Don’t Want to Get Fired? Best and Worst Cities for Firing Professional Coaches

Mike_Shanahan_RedskinsIt’s “Black Monday” in the NFL.  The Vikings, Redskins, Lions, and Bucs have already fired their coaches today, and more firings are possible before the day is done.  There are many variables that can affect the firing of a coach in professional sports.  Of course, three easily observable factors are the performance of the coach (winning percentage, playoff appearances, and championships), the investment by the ownership (team payroll), and the sports league (NFL, MLB, NBA, and NHL).   There are also intangible factors endemic to each city in America and Canada with a professional sports team that can influence the probability of a coach getting fired.

We decided to estimate a logistic regression model that could explain the probability of getting fired as a function of performance, investment by ownership, and professional league affiliation.  We looked at data from all four professional sports leagues over the last twelve years.  We then compared the predicted probability from our model of getting fired with the actual firings in each city.  In theory, cities with intangible characteristics that make it more likely for a coach to get fired would have actual firings at a higher probability than predicted through our model of performance and investment.  We tried several specifications of our model, and these rankings are robust.

Based on our study, the Top 8 Worst Cities (Highest probability for getting fired above predicted) are:

  1. Orlando
  2. San Francisco
  3. Montreal
  4. Sacramento
  5. Milwaukee
  6. Oklahoma City
  7. Jacksonville
  8. Miami

The Top 8 Best Cities (Lowest probability for getting fired below predicted) are:

  1. Winnipeg
  2. Nashville
  3. Salt Lake City
  4. Memphis
  5. Los Angeles
  6. Portland
  7. Buffalo
  8. Minneapolis

It’s interesting to note that the top 8 worst cities does not include big media markets like New York, LA or Chicago, where one might think there is large expectation for winning.

Mike Lewis & Manish Tripathi, Emory University 2013.

NFL Fan Equity: Maybe the Cowboys are America’s Team?

Note: This was originally published on August 15, 2013

The NFL is America’s favorite professional sports league, but which of its teams has the most loyal and supportive fan base?  This is not a straightforward question.  A ranking based on attendance would be skewed toward teams that play in more populated metropolitan areas, and a ranking based on profitability or revenues would be biased in favor of teams that are currently enjoying more on-field success.

In our series of fan base analyses across leagues, we adjust for these complicating factors using a revenue premium model of fan equity.  The key idea is that we look at team box office revenues relative to team on-field success, market population, stadium capacity, median income and other factors.  The first step in our procedure involves the creation of a statistical model that predicts box office revenue as a function of the aforementioned variables.  We then compare actual revenues to the revenues predicted by the model.  Teams with relatively stronger fan support will have revenues that exceed the predicted values, and teams that under perform have relatively less supportive fan bases. We provide more details on the method here and here.

The top fan base was the Dallas Cowboys.  Professor Lewis grew up a Steelers fan in the 1970s so this was a bit of a painful result.  Professor Tripathi grew up as a Redskins fan, and is terribly disturbed by the results of the study.  What are keys to the Cowboys’ ability to create a passionate and supportive fan base?  We think it’s a long legacy of success, a football mad Texas culture and a state of the art stadium.  Over the last three seasons (the time period used to calculate fan equity) the Cowboys have played sub .500 football but generated above capacity attendance (at least according to ESPN).

In positions two and three we have the New England Patriots and the New York Jets.  New England has an all-around strong fan base, while the Jets are somewhat similar to the Cowboys in that they draw consistently well, regardless of the on-field product.  In fourth and fifth place we have the New Orleans Saints and the New York Giants.  The Saints are a more recent success story, but the team’s new success combined with limited professional sports options in New Orleans has created a very strong fan base.  Two New York teams in the top five is an interesting result when viewed in relation to our college football fan base analyses.  New York is (no surprise here) a pro sports town.  As an aside, we will be interested to see how much value the Big Ten gains from acquiring a foothold in the NYC market starting in 2014.

At the more unfortunate end of the scale we have a bottom five of Detroit, Tampa Bay, Arizona, Atlanta and Oakland.  Detroit, of course, suffers from a relative lack of on-field success and a struggling local economy.  But we should note that our method does explicitly control for these factors.  It may well be a matter of the Wolverines & Spartans winning the battle for fans against the Lions.  Similarly, teams like Atlanta and Tampa Bay may suffer from being located in SEC territory.

