Value determination in the secondary ticket market: a quantitative analysis of the NFL playoffs.

Drayer, Joris ; Shapiro, Stephen L.

Introduction

Ticket prices in the live entertainment industry, including team sports, are generally priced at a suboptimal level. According to Volpano (2003), the evidence of sub-optimal pricing is demonstrated by a combination of empty seats at venues and the existence of ticket brokers. There exists an ambiguous area in ticket pricing where team officials want to charge enough so that they maximize revenue but not too much so that fans stop attending games. The balance between optimal ticket pricing and providing opportunities for a wide variety of fans to attend games is challenging within the realm of the sport industry.

Teams use a variety of methods to determine ticket prices. For some teams, the pricing strategy is simply a result of the revenue demands of ownership ("Price is Right," 2004). However, in a study conducted on NFL ticket prices, Reese and Mittelstaedt (2001) found that ticket prices were based on a certain set of criteria. Those criteria, in order of importance, were: team performance, revenue needs of the organization, public relations issues, toleration of the market regarding price increases, fan identification, and the average league ticket price. Clearly, each organization has to consider unique situational factors that may affect the order of importance regarding these criteria.

Pricing tickets appropriately is a critical decision for sport properties. Steve Swetoha, Senior Vice President of Ticket Sales for the NBA's Charlotte Bobcats, claimed that "the difference between black and red with ticket prices can be about as little as 50 cents" ("Price is Right," 2004, [paragraph] 31). In the article titled "Price is Right," it is stated that "most ticketing executives fear the day when their ticket prices are considered too pricey for the market" ("Five minutes for slashing," [paragraph] 1). This can be due to a variety of factors such as team performance, economic conditions, and market conditions. As a result, the popular opinion is that event promoters/organizers intentionally underprice tickets (Courty, 2003). Based on the existing literature, Courty concluded that underpricing may be due to uncertainty of sales, the value customers place on fair treatment, the idea that consistency in pricing is necessary to attract loyal fans, and the notion that a fuller stadium enhances the fan experience.

It appears that ticket pricing is influenced by economic and organizational factors combined with the perceived value that fans place on individual tickets. The majority of research has focused on the factors that influence ticket prices in the primary market (Reese & Mittelstaedt, 2001; Rishe & Mondello, 2003, 2004). However, the literature on the factors that influence value, or the price that fans are willing to pay in the secondary market, is underdeveloped. The purpose of the current study is to examine the potential factors that impact ticket value in the secondary market, where prices fluctuate based on fans' willingness to pay. An analysis of traditional price determinants, combined with factors that influence online ticket auctions, will provide a more complete understanding of the true value placed on tickets by fans. Closely aligning ticket prices with demand conditions will help teams to maximize their revenue from ticket sales and perhaps decrease the presence of sellers in the secondary market, who have long been associated with unethical practices including price gouging, selling fraudulent tickets, and the utilization of computer programs designed to buy blocks of tickets in the primary market (Drayer, Stotlar, & Irwin, 2008).

Review of Related Literature

The following review of previous literature is broken into two main topic areas. First, ticket price determinants and pricing inefficiencies were reviewed to provide an understanding of the traditional factors that influence ticket prices in the primary market and the gap that exists between pricing in the primary market and perceived value in the secondary market. Second, research on the secondary ticket market is examined. This literature provided the foundation for the development of a model to examine the factors that influence perceived value of NFL playoff tickets in the secondary market.

Ticket Price Determinants and Pricing Inefficiencies As mentioned previously, ticket pricing studies have focused almost exclusively on the primary market. According to Rishe and Mondello (2004), a variety of factors impact primary ticket prices, including previous year's success, income level, and playing in a new stadium. The authors also found that population size was positively correlated with ticket prices in all leagues with the exception of the NFL where sellouts are common regardless of market size. However, Rishe and Mondello (2003) acknowledged that the process of price determination will vary from team to team and league to league making it difficult to standardize the procedure.

