National Sports Networks rights fees: the influence of fantasy sports participation and women’s sports content Open Access

In order to better understand the dynamics of the negotiations around television rights fees for a particular sports property, the authors reviewed previous research on the topic and a set of more than 100 publicly available NSN deals to identify variables for consideration as drivers of rights fees. Although it is widely believed that the value of rights fees is highly correlated with the television ratings of a particular sports property, this assumption had not been previously empirically tested.

This data analysis resulted in the identification of three general themes (see Figure 1) consisting of eighteen potential drivers (see Table 1) for predicting rights fees for televised sports programs on NSNs. The outcome, or the dependent variable in the model, annual rights fees paid for a given sports property, operationalized as the dollar amount (USD) paid or received for a given property, as provided by a prominent North American NSN. In the case of dollars received, this was found to be for properties such as darts and poker, which typically pay the network a fee in return for slots on the network.

Data on six independent variables was collected for each sports property: TV Ratings (i.e. viewership), Sports Property Stability, Twitter (now known as “X”) Following, Facebook “Likes”, Importance of Property, and Indoor/Outdoor. TV Ratings are a measure of how many viewers are watching a particular content at a particular point in time. In previous research on Regional Sport Networks (RSNs), a strong correlation between TV ratings and rights fees was found (Foster et al., 2014). In the current research, TV ratings share (for all ages of viewers) is used. TV ratings is defined per the past research of Foster et al. (2014), who noted that ratings are tracked using boxes connected to televisions in a representative sample of homes in the United States (Tainsky, 2010) by Nielsen using an approach where a single rating point represents 1% of the total number of households in the area of interest.

The total number of viewers (or the actual number of viewers) was not provided by the NSN, so average ratings share across the age-groups was used. Given that sports properties do have a history of failures and a track record of financial uncertainty (Keshkar et al., 2021), the Property Stability variable is measured by the age of the organization, which is based on the idea that the longer the organization has existed, the more of a “going-concern” (i.e. likely to continue to exist) it is. Notably, given that relocation happens in professional sport, this variable is measured by age in current location on the premise that the longer a property has existed in one location, the more influence it has on the rights fees.

Indoor/Outdoor of a sports property’s physical location may influence television viewership. A binominal measure of whether the sports property is held indoors or outdoors is included in the analysis, founded on the idea that this could influence television viewer appeal for weather or aesthetic reasons. Social media provides a way to represent the presence a given sports content has in the marketplace (Abeza et al., 2020; Lee and Kahle, 2016). It follows that sports properties that have a greater presence and following on social media will have more fan interest and potentially the ability to attract stronger television audiences and higher rights fees. Hence, we measure the impact of social media by including two variables: (1) by the number of Facebook page likes and (2) the number of Twitter followers that the sports property has. The focus on Facebook and Twitter is justified as these were the two most used social media outlets in North America at the time of data collection (Hull and Lewis, 2014). The final variable in the Sports Property theme is the Importance of Property. Beginning in 2003 and until 2018, ESPN published an annual fan survey which includes an overall fan rank of properties across the “Big 4” professional North American sports leagues in eight categories: Affordability, Coaching, Fan Relations, Ownership, Players, Stadium Experience, Bang for the Buck and Title Track. These results outline the local market strength of clubs in these particular leagues. However, our sample included many observations that were sports properties outside of the Big 4 sports, where comparative data was not available. As such, this variable was designed as a binary variable with a “1” recorded for a Big 4 sport and a “0” for all other sports.

The Deal is the second theme area. It is measured by three variables: (1) Length of Deal which is the number of years of the current deal with the NSN (2) Collective Deal which reflects if the deal is independent or part of a large collective agreement, (3) Timeslot Attractiveness, which is the strength that the content of interest has in comparison with other content that is broadcast at the same time, in the same slot. Particular aspects of a deal between the NSN and the sports property can create fluctuations in rights fees. If a sports property has never had a prior NSN deal compared to a renewal situation of a pre-existing deal, variability in price could result. Similarly, if the NSN is entering a deal for the first time with a sports property that had recently been with a competing NSN, this could also have an effect on rights fees (i.e. competitive factors, bidding impacts). It is important to note that a sports property could have multiple deals with multiple NSNs (or broadcast networks), which is normally only the case for super value content, such as the NFL who has separate deals with all the major cable television networks in the United States. Situations involving high value content, such as the NFL, and multiple NSNs will likely lead to higher rights fees, whether due to the value of the sports property, inter-NSN competition or both.