Note: Here are the first and second follow-ups to our study.  For an alternative fan ranking using “Social Media Equity,” click here.

Mike Lewis & Manish Tripathi, Emory University 2013.

Bowl Game Apathy via Twitter

Tonight’s bowl games include the Military Bowl, Texas Bowl and the Fight Hunger Bowl.  We took a quick look at Twitter activity and Twitter share of voice (the number of Tweets referring to each bowl divided by total Tweets) for the cities represented (over the  last week).

First, we found significant differences in the Twitter activity across the three bowls.  The Military Bowl has generated the most Tweets (2215 for the search term “Military Bowl).  In contrast the Texas Bowl has generated just over 1500 tweets and the Fight Hunger Bowl fewer than 1000.  To some extent these numbers really speak to the importance of creating “local” match ups when scheduling minor bowls.  Only the Military Bowl is relatively local for the two teams playing.

The figure below shows the Twitter share of voice for the home towns of each team playing.  Again we see much more activity for the Military Bowl teams.  This further suggests that more geographically accessible games will generate more local interest.  The other striking feature is how little interest occurs in the big cities represented.  It appears that in pro sports towns having the local college play in a minor bowl barely moves the needle.

So what is today’s marketing message?  When scheduling a minor bowl make sure to cater to the right customer segment.  Choose schools in areas where the team is the big show and make match ups with convenient travel.

bowlgame

Twitter College Football Review: A Tale of Two Heisman Winners

College football is a business and college football fans are a vocal group of customers.  Like many businesses, college football has a new opportunity to track customer opinion: The Twitterverse.  For a bit of a of “Holiday diversion” we are going to take a look at several college football stories from the past season using Twitter as the data.Manziel vs Winston

The first story is a look at the past two Heisman trophy winners.  The chart shows something call Twitter sentiment for Johnny Manziel and Jameis Winston.  Sentiment analysis is basically the ratio of positive to negative Tweets. The higher the score the more positive the Twitterverse is about a subject.

What we have done is track sentiment for a one year period for each player. The blue line shows the weekly sentiment for Johnny Manziel.  This starts off very high following his historic Heisman victory.  But there are some ups or downs in the off season.  The downs can be linked to some of the news stories about Manziel gambling in casinos or showing up at high profile events.

The big dip for Manziel occurred in June.  Following his comments about being eager to leave College Station, his sentiment dropped from 93 to 20.  “Johnny Football” did recover much of the lost sentiment but then dropped to an even lower score of 18 following the mini-controversies surrounding his appearances at UT frat parties, and leaving the Manning passing academy early.  Interestingly, his autograph controversy in August did not generate an equal backlash as his sentiment bottomed out at around 30.

From a marketing perspective, this is a fascinating story.  The biggest damage to his brand occurred when he offended the locals by talking bad about College Station and partying in Austin.  The impact of breaking NCAA rules was not as large an issue.  It’s one thing to break a rule, but it’s much worse for a brand to insult it’s primary customers.

In contrast, there isn’t much to say about Winston during the off season.  While his sentiment also bounces around, this is more of a problem of small numbers (minimal data) rather than anything else.  Manziel’s Twitter traffic absolutely dwarfed Winston’s.  In marketing language we would say that Manziel had much higher brand awareness.  This is important in the context of the Heisman race since publicity matters quite a bit.

winston_manzielDuring the season, we see a steady climb for Winston and an up and down pattern for Manziel.  These patterns obviously have a lot to do with how the teams played.  For example, A&M’s late season losses dropped Manziel’s sentiment to the mid twenties.   Departing from college football for a moment, think about what this means for companies interested in tracking customer satisfaction!  The Twitter data almost tells us exactly what happened in games each week.   We say almost because one of Manziel’s biggest gaining weeks followed a close loss to Alabama.  In other words, Twitter gives us instant feedback about team quality.