Ultimately, Rishe and Mondello (2003) suggested that "only changes in the demand for a given team's tickets will cause that team's ticket prices to change from season to season" (p. 73). They also stated consumer tastes/preferences and consumer expectations are the two factors that shift consumer demand. However, the nature of the sports industry is that demand often fluctuates from game to game or even day to day (Borland & MacDonald, 2003; Rascher, McEvoy, Nagel & Brown, 2007). Given these frequent changes in demand, ticket prices in professional sports do not accurately reflect the demand conditions. Instead, they most often reflect previous year's performance and revenue needs of the organization (Reese & Middlestaedt, 2001). While previous year's performance will certainly influence demand conditions, it is by no means the only factor that will do so. Since these prices are set long before the season starts, they cannot accurately reflect the demand conditions throughout the season. The optimal way to measure demand is by looking at prices that fans are willing to pay in the secondary market and not what is on the face of the ticket (Drayer et al., 2008).

The limitation in the aforementioned price determination studies is the assumption that the face value of the ticket is the same as the perceived value for consumers and therefore reflects the demand conditions for those tickets. However, price is only an objective characteristic that influences consumer perceptions of a product (Zeithaml, 1988). According to Zeithaml, "Objective monetary price is frequently not the price encoded by consumer" (p. 10). Indeed, some consumers may see the exact price, some may perceive the product as "expensive" or "cheap," and some may not code the price at all. Further, according to Dodds, Monroe, and Grewal (1991), "[p]erceptions of the same price stimulus may vary across consumers and, for one consumer, across products, purchase situations, and time" (p. 308). Applied to tickets, fans may differ in their perceptions of actual value of each ticket. For each individual consumer, this perceived value may change based on factors such as how well the team is doing, the opponent, the status of individual players, the quality of the seat, the time of the season (pre-season, regular season, or playoffs), or even the weather. Clearly, in a sport context, these factors change more often than from one season to the next.

Secondary Ticket Market

Boyd and Boyd (1998) noted that "the existence of scalpers and 'ticket agents' who sell game tickets at prices significantly in excess of face value could suggest that teams set prices too low" (p. 169). As a result, some teams, such as the St. Louis Cardinals, are setting ticket prices based loosely on what tickets have sold for in the secondary market (Fisher, 2005; Johnson, 2005). This realization that face value is not always the same as the ticket value for a fan is an important concept that the secondary market illustrates well. In academia, the research on ticketing reflects the determinants of the face value of tickets (Reese & Middlestaedt, 2001; Rishe & Mondello, 2003, 2004). However, as ticket prices are typically set using a cost-oriented strategy (Howard & Crompton, 2004; Kotler, 2003), they reflect little besides the revenue needs of the organization, which is consistent with the findings of Reese and Middlestaedt.

In order to fully shift to a demand-oriented strategy, research must be conducted to determine which factors predict higher perceived value by fans. By looking at the activity in the secondary market, this study is the first "value determination" investigation, which looks at the prices fans are willing to pay and the factors that influence their willingness to pay. This study used the sell price of individual eBay transactions as the dependant variable instead of the average ticket prices set by teams in the primary market. The sell price of eBay transactions is most often the result of a 3-, 7-, or 10day auction and represents the highest price a group of consumers is willing to pay within a particular online environment. Therefore, we believe that these prices most accurately reflect the demand conditions for the games selected. As Rishe and Mondello (2003) stated, "the more intense the demand, the more willing fans will be to pay higher prices" (p. 73).

Methods

The Secondary Pricing Model

A regression model was developed to understand the factors that affect the variability in NFL playoff ticket prices within the secondary market. Due to the nonexistence of a model that examined secondary tickets prices in the NFL, the current model was created through an evaluation of previous literature in the areas of ticket price determinants in the primary market (Reese & Middlestaedt, 2001; Rishe & Mondello, 2003, 2004) and pricing through online auctions (Lucking-Reiley, Bryan, Prasad, & Reeves, 2007). NFL playoff tickets were chosen for two reasons. First, the demand for NFL playoff tickets is extremely high, which increases the likelihood of these tickets being resold in the secondary market. In the current study, a total of 1,725 tickets were transacted on eBay for the first three rounds of the 2007 NFL playoffs. Second, the researchers had access and availability to all 2007 "By looking at the activity in the secondary market, this study is the first 'value determination' investigation, which looks at the prices fans are willing to pay and the factors that influence their willingness to pay."