The objective of including competing content in the analysis is to measure the strength that the content of interest has compared to other content that is broadcast at the same time in the same slot. If content performs well versus other content shown in the same time slot, it is likely that the price of the former content should be higher as it has the ability to generate “share” for the NSN in that slot versus other NSNs. Thus, the Timeslot Attractiveness variable is measured by indexing the ratings of the sports program of interest versus the average of the content of the NSN in the same timeslot. To control for seasonality, the entire year of data (24 h/day, 365 days/year) was used to determine the average ratings for each slot. Ideally, the index would consider all television networks or even all NSNs, but that data was not available to the researchers and thus the ratings data of only the single NSN was used.

The third theme area is Sports Content Attractiveness. At the regional level on RSNs, content which has the ability to appeal to niche markets and local demographics attracts viewers in that market (Foster et al., 2014). However, in the context of a NSN, a national audience is sought without attachments to local properties and market. Further, since advertisers have the ability to reach a wider group of people via the NSN than via the RSN, they can charge more for the advertising and, in theory, ultimately pay more for content based on the type of audience that the content reaches. This variable was calculated as a sum of z-scores, where z-scores explain the exact distance of a statistic from the mean of the normal distribution, for the full year of NSN ratings share across each of the different 14 demographic groups (i.e. children 2–11, teenagers 12–17, M18–24, M25–34, M35–49, M50–54, M55–64, M65+, F18–24, F25–34, F35–49, F50–54, F55-64 and F65+) to provide a measure of the appeal of a given sport property to a broad range of demographic groups. The rating for the demographic group for each record received a z-score, and then all were summed to show the popularity (or not) of the given content across gender demographics. The overall entertainment value, as well as the ability to attract viewers and influence rights fees, is impacted by the attractiveness of the content to viewers.

The “celebrity” factor of the sports content, or the presence of celebrities in a particular event (see Foster et al., 2014), is considered in this study. Since published records of celebrities (e.g. Forbes Top 100) do not include many athletes from content in this data set, this is measured via the social media analytics of the top players/athletes within the sports in 2015, as a binary variable defined as “1” for top half (higher levels of celebrity following) and “0” for bottom half (lower levels of celebrity following) used.

The popularity of a sport overall can impact television viewership and potential television viewership of a particular event, and could, in turn, have an effect on rights fees. The major global professional sports leagues (four in North America; five in Europe) are popular on a global scale (major), but the more niche sports tend to have specific followings (minor), leading to the use of a binary variable of major/minor, which was used to assess the influence of popularity of a specific sport. Participation levels across sports around the world could have an effect on their popularity and thus, potentially, their viewership. The ability of a person to access these sports, either through playing or watching, could influence their willingness to watch. Sports accessibility is measured via three different variables: (1) the number of participants (i.e. the total number of people playing the sport), (2) the cost of entry to participate (i.e. an estimate of minimum costs required to play the sport) and (3) the number of national sports organizations globally in that sport (i.e. membership in the sport’s International Federation (IF)). Further justification for the second and third variables under sport accessibility is provided here. First, the basic equipment cost of participation for each sport in North American city was estimated based on public data on the costs of essential gear to play the sport at a development level. For example, for soccer (the lowest participation cost sport), the costs for player jerseys, boots, shin pads and training balls for a team of 11 players was calculated. The cost per player was $430, and team training balls came up to $270, for a total of $5,000 per team. We used per-team estimates if participation required one to be part of a team. Second, for the members of the sport’s IF variable, the rationale is based on the fact that every IF is structured differently due to a number of factors, including politics, geography and country-by-country interest in the sport. For broadcasters, the IF structure is an important means by which to inform a favorable broadcast agreement, where the IF’s membership size (i.e. the number of NSOs) is a measure of the popularity of the sport, and a proxy for global interest and potential “eyeballs” that could be reached both globally (NSO’s) and domestically (e.g. immigrant populations that have an affinity for a sport based on their home country). The fantasy and gambling aspects of a given sports property are assessed based upon the idea that involvement with a fantasy team or wagering on the competition will increase an individual’s interest in viewing (Martin and Nelson, 2014) and, in turn, increase the rights fees. This is measured using fantasy data (two different variables) as gambling data is not readily available for many sports properties. According to the Fantasy Sports and Gaming Association (2023), 55% of fantasy sports players (reported to be 62.5 million North Americans in 2022) report watching more sports on television since they started playing fantasy sports. The first measure used (binomial) is the availability of fantasy sports via a binary representation that there is an active fantasy industry for that sport (1) or that there is not (0). The second measure is the estimated number of participants in fantasy for that specific sport as provided by a Fantasy Sports and Gaming Association (2023) report.