Just like Manziel, Winston also had a huge late season drop in popularity.  He bottomed out with a sentiment score of 13 in the middle of November.  In Winston’s case, the drop can be attributed to legal issues.  The interesting thing about Winston’s drop is how quickly the public forgave.  The combination of “no charges” and the Heisman victory drove Winston’s sentiment to 97 in December.  Winston also finally passed Manziel in terms of total Twitter activity in December as well.

In terms of the final word – a look at these two brands / quarterbacks over time says something profound (and obvious) about sports brands: If you win then the public will be very forgiving.

Mike Lewis & Manish Tripathi, Emory University 2013.

The Financial Impact of Mascots on Sports Brands

MascotValue

When we started this endeavor, we had no intention of spending time thinking and writing about mascots.  What we did plan on, was writing about sports marketing assets such as team brands.  However, as we have progressed we have found ourselves going beyond measurement of team brands to also look into how valuable brands are created.  Since mascots are an element of teams’ brands it makes sense for us to spend time on the topic.

We have been surprised by the interest generated by our previous work on mascots.  This interest is likely due to the fact that we go beyond emotion-based arguments, and try to examine how mascots affect the bottom line.  We should also emphasize that this is a statistically “tricky” area.  In general, there just isn’t enough variation in the world for us to perfectly identify how a specific or even a type of mascot impacts the fortunes of a given team.  For example, in the case of Native American themed mascots our perfect “data” would include examples of teams switching back and forth between Indian and non-Indian mascots.  This doesn’t mean that it is impossible to study how different types of mascots impact financial performance.  It just means that we have to make some assumptions and we have to make clear how these assumptions limit our results.

Today’s post is a bit more long-form than our usual entries.  This is because we have multiple issues, and because we want to be transparent regarding our assumptions.  The two issues that we address in this post arose from conversations with readers of our previous mascot analyses.  The first was a question related to some work we did related to the financial value of Native American mascots in professional sports.  In the previous work, we had simply looked at how teams with Native-American mascots performed relative to all other mascot types.  Our readers were interested on the impact of other classifications of mascots.  The second question was related to our previous work on college mascots.  Specifically, the interest was in the financial impact of using ‘live” animal mascots.  Frankly, this was a controversy of which we were unaware.

Why do Mascots Matter?

Before we get into the analyses, it may be useful to make a couple of comments regarding why or why not mascots matter.  There are a variety of theories about sports fandom, and almost all emphasize the importance of factors such as team history and fan community.  These are related because it is often the historical accomplishments of a team that provide a basis for fan communities.  For example, in Chicago fans still talk about the 1985 Bears, and it is doubtful that you can find many Steelers fans that don’t know about the “Steel Curtain.”

Mascots provide a symbol that can be a focal point for a fan community.  At a very simple level, when fans wear a jersey with a Redskins or Cowboys logo they are identifying themselves as part of a fan community.  There is research in psychology that that has studied the wearing of team symbols following wins and losses.  Researchers, unsurprisingly, find that team logos are worn more frequently after victories than after losses.  The term “Basking in Reflected Glory” has been used to explain this phenomenon.

Mascots may play a similar role in that they provide a shared experience.  When the University of Illinois dropped the “Chief,” t-shirts that commemorated the “last dance” of the Chief quickly appeared.  Illinois students witnessed the Chief’s halftime dance for decades, and this experience has therefore been shared across generations of students.

Teams’ and fans’ reluctance to drop or change mascots may be based on fears about how losing a focal symbol will alter the fan community.  In our first analysis of “Native American” mascots we looked at college basketball revenues for schools with and without this type of mascot.  We also included time since mascot change in our statistical models.  The key result was that switching away from a Native American mascot didn’t have a long-term negative effect.

Classes of Mascots in Professional Football and Baseball

But the college environment is unique in that we have a fair number of schools that have made switches.  At the professional level there isn’t a similar body of data that exists.  Not having perfect data doesn’t mean that we can’t study an issue (though many unimaginative academics might say so).  We just need to use a bit of theory to structure the problem and then be clear about the assumptions to avoid over-interpretation of the results.

We did perform a preliminary analysis related to the financial impact of Native American themed mascots.  That analysis was based on the simple idea that we could build a statistical model of team box office revenues as a function of team quality (winning percentage, playoff participation, etc…) and market potential (market population, median income, stadium capacity, etc…).  We included a binary (i.e. dummy or indicator) variable in these regressions to indicate if the team had a Native American mascot.  We also included an interaction variable between the Native American dummy variable and the year to account for changing consumer preferences.

