Moneyball Madness

In the first line of his recent Grantland article, Bill Barnwell proclaims, “It's just about impossible to statistically measure the return that each draft pick provides after being selected in the NFL draft.” After giving reasons why this is the case, Barnwell then goes on to…measure the return of each draft pick selected from 1997-2007 (he wanted to look at the first five years of each player’s career). He gives a player one point for starting a game, one-half point for coming in as a substitute, ten points for each Pro Bowl Appearance, and twenty points for Each All-Pro appearance (those players who were both All-Pros and Pro-Bowlers were only given points for their All-Pro appearance). He added up all the of players scores and see how each draft pick from this time period stacks up within his rating system. This would allow him to compare how all players against each other even if they competed at different positions.

Barnwell says acknowledges the flaws in his system, but suggests that it is more of a starting point for discussion. Barnwell is right that there are flaws in his scoring system. The biggest problem is the fact that he is trying to do a relative valuation without actually giving reasons for what a “point” means. For example, what is the reasoning behind giving ten points for Pro Bowlers and twenty points for All-Pros. Are All-Pros really twice as good and/or twice as valuable as Pro Bowlers? It would be very difficult to make any argument that is proves that is the case.

The bigger problem with Barnwell’s approach is that it reflects a trend in the sports industry. Since the publication of Moneyball the book and the release of the Moneyball movie, there has been an explosion in the use of quantitative analysis in sports. This has largely been a good thing as teams like the Oakland A’s, Tampa Bay Rays, and Houston Rockets have been able to effectively compete against teams with more resources. The problem is that now everyone who works in or around the sports industry is trying to find their own version of Moneyball. Yet, finding the right form of quantitative analysis for each situation is very difficult. For example, Bill James (the godfather of sports analytics) published his first book 1977. He is still finding ways to better evaluate baseball from quantitative perspective over 30-years later.

Instead of taking the appropriate approach to this problem, many sports team and leagues take Barnwell’s approach. While it appears to have surface validity, Barnwell’s rating system fails to hold up to even a little scrutiny. It is not just that his rating system is suspect, but the idea that simply starting games makes you a more effective player is questionable. For example, Wins Above Replacement Players statistics in baseball and Player Efficiency Ratings statistics in basketball were developed in part because some players who start are not helping their team win games. In essence, these starting players were less valuable than reserve players and actually could damage a team’s overall performance.

This is not to say that sports organizations should not continue to try to employ quantitative analysis for both on and off the field issues. They should take the right approach to solving the problem. To accomplish this goal, sports organizations need to identify their target metrics and determine what factors help maximize results. In many team sports, the objective of the game is to score more points, goals, etc. than the other team. Therefore, sports organizations should find players who can maximize point differential whether on offense or on defense (or both).

On the business side, one area that has received more attention from a quantitative perspective is sports sponsorship. The primary goal of many corporate partners is to work with teams to grow their revenue streams. Therefore, sports organizations should determine how they can help their partners achieve this goal and track this ability throughout the course of the year. Unfortunately, many sports organizations do not how their corporate partnership inventory drives revenue or delivers a tangible ROI to their sponsors. Block Six Analytics Partnership Scoreboard and Corporate Asset Valuation Model can help sports teams and leagues achieve this goal.

Quantitative analysis is and will be a critical part of the future of sports. Making sure to evaluate the quantitative analysis is something that every sports manager, fan, or sponsor should do to avoid the Moneyball madness.

Valuing Intelligence and Decision Making In Athletes

When most people think of athletes, they mainly believe that thinking is not a critical part of being an elite athlete. However, a recent CNN article highlighted a Swedish study that “suggests that elite soccer players outperform players in lower divisions in tests of certain cognitive abilities, and both groups bested the general public.” While this was a small study (only 57 male and 26 female Swedish professional soccer players), it does highlight an interesting idea that Block Six Analytics has started to examine as well. Athletes are continually becoming bigger, stronger, and faster than compared to their counterparts in the past. For example, the Washington Redskins offensive linemen called the “Hogs” anchored the team’s three Super Bowl Winning teams in the 1980s and 1990s. Hogs are not known for the svelteness and these massive players were not different weighing in at an average of 279 pounds. Yet the hogs would be considered lightweights in today’s NFL. Last year’s Green Bay Packers’ offensive linemen weighed an average 314 pounds.