NFL playoff tickets sold on eBay. This provided a complete sample of ticket transactions from a particular online auction source for the first three rounds of the 2007 post-season. A total of 17 explanatory variables were used to investigate the determinants of secondary market ticket value in the NFL.

This model is unique compared to previous models that have examined the factors that affect ticket pricing in professional sport due to the focus within the secondary market and the use of online auctions. The following section defines the dependent variable and each of the explanatory variables that make up the model.

Dependent Variable

Secondary Ticket Price at Auction (TIXAUC)--The final price that an NFL ticket was sold through online auction for all 2007 NFL playoff games (excluding the Super Bowl).

Explanatory Variables

1. Transactions (TRANS)--The total number of transactions on eBay for each game.

2. Per Capita Income (INCOME)--Per Capita income was used to measure potential purchasing power in a team's area.

3. Population (POP)--The population, as reported by the U.S. Census Bureau, of the city where the home team plays.

4. Day of the Week (GDAY)--The day that a specific playoff game was held.

5. Playoff Round (ROUND)--The NFL playoffs have three preliminary rounds: Wild Card round, Divisional round, and Conference Championship round.

6. Face Value of the Ticket (TIXFACE)--The face value price of an individual ticket within the primary market.

7. Current Season Winning Percentage (CURWIN)--The winning percentage at the time of the game for the home team during the current season, including the playoffs.

8. Previous Winning Percentage (PREVWIN)--The winning percentage for the home team in the previous season.

9. Division (DIV)--A categorical variable that establishes whether or not two teams are in the same division.

10. Wagering line on the game (SPREAD)--A variable that represents the betting line for a specific game. The spread indicates the perceived competitiveness of the game.

11. Game Certainty (CERTAIN)--A categorical variable that establishes whether or not a game is certain at the time of an auction. Tickets can be bought and auctioned for NFL playoff games prior to a team qualifying to participate in a specific game.

eBay Auction Explanatory Variables

12. Total Number of Bids (BIDS)--Represents the number of bids that were made for a ticket and/or group of tickets in a specific auction.

13. Days Before the Game (DAYSBEF)--The number of days prior to the game that a specific auction for a ticket and/or group of tickets began.

14. Auction Format (FORMAT)--A categorical variable that establishes the type of auction being used. eBay allows sellers to choose certain formats that provide opportunities for auctions and/or direct sales.

15. Seller Positive Feedback (PFB)--The number of positive feedback ratings the seller has received in the past. eBay provides opportunities for buyers to rate the quality of a seller in an effort to help future buyers.

16. Seller Negative Feedback (NFB)--The number of negative feedback ratings a seller has received in the past.

17. Auction Length (LENGTH)--The total number of days that the auction lasted.

Sample

Data were collected for all NFL post-season games during the first three rounds of the 2007 playoffs from November, 2007 through January, 2008. Data from the Super Bowl was not collected because the unique nature of this particular game would potentially skew the results. A total of 750 completed eBay auctions were collected during this time period (n=750). Some of the auctions provided the sale of multiple tickets. A total of 1,725 individual tickets were sold on eBay during this time period.

Data on the variables in the pricing model were collected from a variety of sources. Team success and spread data were collected from ESPN.com. Population and per capita income data were both collected from the U.S. Census Bureau. The data for all auction-related variables were collected from www.ebay.com. The individual auctions provided detailed data regarding pricing of tickets and the auction format and guidelines. In addition, the face value of each individual ticket was collected from www.ebay.com as well as official team ticketing websites.

Statistical Design

A standard multiple linear regression equation was initially developed to examine the relationship between the 17 variables identified in previous research and NFL playoff ticket value in the secondary market through online auction:

[TIXAUC.sub.j] = [A.sub.0] + [B.sub.1] [SPREAD.sub.j] + [B.sub.2] [INCOME.sub.j] + [B.sub.3] [POP.sub.j] + [B.sub.4] GDAY + [B.sub.5] ROUND + [B.sub.6] TIXFACE + [B.sub.7] [CURWIN.sub.j] + [B.sub.8] [PREVWIN.sub.j] + [B.sub.9] DIV + [B.sub.10] BIDS + [B.sub.11] DAYSBEF + [B.sub.12] CERTAIN + [B.sub.13] FORMAT + [B.sub.14] PFB + [B.sub.15] NFB + [B.sub.16] LENGTH + [B.sub.17] TRANS

The Bs represent the Beta coefficients assigned to each of the independent variables during the regression analysis. The home team is represented by j in the equation.