One common response to this analysis was to ask how other types of mascots influence financial results.  We thought this was an interesting question.  But it was also a question that wasn’t straight-forward to address.  Our first stumbling block was how to determine the different mascot categories.  For example, we could have a classification of “human” mascots but then the question arises of if we should differentiate between aggressive humans such as Pirates or Raiders and the gentler Padres or Saints.  Similar questions occurred with animal mascots: should we have a separate category for birds and what about aquatic animals?

To get a handle on these questions we created something called a perceptual map.  Perceptual maps are used in marketing to visually display the perceptions of customers or potential customers along a number of dimensions (e.g. affordability, social appeal, etc…).  For our mascot study, the map was based on survey data that asked subjects to rate the similarity between team names.  The survey involved 18 team names split between the NFL and MLB.  We tried to assemble a cross section of names that included different types of animals (Tigers, Bears, Dolphins, etc…), humans (Rangers, Packers, Pirates, etc…), miscellaneous names (Rockies, Giants) and a split between baseball and football.  The technical term for the procedure is Multidimensional Scaling (MDS).

MDS is great in that we allow subjects the freedom to rate items however it makes sense to them, but this freedom comes with a cost: the perceptual maps generated do not come with labeled dimensions.  We generated a three dimensional perceptual map (using SAS software).  Dimension 1 (the horizontal axis in the chart below) seems to roughly correspond to human versus animal mascots.  We say roughly because Cardinals are rated more “human” on this axis than Packers.  A potential issue with our study is that subjects are rating the team names based on factors beyond the literal meaning of the name.  This is probably unavoidable given the focal nature of sports teams in American culture.  The second dimension (not displayed) was difficult to interpret.  At one extreme we had the Padres and Rockies.  At the other, it was the Dodgers and Packers.  One thought was that this dimension was about historical success.  However, the Steelers were in the middle of the scale.

The third dimension (the vertical axis in the chart below) was also difficult to interpret.  The Redskins and Indians are at the top of the scale while the Tigers, Cardinals, and Dodgers are at the bottom.  While we will not try to name this axis, it is interesting that the two Native American mascots were viewed as extreme on this dimension.

MDS Mascots

The fundamental point to the MDS exercise was to develop an understanding of how fans perceive different types of mascots.  Based on the preceding, we decided to evaluate four mascot types: Human, Native American, Animal and Other.

We conducted statistical analysis separately for the NFL and for MLB.  Our logic is that because the games are very different and played at different times of year, the effect of different types of mascots may vary.  For each league, we created statistical models of revenue as a function of winning percentage, winning percentage squared, playoff participation, relative payroll, population, population squared, median income and stadium capacity.

A baseline model (without mascot dummy variables) for the NFL yielded an R-squared of 0.44.  R-squared provides information about the goodness of fit of a model (the higher the R-squared the better the model fits the data).  This model was estimated using data from the 2002 to 2012 seasons.  In addition, all coefficients were of the expected sign.  For example, winning percentage was positively correlated with box office revenue.  We next estimated the same model but included the mascot dummy variables.  Including the mascot dummies increased the R-squared to 0.51.

The coefficients associated with each class of mascot are provided in the table below.  The model suggests that over this time period, having a Native American mascot had a significant positive revenue impact relative to the “other” category of mascots.  Animal mascots had a negative impact.

Mascot Type

Coefficient Value

T-Stat

P-Value

Native American

12,117,107.2

4.86

<.0001

Human

1,353,243.8

0.83

0.409

Animal

-3,567,963.7

-2.49

0.013

However, as we noted above, our analysis includes some strong implicit assumptions.  In the case of the NFL results above, the Native American variable is associated with just two cities: Kansas City and DC.  The danger is that this variable may be picking up some common trait of the two cities other than the mascot.  An additional concern is that the preceding model treats the mascot issue as staticIt seems more likely that opinions change over time.  To account for these issues we next re-estimated the model but now included interactions between time and the mascot indicators.  This model yields an R-squared of 0.55.  Again all of the control variables (win percent, population, etc…) are of the expected signs.