An important point to consider when assessing rights fees is the additional benefits that may be included in a contract beyond the base fee. For example, the current NFL-ESPN deal provides a chance for ESPN to add a wild card playoff game to its coverage and includes an option for the NFL to unconditionally give one game to ESPN, taking a game away from another network (New York Times, 2019). These dynamics within NSN-sports property negotiations highlight the multivariate nature of determining a rights fees amount, supporting the consideration of the wide array of independent variables described here and summarized in Figure 1 and Table 2.

An anonymous NSN provided a full year of ratings for all sports programming on each of their channels and subsidiaries. The rights fee paid (or received) for each program was also provided. The original data set contained each and every program on the NSN (all of its channels) for the 365 days of 2015 with more than 40,000 entries. Data included name of event, time start/finish, ratings (overall, share), ratings share by demographics (14 groups by age/gender) and rights fees. All RSN programming and non-live programing (e.g. news shows, talk shows, documentaries) was removed, as was programming without rights fees (i.e. content owned by the NSN). This resulted in a data set of 1,952 entries of live sports content (with rights fees data) that formed the basis for analysis. These entries included data points from North American leagues, European leagues, golf, tennis, auto racing, mixed martial arts, boxing, Olympic sports, NCAA sports, youth sports and more. The data provided by the NSN, in addition to data provided by secondary sources, industry reports and public websites was used to populate the dependent variable (annual rights fees) and as many of the independent variables identified in Figure 1 as possible.

The data was analyzed using univariate and multivariate approaches to assess the determinants of rights fees paid for by NSNs for sports content. Previous research in sports management has adopted a similar methodological approach (i.e. a full-season analysis) including Foster et al.’s (2014) study of RSN ratings. A principal components analysis via a varimax rotation was used to identify factors of high commonality. Factors with eigenvalues of 0.7 or higher were retained with the objective to provide understandable results (Camiz and Pillar, 2018; Weir and Vincent, 2020) and account for multicollinearity. For each factor, a proxy (i.e. the most relevant variable) was identified (Nadeau and O’Reilly, 2006). Following data reduction, a univariate correlation analysis (both Pearson and Spearman) was used to determine the correlation between each of the independent variables and the dependent variable. The correlation analysis provided insights into the relationship(s) between the independent variables and rights fees. Finally, a linear regression predictive model was built using the proxy variables for each of the identified factors to estimate the rights fees for any NSN-televised sports product.

For each of the independent variables and the dependent variable, descriptive measures are provided in Table 2.

Figure 1 presented 18 potential independent variables as drivers of rights fees in NSN-sports property deals. Data for the independent variables and the dependent variable was provided by a single NSN (on the condition of anonymity) for all 1,952 observations and analyzed using SPSS 26 for analysis. Before reducing the data into factors for a linear regression model predicting rights fees, the sample was evaluated for its appropriateness for data reduction to address multicollinearity and endogeneity. Following this analysis, the data set was deemed appropriate based on three factors: (1) the sample size of 1,952 is large enough (Comrey and Lee, 1992), (2) the Bartlett’s Test of Sphericity resulting in a Chi square of 33,575 (p < 0.001) and (3) the Kaiser–Meyer–Olkin measure of sampling adequacy was greater than the accepted threshold of 0.6 (0.730 (p < 0.001)).