Due to the exploratory nature of the study and the lack of existing theory on determinants of ticket value in the secondary market, a long list of independent variables were included in the model. Seventeen variables were included in the initial regression equation in an effort to include variables from previous literature along with factors that have the potential to influence ticket value in the secondary market. However, if interpretation is the objective and the number of variables is greater than 10 to 15, the model should be reduced to make the computations more feasible (Cox & Snell, 1974). In an attempt to develop a model from a considerably large set of explanatory variables that were selected from multiple studies, an all-subsets regression was used as a variable selection procedure to establish the best fitting model. It is best to have the most parsimonious model that explains the variation in the dependent variable while reducing multicollinearity (Sen & Srivastava, 1990). A predictive model was used due to the absence of an existing theoretical model. The model development in the current study is exploratory in nature and can be cross-validated in future studies using explanatory regression procedures.

An all-subsets regression was chosen because it allows for the examination of all potential regression equations from the chosen variables (Sen & Srivastava, 1990). A measure of fit is provided for each equation, and the researchers can examine the model that provides the best fit from both a theoretical and statistical standpoint. The all-subsets regression was chosen over the stepwise variable selection procedure because stepwise regression does not take into consideration high levels of correlation between independent variables and independent variables are chosen exclusively on mathematical computation rather than logical interpretation (Cohen, Cohen, West, & Aiken, 2003).

The [R.sup.2] statistic and the Mallow's Cp statistic were used to determine the most appropriate model from the all-subsets regression analysis. According to Draper and Smith (1998), the [R.sup.2] statistic is a measure of explained variance within the model. The best fitting model will maximize [R.sup.2] while eliminating variables that explain little or no variance in the dependent variable. The Mallow's Cp statistic can be used in subset regression as a criterion, so that the parameters estimated in the model are not larger than the number of observations. A Mallows Cp statistic that is close to the number of predictors in the model +1 will reflect a lack of bias regarding standard error within the model (Mallows, 1973). The [R.sup.2] and Mallows Cp statistics, combined with logical interpretation of the variables included in the equation, provided the most appropriate exploratory model that explains the variance in secondary ticket value of NFL playoff tickets sold through online auction.

All of the multiple linear regression assumptions (Linearity, Independence, Normality, and Equality of Variances) were examined. Descriptive statistics, residual plots, and statistical tests for normality and equality of variances showed that none of the assumptions were violated in the regression equation. In addition, potential multicollinearity issues within the model were examined. Both variance inflation factors and tolerance statistics were examined for multicollinearity issues. The results suggested there were no multicollinearity issues in the regression equation used in the analysis. A significance level of 0.05 was established a priori in analyzing the regression model and related variable correlations.

Results

Table 1 provides descriptive data for all continuous variables within the regression model. The results showed that the average value of an NFL playoff ticket sold through online auction was $257.75. The average face value of an NFL playoff ticket during this time period was $124.26. These findings represent a 107% increase in ticket price from the primary to secondary markets. The average spread for the playoff games was -7.73 (approximately a touchdown). For most of the playoff games, the spread was generally small, which represents the anticipation of a close game. All home teams were favored in the sample with the exception of one game. All teams in the sample had high winning percentages for both the current season (M = .781) and the previous season (M = .608). This was expected because the sample included only playoff teams. The high win percentages indicated a larger demand for these games. Other important descriptive findings included average population and per capita income. The average population in the sample was 672,734 with a mean per capita income of $21,993.

Table 2 provides correlational data for each of the independent variables in the study and the final ticket price at auction. Fourteen of the 17 independent variables were significantly correlated to ticket value at the 0.05 level. The independent variables that correlated most strongly with ticket value were playoff round (r= .563), city population (r= -.443), face value of the ticket (r=.341), and per capita income (r= .338). It is important to note that while it is anticipated that the majority of ticket buyers resided in the home team's city, eBay did not allow for the tracking of buyer location. So, while city population and per capita income were strongly correlated variables, interpretation of their significance should consider the possibility that buyers may have been located outside of the home team's city.