This model is the most instructive of the three models as it allows for both dynamic effects and lessens the concern about a shared latent factor between Kansas City and Washington DC.  The key result is that there seems to be a shift in preferences.  In particular, the Native American mascots seem to be becoming less popular over time.  Historically, the Chiefs and Redskins have been strong franchises so it makes sense that the static Native American indicator would be positive.  Given the increased scrutiny applied to Native American mascots it also makes sense that we observe a negative long-term trend.

Mascot Type

Coefficient Value

T-Stat

P-Value

Native American

21,861,806.2

4.89

<.0001

Human

-2,924,904.4

-1.19

0.234

Animal

-6,616,731.1

-3.32

0.001

Native American*YR

-1,636,981.4

-2.6

0.010

Human*YR

722,698.9

2.31

0.021

Animal*YR

508,348.0

2.15

0.032

In the preceding model the dependent variable is box office revenues (in constant 2008 dollars).  The interaction between time and the Native American dummy variable suggests that the value of having a Native American mascot is dropping by about $1.6 million per year.  Again, we fully admit that this is a messy statistical problem and readers may be able to construct alternative explanations for the findings.  But the KEY point is that we have intentionally performed a simple analysis in an effort to just let the data speak.  The data seems to be saying that considering mascot type significantly improves model fit and that Native American mascots are becoming less valuable brand assets over time.

In the case of MLB we executed a similar procedure.  The baseline revenue model for MLB used the same variables as the NFL analysis.  The R-squared of the baseline model was 0.627.  In the second analysis, we added dummy variables for the three classes of mascots: Native American, Human and Animal Mascots. In this case, the improvement in the model is minimal as the R-Squared increases to just 0.631.  None of the mascot dummies are significant.

Mascot Type

Coefficient Value

T-Stat

P-Value

Native American

-8,494.4

-1.64

0.1015

Human

-2,822.0

-0.92

0.360

Animal

3,782.2

1.22

0.224

 

However, adding the interactions between time and mascot type produces an interesting set of results.  In particular, we find the same pattern of results for the Native American mascot terms.  In both leagues these mascots have positive coefficient associated with the static dummy variable but a negative interaction between the dummy for Native American mascot and time.

Mascot Type

Coefficient Value

T-Stat

P-Value

Native American

24,567,815.9

2.34

0.0196

Human

697,834.0

0.11

0.909

Animal

22,957,750.4

3.48

0.001

Native American*YR

-2,675,563.5

-3.6

0.000

Human*YR

-260,405.6

-0.59

0.555

Animal*YR

-1,523,533.9

-3.28

0.001

In the case of MLB, the model results suggest that having a Native American is also driving lower box office revenues over time.  The effect is bit higher in MLB with the trend being a loss of about $2.6 per year.

Despite the limitations inherent to our analyses, the consistency between the NFL and MLB findings is in accordance with a trend of growing opposition to these mascots.  However, we do acknowledge that our claim of a trend of “growing opposition” is based largely on anecdotal data such as retirements of prominent Native American mascots in college sports, journalists dropping the use of “offensive” nicknames and politicians beginning to weigh in on the issue.  Our results imply that fans are also becoming less enthusiastic about these mascots.

To be blunt, the implication is that the trends suggest that keeping a Native American mascot is reducing financial performance and harming team brand equity.

Live Animal Mascots in College Football

Bulldog MoneyWe also had a brief correspondence from a reader asking if we had ever investigated the financial consequences of “Live” animal mascots.  At the time of this question, we were basically unaware of the controversy surrounding the use of this type of mascot.  We were familiar with some of the more spectacular live mascots such as Bevo, Uga and Ralphie.  In hindsight, it does make sense that animal rights activists would be concerned about the welfare of these living symbols.

For this study, we used publically available data on college football team revenues.  We decided to restrict the analysis to football because many of the most notable animal mascots only appear during the football season.  But, we should note that we do not know if Colorado has ever run Ralphie across the basketball floor.