Next, a principal components analysis was undertaken following past research (see Nadeau and O'Reilly, 2006). Table 3 presents the results of the principal components analysis which reduced the 18 variables into 7 factors due to a number of variables having the “same” influence as others on the variance within the data set. The seven variables selected to be the most appropriate proxies for the seven factors explain approximately 83% of the variance (see Table 3). Each factor and proxy was reviewed and given a theme. The strongest component of the first factor is “fantasy sports participants”, which was retained as the proxy variable. The second factor is “low social media use” (reverse of Facebook page likes and Twitter followers being the two strongest variables). The proxy used is Facebook page likes. The third through seventh factors have a clear variable that is the proxy, namely “television ratings”, “cost to enter sports as a participant”, “demographic reach”, “deal length” (in years) and “indoor/outdoor venue” (themed as the geography of the telecast).

In order to further investigate the relationships of each factor with the dependent variable, each of the seven selected factors/proxy variables was analyzed using both Spearman and Pearson correlations vs the dependent variable. The results are presented in Table 4.

As reported in Table 4, there are some differences in the Pearson and Spearman correlations for each of the seven proxy variables. Both Pearson and Spearman correlations are reported for each variable since the format of the data differs for each variable. Spearman rank correlation is appropriate for the proxy variables that are not continuous (i.e. Indoor/Outdoor). Pearson rank correlation should be viewed for those proxy variables that are continuous (i.e. Active Fantasy Participants, Facebook likes, TV Ratings, Cost to Participate, Demographic Reach and Deal Length).

A few notable observations from the correlation analysis are reported in Table 4. First, TV ratings has a strong correlation to rights fees but it is not the only item explaining what is spent. Specifically, the correlation between TV ratings and rights fees is 0.714 (Pearson) and 0.428 (Spearman), both significant at the p < 0.001 level. Second, the correlation between active fantasy sports participation and rights fees is also very high (0.619 Pearson and 0.705 Spearman), supporting the result of the data reduction presented earlier. Finally, the only one of the seven factors/proxy variables that does not have a significant univariate correlation with rights fees is demographic reach, which could be a function of the NSN’s national reach.

The multivariate analysis involved regressing (linear) the seven proxy variables against the dependent variable to assess the relative magnitude of their influence on rights fees. A strong model resulted (R Square = 0.614; F = 442.060; p < 0.001) and it should be noted that four of the seven variables have a significant influence (three at the p < 0.001 level, and one at the p < 0.01 level) on rights fees. As noted in Table 5, the four dominant influencers on rights fees are TV Ratings (t = 28.045, p < 0.001), Number of Active Fantasy Participants (t = 17.437, p < 0.001), Deal Length (t = −2.963, p < 0.01) and Indoor/Outdoor Venue (t = −9.213, p < 0.001). It is important here to note the direction of these four influencing variables; with higher rights fees being associated with (1) higher TV ratings, (2) higher number of fantasy participants, (3) the event being held outdoors and (4) shorter deals.

All four variables are key drivers of rights value, because they are both statistically and economically significant. In terms of effect size, 1 point increase in ratings results in $539,515 increase in rights value, 1 additional person participating in fantasy sports participation, results in $360 increase in rights value, while 1 year increase in length of contract (deal) results in decrease of rights value by $141,347. Each indoor sports event also decreased the rights value by $3,399,796. It is clear these are important effects on the overall value of the broadcast rights value.

The other three variables – increased social media use (Facebook likes), cost to participate and demographic reach – are not significant predictors of rights fees.

Normality tests validate statistical modeling (Weir and Vincent, 2020). We tested whether the assumption of normality was valid using the Kolmogorov–Smirnov test. We could not reject the null hypothesis of normality (p = 0.000). Supporting that results were not in infringement with the presumption of normality.

This research set out to identify and quantify the magnitude of the key influencers on rights fees in NSN-sports property negotiations. The impetus of the research was from executives of the NSN who were experiencing challenges with pricing rights fees and negotiating with sports properties around price. These NSN executives seek a more sophisticated way to price and negotiate. This discussion can be found in both media reports and the academic literature has related work (Foster et al., 2014). Although findings are that ratings and rights fees are correlated (0.741), this research finds considerable variability between ratings and rights fees from a multi-variate perspective, justifying the need for a more holistic approach. Supported with proprietary data and structured per previous literature (Foster et al., 2014), this study includes an analysis that identifies the drivers of rights fees and builds a predictive model for use by practitioners and researchers alike.