All-Subsets Regression

Based on the all possible subsets regression results, a 10 variable representation was selected as the most appropriate model for explaining the variability in NFL play-off ticket value within the secondary market. The 10 variable model was found to be significant (F= 129.56(10,740), p < .01) explaining 63.6% of the variability in playoff ticket value through online auction. The 10 variable model was selected for multiple reasons. First, the [R.sup.2] statistic had considerable increases as variables were entered into the model until 10 variables were reached. After 10 variables, the [R.sup.2] only increased slightly due to the fact that the original variables that were not selected in the model contributed only marginally to the explained variance. Of all the 10 variable models, the selected model had the strongest [R.sup.2] statistic. Second, the [C.sub.p] Mallows statistic was examined in addition to [R.sup.2]. The 10 variable model selected had a [C.sub.p] value of 11.766, which was the closest to the recommended value (number of predictors + 1). This result indicated a lack of bias in the model compared to other potential models. Finally, logical interpretation of the selected variables provided evidence of a satisfactory model. None of the variables deleted from the original list of independent variables was presumed to have a considerable effect on ticket value in the secondary market.

As displayed in Table 3, the variables included in the selected model were population (POP) (p = .008), total number of transactions per game (TRANS) (p < .01), face value of the ticket (TIXFACE) (p < .01), winning percentage from the previous year (PREVWIN) (p < .01), day of the game (GDAY) (p < .01), number of days before the game (DAYSBEF) (p <.01), per capita income (INCOME) (p <.01), playoff round (ROUND) (p < .01), total number of bids (BIDS) (p < .01), and current winning percentage (CURWIN) (p < .01. The majority of variables had a positive relationship to secondary ticket value. Only population (POP) had a negative parameter estimate.

Population, total number of transactions, and the face value of the ticket appear to have the greatest influence on secondary market ticket value in the model. A total of eight variables in the selected model can be attributed to the game environment: current and previous winning percentage, population and income of the home city, face value of the ticket, day of the game, round of the playoffs, and number of days before the game that the ticket was sold. Total number of bids was the only variable that can be exclusively attributed to the online auction environment that the tickets were sold in. In addition, total number of transactions can be attributed to both game environment and the online environment. These results suggest that a larger portion of the selected model is influenced by game-related elements. However, two elements provided evidence of an influence from the online auction environment as well.

Many online auction environment variables from the original list were not selected in the model, including buyer feedback and the length and format of the auction. There were also game environment variables that were not selected (game spread, game certainty, and if the game involved divisional opponents).

Discussion

The purpose of the current study was to develop a model to measure the determinants of NFL playoff ticket value in the secondary market. The findings in this study illustrate the importance of establishing a model for understanding secondary ticket value in an online auction environment. A conceptual model of secondary ticket value determinants is non-existent. Therefore, an all-subsets regression was developed in an exploratory capacity to reduce a large number of potential predictors of ticket value into a more stream-lined model for future investigation. Replication is necessary to cross-validate a selection procedure such as all-subsets regression (Snyder, 1991). However, the development and analysis of an initial model is a critical step in the process of understanding ticket value in the secondary market. The current analysis provides a regression model for future explanatory cross-validation efforts such as OLS regression.

The all-subsets regression model produced several findings that were anticipated. For example, it was anticipated that income would be a significant factor in the final sale price of tickets (Rishe & Mondello, 2003, 2004). Significance was also expected with regard to the round of the playoffs. As teams get closer to winning a championship, fans increasingly want to see their team win and are willing to pay higher ticket prices. Not surprisingly, face value was also a significant predictor of secondary ticket value. Face value, while not reflective of demand conditions, typically reflects the quality of the seat in terms of location and the round of the playoffs. It is not surprising that fans are willing to pay higher prices for better seats and for games further into the playoffs. Finally, total number of bids was also found to be related to higher tickets prices. More people bidding against each other indicated a higher demand for that game. Higher demand typically leads to higher prices (Azjen & Peterson, 1988).