Even as size and overall athletic performance continues to increase, the proliferation of strength and conditioning coaches as well as advances in nutrition and dietary habits (and performance enhancing drugs in some cases) make the differences between athletes increasingly small. In the NBA, the average player height is 6’6”. Yet, a player who is 6’8” is often considered too small to be a center or even power forward despite there is only a 5% difference in height between a 6’8” player and a 7’0” player. During the NFL Combine, running a 4.3 second 40-yard dash and running 4.6 second 40-yard dash (or a 7% difference) could earn a draft pick hundreds of thousands (if not millions) of dollars. In baseball, a “power” pitcher throws 95 mph while pitchers who throw 90 mph need to focus on their “control”. Because the difference in athletes has become smaller, many professional sports organizations are making spending millions of dollars for only relative minor perceived or actual increases in performance.

Making larger bets on smaller differences is similar to what has caused many of the greatest financial disasters in American history. In 1998, a company with future Nobel Prize winning economists called Long-Term Capital Management (LTCM) almost caused global panic. Prior to 1998, the firm made been successful making bets on the fluctuations of different currencies that generated outsized annual returns (around 40%) for its investors. As more people discovered and employed LTCM’s strategies, the returns on these trades decreased. This required the firm to invest capital in much riskier transactions with smaller margins. The company went bankrupt and lost hundreds of millions dollars of investors money because it made a large bet on the increase in value of the Russian Rouble that failed after a default by the Russian government on its outstanding loans (meaning the currency rapidly declined in value). Because many of the nation’s largest investment banks had invested millions of dollars in LTCM, they also suffered huge losses. There were have been a global financial panic had the US government not stepped in to help these companies (does this action sound familiar to the bailouts of the financial services and automobile industries that are going on today?).

Unfortunately for sports organizations, it is unlikely that the government will bail them out if they keep making decisions primarily focused on athletic performance. Therefore, teams should examine new areas where they can gain competitive advantage when evaluating current and perspective players other than athletic performance. Identifying differences in intelligence and decision-making ability is likely to be an area where teams can receive significant return on investment. Teams that can identify athletes who are smarter and make better decisions, even if they are not superior athletic talents, will be able to make more cost-effective personnel decisions. For example, quarterbacks have the highest annual salary of any position in the NFL. Yet, an examination of the league’s best quarterbacks shows that most of the top-performers are not top-athletes. Aaron Rodgers, Drew Brees, Tom Brady, Philip Rivers, Matt Schaub, Eli Manning, Peyton Manning and Philip Rivers may be underrated for their athletic ability. However, it is difficult to argue that their athletic ability is the reason they are successful at their position. It is their ability to analyze the field and make quick, accurate decisions that has been the critical element to their success. A recent Grantland article about Gerald Greene supports this idea. While Green has always had superior athletic ability (he once dunked over a line of teammates during a high school dunk contest) his current success now comes from being a “student of the game”. He also said that “These guys at this level…are too smart” to allow someone with great athletic ability to become a star player without taking time to understand the game. Green, the ultimate athlete, has recognized how important intelligence, preparation, and analysis is to being a successful NBA player.

Numerous intelligence and decision-making tests are already on the market that teams and leagues can use to examine a potential player’s intelligence. For example, The NFL requires all draft picks to take The Wonderlic Cognitive Ability Test. A recent controversy emerged when it became public knowledge that LSU cornerback Morris Claiborne had scored a 4 out of 50 on his Wonderlic exam. On the Yahoo! blog Shutdown Corner, author MJD stated, “Why I know [Claiborne’s score], I'm not sure. Why I'd ever want to know it, I have no clue.” MJD is right in saying that the Claiborne’s Wonderlic score should not have been released to the public. However, there is significant value in knowing a player’s cognitive abilities especially related to decision-making ability. While it makes common sense to judge athletes on their athletic performance, sports organizations should think more about evaluating players on their intellectual ability.

