Download the article and view the data below:
article.docx |
sabermetrics.xlsx |
It’s the Midsummer Classic, the highest-rated all-star game in sports; in just 24 hours the American League and National League will faceoff in the greatest nine-inning exhibition of the year. The best of the best will be there. Perennial favorites like Albert Pujols, Miguel Cabrera, and Prince Fielder, among others, made the cut yet again. Reigning MVP, Mike Trout, will be making his 4th consecutive appearance and 3rd straight start this year. Fellow superstars, Bryce Harper and Giancarlo Stanton, were each selected for their 3rd game as well. Josh Donaldson smashed the all-star vote record, receiving an unprecedented 14 million of them, and his 2nd straight appearance. 29 guys will be making just their first all-star appearance as well. It will certainly be an incredible night for baseball, and always an important one (the winning league receives home-field advantage in the World Series). Download the article and view the data below:
2 Comments
Using a linear regression model with data dating back to 2005, there appears to be a statistically insignificant relationship between the number of homeruns a player hits during the regular season and the homerun derby. However, with the addition of a second round in 2014 and then completely overhauling the format in 2015, we might see a shift in the type of people who win the derby. Batters only have 5 minutes to swing per round. This means batters will be swinging as hard as they can and as many times as they can. We might even see some players break a sweat. My prediction: the 2016 rule change will require contestants to juice up #letbarryplay
Now that the playoffs have concluded, many teams are faced with the challenge of refining, or in some cases, completely rebuilding their teams with the hopes of achieving what the Chicago Blackhawks have done thrice in the past six years: winning the Stanley Cup. Many GMs will try to improve their teams in a variety of ways, either via the NHL draft, trades, or free agency. However, for fans and analysts alike, the NHL offseason is predominantly a period for speculation on what moves will be made, how newcomers will fit into their new teams’ system, and what the next season will hold. So instead of using this time to predict what will happen in the coming months, I’d prefer to use this as a period for reflection. Read the full article by downloading the word doc below.
With the 2015 NBA Draft approaching, executive board members Austin Feinstein and Grant Goldman took a look at this year's top prospects. Below are their predicted top 10 picks. #1 - Minnesota Timberwolves Karl-Anthony Towns Karl-Anthony Towns, the predicted first pick of many NBA analysts, will almost certainly go to Minnesota at number 1. Towns, while limited in minutes on a deep Kentucky squad, showed great promise shooting 56% from the field and 81% from the free throw line. Furthermore, Towns showed his value grabbing nearly 19% of all rebounds while on the court. #2 - Los Angeles Lakers Jahlil Okafor Standing at 6' 11" and 272 pounds, Jahlil Okafor has an NBA ready body that can keep up with the dominant big men in the league. The Lakers have won many of their championships on the backs of names like Shaquille O'neal, Wilt Chamberlain and Kareem Abdul Jabbar. Will Okafor be the next center to give the purple and gold a ring? He certainly has the tools to do so, Jahlil draws fouls at 5.1 FTA/game and converted 66.4% of his 11.1 FGA, to average an impressive 17.3 points. With the impressive offensive tools comes defensive faults, Jahlil is a limited athlete which hinders his ability to guard the pick and roll and more mobile centers while he struggles at the line shooting just 51.1% #3 - Philadelphia 76ers Kristaps Porzingis Sam Hinkie has proven through the previous two drafts that he is willing to take a risk and willing to be patient, very patient. Kristaps might be the most intriguing prospect as he stands 7 feet 2 inches tall with an enormous 7 feet 6 inch wingspan, while managing to shoot 36.7% from NBA 3pt range. Simply watching Porzingis' film, it is easy to see the high potential of the 19 year old Latvian. The largest concern is whether he can keep up with the power forwards of the NBA as Kristaps has a slim frame at 220 pounds. #4 - New York Knicks D'Angelo Russell Phil Jackson may surprise us here and trade out the highly coveted #4 spot, but if he decides to stay put D'angelo Russell is a no-brainer with the three big men off the board. Russell is a prolific scorer averaging 19.3 PPG while shooting 41.1% from 3 on 6.6 attempts per game. He can create his own shot using his impressive ball handling skills and has the playmaking ability to run an NBA team as evidenced by his 5 assists per game. D'angelo does lack some explosiveness that NBA fans have grown so accustomed to with Russell Westbrook, Kyrie Irving and Derrick Rose (ok...maybe 3 years ago). D'angelo will have to make up for the lack of athleticism with his craftiness and basketball IQ if he wants to be an elite point guard. #5 - Orlando Magic Justise Winslow Orlando has a young, dynamic and exciting core in Victor Oladipo who was drafted 2nd overall in the 2013 draft, Elfrid Payton who was acquired in a draft night trade with Philadelphia during the 2014 draft and Nik Vucevic who was also acquired from Philadelphia in the blockbuster Andrew Bynum deal. The final piece to this young eastern conference team could just be the elite wing out of Duke, Justise Winslow. Winslow has the size to guard the SF position in the NBA but the quickness and lateral speed to cover the top 2-guards. He is a ferocious finisher and lock down defender. Justise will have an immediate impact in the league and just may be the most NBA ready prospect. #6 - Sacramento Kings Emmanuel Mudiay Emmanuel Mudiay, following the same path as former lottery pick Brandon Jennings, chose to forgo college and instead joined the Guangdong Southern Tigers of the Chinese Basketball Association. While statistics are certainly available for Mudiay's play, it's difficult to read into them, given he only played 12 games. In those 12 games he shot only 57% from the free throw line and 34% from behind the arc. In the event that the Kings deal their