Results indicate that TV ratings are a key driver, as expected, and contribute to previous literature on the importance of TV ratings on rights fees. However, importantly, the very strong influence of off-the-field engagement (measured by fantasy participants, representing a factor with many embedded variable) is a key finding for NSNs and sports properties alike. Results indicate that valuable sports content goes beyond televising a live game and suggests that NSNs should consider incorporating similar initiatives that can increase viewing demand. For example, previous research has shown that participation in fantasy sports complements sports consumption, including television viewing (Martin and Nelson, 2014). Embracing fantasy sports and technological advances (e.g. esports, online gambling) within off-the-field engagement may contribute greatly to the valuation and negotiation of NSN rights fees. Promotion of fantasy sports, or creating fantasy sports incentives, may maximize and increase the value of sports content. This could be applied outside the Big 4 leagues to positively influence interest in other sports properties, such as tennis, running, cycling, cricket, rugby, and other sports properties viewed as “less popular” in North America.

Another notable finding is the negative influence that deal length has on rights fees. This indicates that multi-year contracts might actually devalue a sports entity’s future worth. With multiple NSNs, RSNs, sub-channels and digital channels in the market, and with new entrants expected, pushing toward a high level of saturation of sports content, there is a pressure to lock sports entities into longer deals guaranteeing the NSN longevity by filling relevant programming slots long term and ensuring sports property content to build the NSN. This is reflected in the results, which indicate that there may be a trade-off in cost/price in return for the longer term control of content. This could lead to particular leagues, clubs and other properties finding themselves stuck in a contract that does not reflect their true value, although they will have the security of the long-term deal and a guaranteed revenue stream. From a practical perspective, the predictive model allows managers to assess deal length as a way to measure risk vs reward tradeoffs of getting paid less (lower rights fees), but having a longer deal (increased security, more time to deliver). Thus, using the predictive model to estimate different deal length scenarios can inform decision making.

Results indicate that NSNs can overpay or underpay for the rights to any given sporting event and should take into consideration multiple factors beyond ratings. For example, the predictive model from the current research suggests that the NSN overpaid for MLB content. Specifically, the predictive model found that a regular season MLB game had an outcome worth 56.5% of the actual rights fee paid. Additionally, the NSN was found to have underpaid for a specific tennis property, with the predictive model finding an outcome of 193.3% of the actual rights fee paid. Both of these tests of the predictive model support the finding that the right value to pay for a given rights fees requires the consideration of multiple variables in research and in negotiations. Perhaps, the NSN has market research to suggest that MLB is an important area of investment for the future, so they had to outbid another network for those rights. Or perhaps no other NSN was interested in tennis, so a lower price was possible in the negotiations.

The television appeal of local geography is another notable finding about a variable that influences rights fees. Since the marketing of attractive destinations appeals to (potential) tourists, relating this to the current context, results suggest that viewers may be attracted to sports content that is presented from an attractive destination. For example, an outdoor event like the Tour de France cycling race may be able to charge more because of its geographic appeal (e.g. beautiful views, historical monuments, exposure to different cultures) (Van Reeth, 2022).

The findings from Study 1 that four variables – Number of Active Fantasy Sports Participation, Deal Length, Indoor/Outdoor Venue and Television Ratings – had significant influence on Sports Media Rights Fee motivated the deeper exploration of these variables in Study 2. We were especially interested in understanding the positive relationship between Active Fantasy Sports and Sports Media Rights. Although in Study 1, there was no significant relationship found between Female Inclusion (i.e. the proportion of network content that is female sport) and Sports Media Rights fees, industry data showing that 33% of fantasy sports participants are women (FSGA, 2023) suggests that Female Inclusion in Sports Network programming could have an effect on Active Fantasy Sports Participation. The second interest examined in Study 2 was to investigate how Active Fantasy Sports Participation contributed to Sports Media Rights Fees Values. Table 6 presents the Theoretical Model for Study 2, which includes four hypotheses.