For this analysis, we used relative revenue as our dependent variable.  This was computed by dividing each team’s self-reported football revenues by the overall average for each season.  Relative revenue was modeled as a function of AQ (automatic qualifying conference) status, winning percentage, level of bowl game participation, local population and student body size.  We included a dummy variable for a “live mascot” and an interaction variable between AQ status and having a live mascot.  The interaction is included to account for the possibility that live mascot effectiveness varies across level of competition.

Mascot Type

Estimate

Standard Error

t-Value

Pr>|t|

Live Mascot

0.018

0.072

0.25

0.1015

Live Mascot*AQ

0.369

0.086

4.28

<.0001

In order to interpret the preceding results we need to remember that the statistical results were generated using relative revenues as the dependent variable.  Again, these coefficients are easily translated into dollars.  In 2010, average revenues across the FBS schools were about $23 million and about $35 million for members of the AQ conferences.  The model therefore suggests that on average an AQ member school with a live animal mascot generates about $8.5 million in incremental revenue!  However, the net effect for a non-AQ school is negligible.

This is an amazing number, but it does have some logic, as live animals may be exceptional community builders.  In the case of mascots like Reveille or UGA it is almost as if the entire student body and alumni base co-owns a dog.  And in the case of Bevo or Ralphie, it is hard to imagine a more spectacular halftime display.

These results highlight the tough battle that PETA and other animal rights organizations fight.  Unlike the Native American mascots, the data suggests that live mascots drive incremental revenue and brand equity.

Conclusion

The preceding analyses will hopefully generate interest and debate.  From our perspective, this type of work is a lot of fun.  We are able to investigate the topic using data and analytical techniques without having to endure a multi-year journal review process.  As we have noted, our work does include assumptions but we have tried to be as transparent as possible.

In our minds, what we have produced are data driven and unbiased analyses of how mascots affect brand equity and revenues.  Could we extend the models?  Absolutely.  We could find more data, we could use more categories of mascots, and we could use a more sophisticated statistical model.  But for now we have put a stake in the ground, and have hopefully provided a basis for extending the conversations surround these two mascot controversies.

Mike Lewis & Manish Tripathi, Emory University 2013.

NFL Fans at the “Twitter Water Cooler”

Note: This was originally published on September 4, 2013

The start of the NFL regular season is upon us.  In cities across America, NFL fans will engage in the practice of “Monday Morning Quarterbacking,” giving an analysis of their team’s performance on the previous day.  Some fans will of course be delighted after a team victory, while others will be dejected after a crushing defeat.  We decided it would be interesting to analyze how the thirty-two NFL fan bases felt the day(s) after their teams played in the regular season.  While we don’t have the ability to observe the millions of “water cooler” conversations that occur every week, we do have access to millions of Twitter conversations about NFL teams.

We used Twitter data to describe fan base reactions to team wins and losses during the sixteen-game 2012 NFL Regular season.  Our process for data collection can be illustrated with an example using the Buffalo Bills.  Imagine that the Bills played a game on a Sunday.  We recorded whether the Bills won or lost the game.  We then collected all tweets in the Buffalo area that mentioned the words “Buffalo Bills”, “Bills”, or other very frequent terms used to describe the team.  We collected the tweets from Monday (one day after the game), Tuesday (two days after the game), and Wednesday (three days after the game).  We then analyzed each tweet and characterized its content as positive or negative.  Next, we calculated the overall sentiment (roughly the indexed ratio of positive to negative tweets) of the Buffalo Bills related tweets for each of the three days.  We repeated this process for all thirty-two teams, and for all regular season games*.

The chart above displays the average sentiment of fans both after wins and losses.  The chart is based on data from the regular season for all thirty-two teams.  It is interesting to note that by three days after a win or loss, fans on average seem to either come down from their win “high” or recover significantly from their loss “low”.  While the chart above looks at all NFL fan bases in aggregate, we thought it would be interesting to classify each NFL fan base on the following dimensions**:

1)   Happiness After a Win (Highest ratio of positive to negative tweets after win)

2)   Sadness After a Loss (Lowest ratio of positive to negative tweets after loss)

3)   Stability (Least difference in positive to negative tweet ratio between after wins and after losses)

1) Happiness After a Win

The New Orleans Saints’ fans seemed to have just over a 9:1 positive to negative tweet ratio in the two days after the team won a game during 2012 regular season.  We believe that rankings on any of these dimensions are most likely driven by fan expectations (which is in part a function of past and current performance) and by the “expressiveness” of fans.  Since we are presenting descriptive statistics, and not explicitly modeling these drivers, it is tough to make a definitive statement as to why we see this particular order of teams.  Although, is anyone really surprised to see Cleveland or Oakland in the top 5?