Microtargeting Equals Macrosuccess

Any article whose title starts out with “The Creepiness Factor” these days usually has sports fans thinking of Penn State, Syracuse, or Bobby Petrino. In this instance, The Atlantic is writing about something a little more benign - how much information political campaigns have about current and potential voters. The “creepy” part of the article is that political pollsters, strategists, and candidates have access to so much data that they can determine which behaviors, actions, or values are associated with specific individuals or groups of individuals. More importantly, political candidates and their staffs can use this data to see how these behaviors translate into voting behavior. For example, this article cites that drinking Diet Dr. Pepper means you are a person who is more likely to be a voting Republican while drinking 7up means you are a person who is more likely a non-voting Democrat. For many people, it understandably seems bizarre that these political campaigns have access to what soda you drink and can tell how this reliably translates into your voting preferences.

For sports organizations, however, understanding and employing microtargeting is one solution to attracting and retaining customers in an increasingly competitive environment. For example, many colleges used to rely on the fact if you put “if you put up enough billboards and sent out enough brochures, people would show up” to games according to Central Florida's DeVos Sport Business Management Graduate Program professor Bill Sutton. Yet, large schools like Georgia Tech and the University of Tennessee saw dramatic declines in their season ticket base 2010 and 2011 according to a recent article in USA Today. If these schools are experiencing declines in season tickets then it is likely that numerous smaller institutions without significant resources to spend on ticket sales are suffering to an even greater extent.

With traditional techniques no longer producing necessary results, it has become time for schools (and professional teams) to look at new solutions. Sports organizations have access to a significant amount of information about their fans, media, sponsors, and employees. For example, professional sports organizations have the opportunity to collect information about season ticket holders when they apply for tickets, from fans on their Facebook pages and Twitter accounts, or through promotions that include calls to action (i.e. when teams ask fans to text them the play of the game). Collegiate and high school teams also can have access to databases kept about alumni or current students. In addition, all organizations can buy information from third-party vendors who track the cookies (markers that track users internet histories and search patterns) to learn more about who is visiting a team or league site and what other sites these users view.

Simply obtaining this information is not enough for it to be valuable to sports organizations. Data mining and analytics allow sports organizations to identify audiences that are most lucrative to a team or league. This gives organizations the opportunity to discover which fans that are most likely to be interested in season ticket solicitations and most likely to make purchases. Organizations at all levels can purchase CRM solutions that allows them to aggregate all of their sales information in single application (such as B6A’s CRM Module) and complete basic analysis on which current and potential customers are most valuable to an organization. In addition, there are numerous organizations (like B6A) that can provide more sophisticated data mining and regression analysis to identify which factors drive sales and how much impact each factor has an organization’s bottom line. Once sports organizations identify which audiences are more likely to contribute revenue to a team or league, it can create customized messages in specific channels that increase the possibility of future sales. Microtargeting and data mining can make any sports organization much more efficient with its limited marketing dollars.

Despite all of these benefits, organizations must make it very clear to all audiences that are doing some form of microtargeting analysis with the data that customers knowingly or unknowingly provide. When American are informed about how data mining works, 86 percent want it to stop even if the data is used to make targeted appeals according to The Atlantic article. This number is misleading because people still routinely visit sites ranging from Facebook to barackobama.com which state that data can be used for the type of microtargeting and data mining described in this blog post. However, one Facebook’s biggest strategic missteps as a company was that it was not clear from the beginning what it is doing with user data. This is mistake that still haunts Facebook and has been one of the biggest blows to the company’s brand image. Sports organizations must be clear with their fans, media, sponsors, and employees that this type of analysis can and will occur with clear privacy statements on their websites or any area where they are collecting data.