Industry data from the Fantasy Sports Gaming Association (FSGA) estimates there are 62 million active fantasy sports participants in North America (FSGA, 2023). Within these statistics, the estimated female participation in fantasy sports has seen a massive surge in the last ten years. For instance, Ruihley and Billings (2013) reported that in 2012 only 10% of all fantasy sports players were women. However, recent FSGA data highlights that women players now represent 33% of participants in North America. Notably, 38% of participants in the single largest fantasy sport (i.e. fantasy football) are women players (Dwyer et al., 2018). We theorize that increased broadcast of female sports during this period (Female Inclusion), may have contributed to increased engagement of women with fantasy sports, therefore resulting in an overall total increase in active fantasy sports participation.

Past research evidence points to the positive relationship between fantasy football participation and NFL television ratings (Fortunato, 2011; Sung and Tainsky, 2014), and as a motivator of additional NFL sports media consumption with a positive relationship between fantasy football participation and the use of second screens for NFL games (Billings et al., 2020). Further, previous research shows that fantasy sports can help increase fan bases, with one specific study noting a positive relationship between fantasy sports participation and the conversion of non-fans to watch NASCAR races (Goldsmith and Walker, 2015). In terms of fandom, fantasy sports digital experiences have been shown to positively enhance the sports consumer journey (Yuksel et al., 2021). We therefore theorize that participation in fantasy sports can result in a positive increase in the fan viewership experience (and thus increased television ratings) of actual sports matches or games.

With the rise of social media and digital sport media channels, there is increased scrutiny as to the influence of so-called “legacy media” such as television broadcasts to impact sports media consumption (Chan-Olmsted and Kwak, 2020). This will be a major factor in negotiating future sports media rights fees. However, research shows fantasy sports programming on television attracts increased viewership from fantasy sports participants (Kupfer and Anderson, 2021). Other researchers (Billings et al., 2021) advance the argument that fantasy sports and sports media have been perfect partners to boost each other’s participation and consumption. There also appears to be increased opportunities for sports broadcasters, including NSNs, to further capitalize on fantasy sports participants (Dwyer et al., 2022). The authors believe that sports programming that is able to incorporate fantasy sports content will attract higher media rights value due to the clear potential for increased viewership from fantasy sports participants.

The fourth hypothesis (H4) is about television ratings. We hypothesize that participation in fantasy sports results in increased viewership of actual matches or games, which in turn drives television rights fees in the market. Sports media rights buyers will seek clear proxy evidence that projections from potential increased viewership are derived from fantasy sports participants. That evidence is television ratings of sports broadcasting.

Study 2 followed the same method for data collection as Study 1. A correlation matrix of results is shared in Table 4. In Study 2, we had four hypotheses which we tested using multiple linear regression models shown in Table 7. The first hypothesis (H1: Female Inclusion has a positive influence on Active Fantasy Participation) was not supported as shown in Model 2 (ß = −0.627, p ≤ 0.01). Instead, we found a significant negative relationship between the level of female sports programming (Female Inclusion) and overall fantasy sports participation. That is, more inclusion of female sports broadcasts resulted in lower overall fantasy sports participation. This finding is intriguing for two reasons: (1) this suggests that even among the increase of female fantasy sports participation over the last 10 years, including more women’s sports in broadcasts may decrease overall fantasy sports participation and (2) it suggests that including more women’s sports in network broadcasts may result in both men and women decreasing their overall interest in viewership, with a proxy of this behavior being decreased fantasy sports participation. If either suggestion is true, it could provide some very early implications for television rights value for women’s sports matches that have historically lagged behind that of television rights value for men’s sports matches.

The second hypothesis (H2: Active Fantasy Participation has a positive relationship with Television Ratings) was supported in Model 4 (ß = 0.704, p ≤ 0.01) as the coefficient was significant and positive. Similarly, the third hypothesis (H3: Active Fantasy Participation has a positive relationship with Rights Fees) was also supported in Model 6 (ß = 0.729, p ≤ 0.01) showing a positive and significant coefficient.