2) Sadness After a Loss

The Pittsburgh Steelers’ fans seemed to take losing really badly in the 2012 season.  This could be because of fan expectations.  The Steelers finished 12-4 in 2011, but failed to make the post-season in 2012.  Early losses to the Raiders and Titans produced especially negative Twitter reaction, as did late season losses to the Browns and Bengals.

3) Stability

We measured “stability” by looking at the difference between average sentiment after wins and average sentiment after losses.  Dallas Cowboys’ fans seemed to never get too negative after losses, nor were they tremendously positive after wins.  Colts’ fans were even more understanding after a loss, but more positive on average than Cowboys’ fans after a win.  This could be due to the Colts being a young team that did not have high expectations going into the 2012 season.    The Atlanta Falcons only lost three times during the regular season, and the last loss was meaningless, as the Falcons had already secured home-field advantage throughout the playoffs.  Thus, there was very little negative reaction to the last loss.  The Philadelphia Eagles’ fan base is an interesting story.  It may be surprising to many to see them in the list of “stable” fans.   A better moniker for these fans in 2012 might be “resigned”.  It seems that as Philly began to lose more games, fans started to look forward to the next season, and a new head coach.  A majority of the fan tweets after a game were about changes for the next season, and not about the most recent loss.

The Oakland Raiders’ fan base best resembled “Dr. Jekyll and Mr. Hyde” during the 2012 season.  Fans were extremely happy when the team won, and terribly negative after a loss.  The Raiders were the only team in the top 5 of the “Happy” and “Sad” fan rankings.

 

Mike Lewis & Manish Tripathi, Emory University 2013.

*There are, of course, several caveats regarding this study.  First, while we only used tweets from the team’s geographic market, there could always be fans of other teams who may have tweeted about the local team.  Similarly, there are fans of the team that do not live in the local market, whose tweets would have been excluded.  Second, though we used terms that were associated/descriptive of the team, there are tweets related to the team that we undoubtedly excluded because they did not mention the terms we were looking for.  Third, the volume of tweets is not the same for each team.  We are confident, however, that a minimum threshold was met for each day, such that the sentiment score was not heavily influenced by a small number of tweets.  Fourth, this study is only over one year; it would be beneficial to perform a multi-year study.  Finally, there was one game in the 2012 season between the San Francisco 49ers and St Louis Rams that ended in a tie.  We have excluded that game from this analysis.  The Twitter data was collected using Topsy Pro Analytics.

**We computed each of these metrics using one day after, an average of one and two days after, and an average of the first three days after.  Since the rankings were fairly robust across these specifications, we only report the average of one and two days after the game.

St. Louis Tops Rankings of Three Team Cities, Tampa Bay is Last

Our city ranking series continues today with a look at cities with three professional sports teams.  These markets tend to be a bit on the smaller side, but many have significant sports histories.  We also fully admit that we struggled a bit with how to classify several of these markets.  For example, what is the city of Milwaukee?  Is Milwaukee a two sport town with NBA and MLB franchises or should we include the Packers and call it a three sport town?  Having lived in Chicago, it always seemed like all the Wisconsin teams should be lumped together.  Toronto was another decision.  Until now, we have only considered US cities, and avoided one professional team Canadian markets such as Calgary and Edmonton.  So before the complaints begin, please realize that we have made some assumptions about markets.

The table on the right provides our ranking of the eight markets with three professional teams.  According to the data, St. Louis is the best of these markets.  Professor Lewis used to live in St. Louis and the first place ranking was a bit of a surprise to him.  While the Cardinals have an amazing following,  Lewis’ sense was that the Rams and Blues only had average fan bases. The Cardinals do have an exceptional fan base ranking 4th in MLB in both fan equity and social media equity.  The Blues have an above average fan base ranking 14th in the NHL.  The Rams do struggle with a fan equity ranking of 22th in the NFL.  So it really is the Cardinals that elevate St. Louis to the top of the list.