While they may seem creepy to some, microtargeting and data mining are important tools that all sports organizations should have in their strategic arsenal.

Did The MLB De-Value Its Regular Season By Adding More Wild-Card Teams?

Most Major League Baseball (MLB) teams will play their open days games for their 2012-13 season this week. One of the biggest changes for this season is the addition of one wild-card team to MLB’s playoffs for both the National and American Leagues. Critics of adding new wild-card teams focus on two main faults with this postseason change. First, baseball is devaluing its regular season by adding more teams that can make the playoffs. Second, adding more wild-card teams would eliminate the drama that occurred at the end of the last regular season where the Tampa Bay Rays and the St. Louis Cardinals overtook the Boston Red Sox and the Atlanta Braves on the final day of the season to win the wild-cards for their leagues, respectively. The late seasons surges by the Rays and the Cardinals and swoons by the Red Sox and Braves were the central storyline of the 2011-12 regular season. With the addition of the new wild-card teams, the Rays, Cardinals, Red Sox, and Braves would all have made the playoffs making the regular season much less exciting.

Did MLB then devalue the regular season and make a mistake by adding a new wild-card team for each league? We realize we need to come up with a definition of what adds “value” to the regular season. For the purposes of this blog post, we wanted to use a definition of value that could apply to multiple sports in different leagues. One of, if not the only, thing that regular seasons across sports have in common is that teams compete to have a chance to participate in the postseason and win a league’s overall championship title. Therefore, we determined that regular seasons have a increasing value to a particular team based on the length of time it has an opportunity to compete in the postseason for an overall league championship.

The “for a championship part” is crucial because of the Bowl Championship Series (BCS) in college football. The BCS guarantees that only two teams will compete for the Football Bowl Subdivision (FBS) national title each season. Most of the supporters and opponents of the BCS agree that its main appeal is that this structure makes the regular season “matter”. No team has ever had more than two losses and competed in the national championship game. In most years, teams in the title game only have one or zero losses. Because a team has to win virtually all of its regular season games, BCS proponents contend this makes the regular season matter more in college football than any other major sport. In fact, the NCAA’s website states, “NCAAFOOTBALL.COM - Where every game counts.”

This contention and the website are wrong. In fact, the exact opposite is true as most game do not really “count” and the regular season means the least in college football than it does in any other major sport. By definition, virtually half of the teams in the FBS lose their first game of the year (it is not exactly half as some FBS teams play teams in the Football Championship Subdivision). Of the 120 teams in the FBS, only five teams had one loss and only eleven teams had two losses or fewer by the end of the 2011 season. This means that most FBS teams have no chance of playing in the FBS national championship game well before the season ends. This also means that much of the regular season does not matter for most FBS teams because they lose their ability to compete for a national championship so early in the regular season as compared to other sports. Therefore, the NCAA slogan should be “NCAAFOOTBALL.COM – Where every game counts for a very small percentage of teams.”

If MLB’s goal is to make the regular season “matter” more then adding more teams that compete in postseason (and thus compete for a World Series title) is the right decision. Not only do more teams compete in the playoffs but also more teams will be able to compete for playoff spots for a longer portion of the season. And its not as if regular season will lack for drama if previous seasons are any indication of future close finishes. At the end of the 2010 season, there was a one game difference between the teams that finished in second place (Boston Red Sox) and third place (Chicago White Sox) for the AL wild-card (i.e. the teams that would be competing for the final playoff spots with MLB’s revised playoff structure). At the end of the 2009 season, there was one-a-half game difference between the second place (Texas Rangers) and third place (Detroit Tigers) teams for the AL wild-card and an one game difference between the second place (San Francisco Giants) and third place (Miami Marlins) teams for the NL wild-card.

There could be many reasons to dislike adding playoff teams to MLB’s postseason but devaluing the regular season is not one of them.