The fourth hypothesis is that Television Ratings will mediate the relationship between Active Fantasy Sports Participation and Sports Media Rights Fees. Baron and Kenny (1986) required that three steps need to be analyzed in establishing mediation. In Step 1, the independent variable (for our case: Active Fantasy Sports Participation) should have an influence on the mediator variable (Television Ratings). This was supported by Model 4 in Table 7. For Step 2, the independent variable (Active Fantasy Sports Participation) should influence the dependent variable (Sports Media Rights Fees). This was confirmed by Model 6 in Table 7. In the final Step 3, controlling for the independent variable (Active Fantasy Sports Participation), we must first show the mediator variable (Television Ratings) influences the dependent variable (Sports Media Rights Fees). This is confirmed in Model 7 in Table 7 (ß = 0.523, p ≤ 0.01). After all these three conditions are met, we need to establish that the influence of the independent variable (Active Fantasy Sports Participants) on the dependent variable (Sports Media Rights) must be less in Step 3 (Model 7) compared to Step 2 (Model 6). This was demonstrated when the size of the co-efficient for the independent variable (Active Fantasy Sports Participation) decreased from Step 2 (Model 6, ß = −0.729, p ≤ 0.01) to Step 3(Model 7, ß = −0.361, p ≤ 0.01) confirming partial mediation.

The results of this research support that NSNs can adopt the predictive model to improve their valuations and decision making around which sports properties to include on their network and how to negotiate the rights fees to be paid for properties they do want to include. Indeed, the use of the predictive model can help NSNs to (1) inform negotiations, (2) prospect potential “deals” at the sports level or the level of the sports property, (3) identify areas of potential future investment and (4) equip advertising sales teams.

The identification of the significant drivers of rights fees is a valuable finding for NSNs. For instance, NSNs can achieve more (viewers, advertising revenue and cable/a la carte subscriptions) from investments in a sports property by creating off-the-field engagement and developing digital content that can be packaged at a premium or that is of high interest to consumers (e.g. College Game Day on ESPN). By extension to the current climate, this reality is further enhanced as digital NSN channels (e.g. ESPN+, Sportsnet One, Peacock, etc.) are continuing to emerge/expand and are taking viewers from traditional cable.

Findings also suggest that the demographic reach of a sports property should be interpreted cautiously by the NSN in negotiations and content selection. Although there is only a modest influence on rights fees reported, technology and demographic shifts should not be disregarded by the NSN. Notably, younger generations not only consume more sports entertainment content through other forms of technology (e.g. mobile phones, tablets, streaming services such as Apple TV), social media (e.g. Facebook, Twitter, Instagram, Snapchat) and sports formats (e.g. eSports, fantasy sports) than traditional cable television, they also have decreased sports participation rates as they grow older (Eime et al., 2016). All of which have the potential to impact the NSN’s ability to attract younger viewers. Thus, content that is appealing to youth and mobile delivery methods are essential in this regard.

Finally, to combat illegal and unlicensed streaming of their content (i.e. revenue lost by the NSN), results of the current study suggest that NSNs need to provide a superior product, through expanded engagement tactics, such as using elite game announcers, diffusing interesting promotional information, providing higher-quality digital streaming, amplifying live coverage with engaging social media content, linking to esports, offering gambling options/links and offering interactive fantasy sports contests. Further, since results show that off-the-field engagement adds to the valuation of rights fees, creating stories around notable athletes and engaging the audience through social media content are just a few off-the-field engagement ideas that the NSN can implement.

Our second study has demonstrated the importance of incorporating more fantasy sports programming segments into sports broadcasting. Exploratory findings show that attracting fantasy sports participants to these special segments can positively increases television rating of sports broadcasts, and in turn increase sports media value of sports properties. However, broadcasters need to better understand our findings that women’s sports content has an inverse relationship with fantasy sports participation. This is a major concern as women sports stakeholders have forcefully argued that sports broadcast rights value for women’s sports programming have been significantly undervalued by commercial sports entities.

The results of the current study have practical implications for sports properties and their managers. The first is the same trade-off in the length of contract as noted for NSN professionals deal previously, where it is clearly a decision balancing risk with revenue maximization for the sports property. Since the results show that longer length contracts, although more secure, are typically of lower revenue for the sports property, the sports property needs to consider their future value, external market trends in media rights and their willingness to take risks. For smaller sports properties, there may be an opportunity to garner additional and/or better coverage with the flexibility of shorter term deals. For example, a property like the World Curling Championships could have an opportunity to generate more revenues and more/better coverage with shorter term deals, vs the financial security that comes with a long-term deal in a sport where television deals can be scarce.