Following St. Louis, we have Toronto ranked 2nd, Milwaukee 3rd and Pittsburgh 4th.  Frankly, we would have predicted Pittsburgh would rank higher.  The issue is that our fan equity metric is based on a “revenue premium” model, and the Steelers don’t seem to price nearly as high as they could.  But, this was a close competition.  Toronto has the best NHL fan base and the Packers and Steelers have devoted followings.

At the bottom of the list we have Tampa Bay.  The Lightning ranked 18th in NHL fan equity.  The Bucs ranked 29th in the NFL and Rays ranked 22nd in MLB.  On a side note, the Atlanta ranking should put to rest any complaints about the Braves relocating.  The Braves have delivered phenomenal quality and have only gained an average fan following.  Add in a history that includes players like Hank Aaron and Dale Murphy, and you would expect that the Braves would have a monster following.  Our expectation is that the move to Cobb County and the building of a mixed use development around the stadium should lead to a stronger fan base in the near future.

Mike Lewis & Manish Tripathi, Emory University 2013.

What if the Heisman Trophy Really Was A Popularity Contest?

Ballots for the Heisman Trophy were due yesterday.  Ostensibly, the Heisman Trophy “annually recognizes the outstanding college football player whose performance best exhibits the pursuit of excellence with integrity”.  However, many have argued throughout the years that the Heisman is essentially a large popularity contest.  This view is supported by the millions of dollars annually spent by universities on publicity campaigns for their Heisman candidates.  There are 928 voters for the Heisman Trophy.  This includes members of the media, former winners, and 1 “fan vote” that represents the public at large.  We were curious to see what would happen if the general public was completely responsible for determining the winner of the Heisman Trophy.  As with past studies, we decided to use Twitter as a proxy for the views of the public.  Below, we present our methodology and results.

The first thing to consider is how does one define “Popularity” on Twitter.  Often, studies will use the volume/number of mentions on Twitter as a proxy for popularity.  However, this measure does not account for sentiment (positive, negative, or neutral), which could be important in the decision to vote for someone.  So, we constructed a daily “popularity” measure that is the product of the volume of tweets mentioning a candidate and the average sentiment of those tweets (Note: we tried several specifications of the “popularity” measure, but the rankings were robust).

Once we had a method for determining popularity, we decided to look at the six Heisman Trophy finalists: Johnny Manziel, Jameis Winston, A.J. McCarron, Tre Mason, Jordan Lynch, and Andre Williams.  The pie chart on the left looks at the sum of the popularity measure for each candidate over the entire season (mid-August to Dec 9th).  Johnny Manziel is by far the leader of the pack.  This could potentially be attributed to the stellar start of his season, as well as his huge following.  Heisman-favorite Jameis Winston is in second place, and A.J. McCarron is third.  It’s incredible that Manziel leads Winston by more than a 2:1 margin.  We realize that Heisman voters mark their ballots for 1st, 2nd, and 3rd place, and we are simply looking at most “popular”.

We performed a similar analysis, looking at only the last month and looking at only the last week.   It’s remarkable to see the variation in “popularity” over time.  Tre Mason had a relative 5% popularity if you look at the full season, but 11% over the last month, and 24% over the last week.  In the analysis of popularity over the last week, James Winston barely edges out Manziel for 1st place.  To better understand the factors behind these movements in popularity, we would have to perform content analysis on the tweets to determine what topics were being discussed with respect to these athletes; that is left for a future study.

It is interesting to note that in the their final straw poll, Heisman Pundit has the following ranking: 1) Winston, 2) Lynch, 3) Manziel, 4) Mason, and 5) McCarron.  The “popularity” measure for over the last week gives the ranking: 1) Winston, 2) Manziel, 3) Mason, 4) McCarron, and 5) Lynch.  Jordan Lynch is the only player of these top 5 that plays for a “Non-AQ” school (Northern Illinois).  Perhaps Lynch in second place is evidence that voters look at performance on the field, and not just popularity, however if Heisman Pundit’s straw poll is correct, it seems a lot can be explained by recent popularity.

Mike Lewis & Manish Tripathi, Emory University 2013.