection that exists between a consumer and a firm as a relationship then these practices are obviously not going to improve the “relationship.”  In the previous paragraph, we placed fair in quotes for a specific reason.  Fairness is subjective and teams do not get to decide how fans feel.

If consumers accept dynamic pricing in travel industries, why should these techniques be a problem in sports?  Let’s consider a couple of key differences between air travel and sporting events.  First, in the airline industry, the consumers who typically pay the highest rates do not pay for their own tickets.  In the airline industry, the business customer often chooses the travel options while the firm pays the price of the travel.  And as a bonus, the employee collects loyalty points.  A second, key difference is that air travel is purchased in order to achieve some other goal, whether it is a business meeting or a visit to a resort.  In other words, air travel is a product that is purchased so that the buyer can do something else.  In contrast, dynamic pricing of tickets affects the focal experience rather than an intermediary transaction cost.

A third issue is that the relationship between a fan and a team is fundamentally different than the relationship between an airline and a traveler.  There is a reason why we call consumers of sports “fans” rather than “customers.”  In what other categories, do consumers proudly wear the brands they consume on their clothing?  For example, Manish’s wardrobe seems to consist of mainly t-shirts emblazoned with team logos and cargo shorts.  In the last week I have seen Cubs, Maple Leafs and Redskins t-shirts.  I raise the issue of clothing because it highlights that consumers want to have a strong relationship that includes being publically associated with the team.  In marketing, we might refer to this as a desire for a “communal” relationship.  In contrast, a reliance on strict demand based pricing will tend to reduce the fan-team relationship to a series of cold economic exchanges.

Mike Lewis & Manish Tripathi. Emory University 2013