The second notable finding for sports properties is that they can increase their revenues from NSN rights fees with enhanced social media activity. Since results indicated that NSNs are paying a premium for content with high social media activity, this should be a priority for a sports property with a decent following. Thus, they should create relevant and enhanced content on their social media platforms. Ideally, in combining this finding with other results from the predictive model, the creation of fantasy leagues, interactive platforms, low-cost participation programs and engaging outdoor events are all tactics a sports property should consider to provide content for and/or leverage social media content.

Finally, results suggest that smaller sports properties should re-assess their benchmarks for getting on the air or being streamed on a network (NSN or, by extension, RSN). Specifically, the finding that poker, darts and billiards all pay the network for the right to be on the air (i.e. purchasing time slots for their content), means that the managers of sports properties who are seeking television coverage for their events need to be cognizant of the fact that they are competing against content that pays for slots, not just other sports properties of similar reach and following, who would like the same slots.

The Study 2 finding about the importance of fantasy sports participation on both television ratings and sports media broadcast values reaffirms that sports media rights owners need to continuously invest in attracting fantasy sports fans. They form a major market segment that is being courted by new media (i.e. social media), while there is strong research evidence that traditional media broadcasts (i.e. television) still is a formidable consumption channel for the segment.

Given the sensitivity around the data used for this research, there are limitations to this research. First, we were provided rights data from the NSN for a 12 month period. We signed an explicit Non-Disclosure Agreement and no contracts or contract-level data was shared with us, other than the rights fees paid for the content. We were not privy to the specifics of the contracts, as it was clear that the NSN had confidential commercial agreements with multiple sports rights owners. Thus, we are unable to explore a number of deeper analyses (such as controlling for the number of hours of content or the value of given timeslots within a contract). If such data was available, future research extending our work would be of high value.

Second, there is no public access to valid data related to NSN and RSN contracts. Details on these deals and data related to ratings are not publicly available. Thus, a major limitation of this research from the onset was the heterogeneity of the data available since there are multiple types of (confidential) contracts spread across multiple sports events broadcast on multiple NSN’s and time slots. A further point of consideration here is that for some content, the event/game was televised not only as a live event but also as a replay of the event later in a non-prime-time (i.e. less attractive) slot. Given that the rights fees received are typically done on an annual basis, and the amount of coverage/allocation of slots differs by property, this is a limitation. Future research where rights fee per game/event data is available could help address this. Another line of future research which analyzes how many minutes of content a given contract delivers and how many distinct events are covered in order to more precisely assess if a given sport is over or undervalued. Third, the current study is also limited by the use of a less than ideal number of independent variables (18) which limits the scope of the research findings. It is suggested that future research incorporate other independent variables such as (1) ownership intent of sports property (win, profit, etc.), (2) non-playing celebrities (coach, owner, etc.), (3) quality of the delivery of the content, (4) the current sponsorship portfolio of the sports property to better understand and (5) the influence of game broadcast time slot. For example, research by Nalbantis et al. (2023) shows that there are important overlap effects of viewership for different European soccer league telecasts on US television. Future research that includes these additional variables would allow for the development of a model that explains more of this variance. Further, having access to the actual number of viewers for a particular program (vs ratings share) would also be recommended for future research. Moreover, the impact of interactions on Instagram and/or TikTok, two increasingly relevant social media platforms, on the rights fees, although insufficient in terms of data for this study, should be included in future and similar research.

A fourth and final limitation of this study is that the data is from a single NSN. Therefore, caution must be taken with any proposed generalizations to a broader set of NSNs. However, notwithstanding the challenges in acquiring such sensitive data, there is considerable opportunity in this regard for future research, in comparing the findings not only to other NSNs but to any national broadcast network that diffuses sports content, including cable networks (e.g. NBC, BBC, CBC) and RSNs (e.g. Yes Network, Fox Sports Ohio).