However, there were several other intriguing findings. Population size was found to be significant and negatively correlated to ticket value. This finding was consistent with that of Rishe and Mondello (2003, 2004). The authors suggested lack of competition as a justification for teams charging higher prices in smaller markets. The lack of competition appears to affect ticket value in the secondary market as well. It is important to note that Green Bay, the smallest market in the NFL, hosted the NFC Championship game, which may have influenced the data for this particular playoff year. As mentioned previously, round of the playoffs was highly correlated with secondary market sell price.

The results also indicated that the total number of transactions for each game was among the strongest predictors of high ticket prices in the secondary ticket market. In the primary market, supply is fixed as teams can only sell one ticket for each seat in the stadium. However, once a game is sold out in the primary market, the secondary market supply does fluctuate based on the number of ticket holders willing to resell their tickets. Generally, as supply goes up, price will go down (Azjen & Peterson, 1988). However, these results suggest the opposite. While the data does not allow for cause-and-effect statements between the variables, it seems reasonable to assume that the increased supply does not cause higher prices. On the other hand, it is more plausible to assume that as ticket holders realize that fans are willing to pay high prices for tickets in the secondary market, they become more likely to sell, thus increasing supply. In other words, each fan places a specific value on their tickets. As soon as secondary market prices exceed that figure, a ticket holder will be more likely to resell. Of course, the price threshold is different for each fan.

There were also interesting findings related to the concept of outcome certainty. Home team winning percentage in the current season and home team winning percentage in the previous season were both included in the final model. This finding suggests that fans prefer to pay more money to watch a better team that is more likely to have success. Ultimately, it seems that fans are basking in the reflected glory of their favorite team's success by placing a higher value on attending their next game (Cialdini et al., 1976).

The selected 10-variable model explained 63.6% of the variability in playoff ticket value through online auction. It is the belief of these authors that some of the unexplained variance is the result of the situational factors associated with each game. Many of these factors have been accounted for in the 10-variable model; however, there are many other game-specific factors that are difficult to quantify. For example, in the case of this playoff year, the New England Patriots were not only successful in their run to the Super Bowl; they were also pursuing an undefeated season which may have impacted the demand for tickets.

Pricing (In)efficiency

As mentioned in the Results section, face value per ticket was less than half of the actual sale price in the online auction, on average. This is the clearest indication of pricing inefficiency in the primary market. Demand for NFL playoff tickets remains strong at higher price points as evidenced by the large secondary market. A large increase in primary market prices will certainly have its influence on demand. However, it appears that tickets can be more effectively priced based on the large gap between primary and secondary ticket prices. A closer look at the descriptive statistics indicates that the biggest inefficiency occurs with regard to round of the playoffs. In the first round, average sale price was $38.79 higher than average face value. In the second and third round the difference between sale price and face value increased to $105.92 and $306.21, respectively. It appears that teams are missing out on a substantial revenue stream by inadequately raising prices between each round of the playoffs. It is necessary to continue to examine the factors influencing secondary market sell price to gain a detailed understanding of what factors teams should consider when pricing their tickets.

Among the teams that participated in the playoffs in the 2007-2008 season, the Pittsburgh Steelers, who hosted a first round game, were the team that appears to price their tickets most accurately. Average face value for Steelers tickets sold on eBay was $161.32, compared to $118.71 for the other three teams to host first round games. However, the eBay sell price was $166.32 for the Steelers and $163.84 for the other three teams. The Steelers appeared to price their tickets right under what the market was willing to bear and maximized their revenue from ticket sales while the other three teams missed out on over $40 of revenue per ticket. All four games were sold out.

Limitations and Suggestions for Future Research

As mentioned previously, this research covered the playoff games for only one NFL season. The unique game factors associated with each playoff game in each playoff year will vary considerably. A future study could track this data over a period of several years to see if these factors hold constant over time. Additionally, as this study only included playoff games, a future study could also examine the factors influencing perceived value during the regular season and across other sport leagues. In particular, the use of this specific sample may minimize the significance of winning percentage, as the importance of winning percentage is likely diminished in a playoff scenario.

This study included transactions only on www.ebay.com. Each secondary distribution channel, either online or offline, has a variety of factors that may influence prices. For example, Elfenbein (2006) suggested that the legal status of ticket resale may impact the online secondary market. The author reported fewer transactions and higher secondary prices in states with stricter regulations. Anti-scalping laws certainly impact ticket brokers as well. Based on the specifics of each location and distribution channel, prices may vary across a variety of secondary ticketing websites and local or regional ticket brokers. Future research could compare prices among these ticket outlets.

References

Azjen, I., & Peterson, G. L. (1988). Contingent value measurement: The price of everything and the value of nothing? In G. L. Peterson, B. L. Driver, & R. Gregory (Eds.), Amenity resource valuation (pp. 65-76). State College, PA: Venture Publishing, Inc.

Borland, J., & MacDonald, R. (2003). Demand for sport. Oxford Review of Economic Policy, 19(4), 478-502.

Boyd, D. W., & Boyd, L. A. (1998). The home field advantage: Implications for the pricing of tickets to professional team sporting events [Electronic Version]. Journal of Economics and Finance, 22, 169-179.

Cialdini, R. B., Borden, R. J., Thorne, A., Walker, M. R., Freeman, S., & Sloan, L. R. (1976). Basking in reflected glory: Three (football) field studies. Journal of Personality and Social Psychology, 34(3), 406-415.

Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis for the behavioral sciences (3rd ed.). Mahwah, NJ: Lawrence Erlbaum Associates.

Courty, P. (2003). Some economics of ticket resale [Electronic Version]. Journal of Economic Perspectives, 17, 85-97.

Cox, D. R., & Snell, E. J. (1974). The choice of variables in observational studies. Applied Statistics, 23(1), 51-59.

Dodds, W. B., Monroe, K. B., & Grewal, D. (1991), Effects of price, brand, and store information on buyers' product evaluations. Journal of Marketing Research, 28, 307-319.

Draper, N. R., & Smith, H. (1998). Applied regression analysis (3rd ed.). New York: John Wiley & Sons, Inc.

Drayer, J., Stotlar, D. K., & Irwin, R. L. (2008). Tradition vs. trend: A case study of team response to the secondary ticket market. Sport Marketing Quarterly, 17(4), 235-240.

Elfenbein, D. W. (June 30, 2006). Do anti-ticket scalping laws make a difference online? Evidence from internet sales of NFL tickets. Retrieved September 17, 2008, from http://ssrn.com/abstract=595682

Fisher, E. (2005). Secondary ticketing [Electronic version]. SportsBusiness Journal. Retrieved March 6, 2006, from http://www.sportsbusinessjournal.com/article/47662

Howard, D. R., & Crompton, J. L. (2004). Tactics used by sports organizations in the United States to increase ticket sales. Managing Leisure, 9, 87-95.

Johnson, A. (2005, April 11). Extra seats: Brokers find selling tickets isn't always easy. The Washington Times. Retrieved April 17, 2006, from LexisNexis database.

Kotler, P. (2003). Marketing management (11th ed.). Upper Saddle River, NJ: Prentice Hall.

Lucking-Reiley, D., Bryan, D., Prasad, N., & Reeves, D. (2007). Pennies from eBay: The determinants of price in online auctions. The Journal of Industrial Economics, 55(2), 223-233.

Mallows, C. L. (1973). Some comments on [C.sub.p] [Electronic Version]. Technometrics, 15(4), 661-675.

Price is right: How to more effectively price your tickets to maximize sales and revenue-potential (2004). The Migala Report. Retrieved April 24, 2006, from http://www.migalareport.com

Rascher, D. A., McEvoy, C. D., Nagel, M. S., & Brown, M. T. (2007). Variable ticket pricing in Major League Baseball. Journal of Sport Management, 21(3), 407-437.

Reese, J. T., & Mittelstaedt, R. D. (2001). An exploratory study of the criteria used to establish NFL ticket prices. Sport Marketing Quarterly, 10(4), 223-230.

Rishe, P. J., & Mondello, M. J. (2003). Ticket price determination in the National Football League: A quantitative approach. Sport Marketing Quarterly, 12(2), 72-79.

Rishe, P. J., & Mondello, M. J. (2004). Ticket price determination in professional sports: An empirical analysis of the NBA, NFL, NHL, and Major League Baseball. Sport Marketing Quarterly, 13(2), 104-112.

Sen, A., & Srivastava, M. (1990). Regression analysis: Theory, methods, and applications. New York: Springer-Verlag.

Snyder, P. (1991). Three reasons why stepwise regression methods should not be used by researchers. In B. Thompson (Ed.), Advances in educational research: Substantive findings, methodological developments (Vol. 1, pp. 99-105). Greenwich, CT: JAI.

Volpano, L. J. (2003). A proposal to rationalise entertainment ticket pricing using price discrimination. Journal of Revenue and Pricing Management, 1(4), 379-382.

Zeithaml, V. (1988). Consumer perceptions of price, quality, and value: A means-end model and synthesis of evidence. Journal of Marketing, 52(3), 2-22.

Joris Drayer and Stephen L. Shapiro

Joris Drayer, PhD, is an assistant professor in the Department of Health and Sport Sciences at the University of Memphis. His research interests include ticket pricing strategies, the secondary ticket market, and consumer behavior, particularly related to fantasy sports participation.

Stephen L. Shapiro, PhD, is an assistant professor of sport management at Old Dominion University. His research interests include college athletic fundraising, ticket pricing in the primary and secondary market, and the relationship between attitudes and sport-related consumption.

Table 1. Descriptive Statistics

Variable   Variable Description     Mean      Standard
                                              Deviation

AUCPRICE   Final price of the       257.75    177.74
           auctioned playoff
           ticket

TRANS      Total number of online   222.37    117.08
           ticket transactions
           per game

INCOME     Home team's per capita   21,993    2,664
           income

POP        Home tema's city         672,734   444,262
           population

TIXFACE    Face value of the        124       45
           ticket

CURWIN     Current winning %        1         0

PREVWIN    Winning % from the       0.61      0.16
           previous year

SPREAD     Wagering line on the     -8        -4
           game

BIDS       Total number of          18        11
           auction bids

DAYS       Number of days before    8         7
           the game that the
           ticket went to auction

PFB        Number of positive       1,409     3,432
           seller feedback
           comments

NFB        Number of negative       3.58      10.375
           seller feedback
           comments

LENGTH     Total number of days     4         3
           of the auction

Variable   Minimum   Maximum

AUCPRICE   58.00     1,250.00

TRANS      85.00     415.00

INCOME     18,816    30,306

POP        102,313   1,223,400

TIXFACE    60        410

CURWIN     1         1

PREVWIN    0         0.875

SPREAD     3         -14.00

BIDS       1         54

DAYS       0         34

PFB        0         25,544

NFB        0         94

LENGTH     1         10

Note: n = 750

Table 2.

Variable Correlations with Final Ticket Price at Auction

Variable   Secondary
           Ticket
           Value

TRANS       .014
INCOME      .338 *
POP        -.443 *
TIXFACE     .341 *
CURWIN      .273 *
PREVWIN     .224 *
SPREAD      .107 *
BIDS        .304 *
DAYS        .108 *
PFB         .025
NFB         .099 *
LENGTH     -.123 *
GDAY        .083 *
ROUND       .563 *
CERTAIN     .076
FORMAT      .112 *
DIV         .164 *

Note: * Correlation is significant at the .05 level.
n=750

Table 3.

All Subsets Regression Results (10-Variable Selected Model)

           Parameter   Standard
Variable   Estimate    Error      F-Value   P-value

GDAY        56.5652     6.1184     85.47    <.0001

BIDS         1.6496      .3825     18.60    <.0001

TIXFACE      .9411       .0913    106.36    <.0001

DAYSBEF      4.8925      .5866     69.56    <.0001

CURWIN     145.1367    43.1272     11.33    <.0001

PREVWIN    434.7698    44.8266     94.07    <.0001

POP          -.0052      .0002    388.89     .008

INCOME        .0130      .0022     35.03    <.0001

TRANS        1.1250      .0721    243.41    <.0001

ROUND       62.3581    11.1646     31.20    <.0001

Note: Based on a p < .05 significance cutoff