star center DeMarcus Cousins, Mudiay will be able to pitch in as an above average rebounder (17% Defensive Rebound Percentage with Guangdong). Look for Sacramento to take the quick, physical point guard on draft night. #7 - Denver Nuggets Willie Cauley-Stein Kentucky big man Willie "Trill" Cauley-Stein has proven to be an excellent defender at the collegiate level. Cauley-Stein has been compared to the likes of Tyson Chandler, while being touted as an excellent athlete. While his block percentage is just over 7%, (down from 12% in his sophomore season) Cauley-Stein is less likely to receive help in the NBA, as he will not be playing alongside Karl-Anthony Towns. Cauley-Stein is an elite defender in isolation, with opposing isolation FG% of 22% according to ESPN. #8 - Detroit Pistons Stanley Johnson We are excited to see what Stan Van Gundy will do in his second offseason as President and Coach of the Detroit Pistons, a team in need of, well, talent and some resemblance of a plan. The pistons roster features one of the more exciting young centers in the league in Andre Drummond and a Brandon Jennings, a scoring but inconsistent guard. Detroit is in need of a wing and no player in the draft fits with the Motor City's blue collar, tough mentality as well as Stanley Johnson. Johnson is the type of guy that you can put on the opposing team's best player and not worry for the rest of the game. Standing 6'7" with a near 7 foot wing span in addition to his combination of strength and speed allow Stanley to be a ferocious defender and a fantastic rebounder as evidenced by his 6.5 Rebounds per game. Concerns of his ball handling are fair and it is a possibility that Johnson will never be a number 1 or 2 option on a winning team. #9 - Charlotte Hornets Mario Hezonja Mario is the second international prospect (excluding Mudiay) in the top 10 of Wash U Sports Analytics' mock draft but he is just as intriguing as Kristaps. Mario has excellent size for the shooting guard position standing 6'8" tall and has shown promise from deep shooting nearly 38% on 6.7 attempts (per 36). Hezonja has a great first step and the confidence and ability to be a great scorer in the NBA, something that Charlotte desperately lacks as their offense ranks 28 in offensive efficiency with a score of 97.6. Hezonja needs to work on his man-to-man defense, he has the size and athleticism to be a much stronger defender but has not shown the desire to at this point in his career. #10 - Miami Heat Frank Kaminsky Miami will look to draft a big, Frank Kaminsky in particular, with the 10th pick. One reason is that their current center Hassan Whiteside is only under contract for the upcoming season and Miami might shy away from paying big money to the break-out center with Chris Bosh set to earn upwards of $20 million in each of the next four seasons. Not only can Kaminsky rebound the ball, but he is also able to spread the floor. Shooting nearly 42% from three in his senior year with Wisconsin, Kaminsky's range could help the Heat as they shot 33.5% from deep, good for 24th in the league. Bonus: This great photo of Frank Kaminsky 2015 NBA Finals Preview Golden State Warriors (W1) vs. Cleveland Cavaliers (E2) Game 1: June 4th, 2015, 9 PM EST, ABC It’s finally here. The dream matchup that NBA fans were hoping for, with the two faces of the NBA squaring off for the Larry O’Brien Trophy. Before I delve into the matchups of this series and give our predictions, let’s first assess how we got here. Cleveland and Golden State have followed similar paths to get to where they are now in these playoffs. Both teams swept their first round matchups, as the Cavs breezed by the overachieving Boston Celtics and the Warriors defeated the New Orleans Pelicans. Each team then faced their toughest challenge in the second round, with Cleveland beating the Chicago Bulls in 6 games in a back and forth series that saw 2 game-winning buzzer beaters from former MVPs, while the Warriors won against the tough, defensive Memphis Grizzlies in 6 games as well. In the Conference Finals, the Cavs again swept their opponent, this time the #1 seeded Atlanta Hawks, while the Warriors beat the Houston Rockets in 5 games (in a series that likely would have ended in 4 games had it not been for a monstrous performance by MVP runner-up James Harden). Now, the Cavs are looking to bring a championship to a city that hasn’t had a sports franchise win a championship in over half a century (more on that later), while Golden State is aiming for their first title since 1975.
Is There a Relationship Between Ejections and Instant Replays in the MLB? By Robbie Steirn5/23/2015 Now that there is instant replay in the MLB, one would think that ejections would decline. However, this is not the case. Ejections increased from the 2013 to 2014 season. It is too early to tell if instant replays adversely impact ejections. We would need a few more years to accurately prove a relationship between the new instant replay rules and ejection frequencies. Once you start peeling the layers of all the ejection data dating back to 2008, there is a bigger story to be told which many analysts have looked over. While ejections have increased since the start of instant replays, I believe we will see a strong decline in ejections for two main reasons. First, the number of ejections made at first base declined from 15 percent to 9 percent. Ejections are less likely to occur at first base due to the introduction of instant replays, since replays decrease the chance of umpires making an incorrect call. So why are were there more ejections made last season? There are two explanations: an unprecedented increase in pitchers being ejected and an increase in ejections due to replays. Pitchers got ejected more last season than in recent history because of intentional hit by pitches. Because this data has no relevance to instant replays, then pitcher ejection rates will decline in future years which means that ejections will be lower in a world of instant replays than without them. Second, many ejections occurred from instant replays themselves. There were some replays that were blatantly incorrect. Even people who take second looks can make the wrong decision. Because 2014 was the first year instant replays were introduced, we will see a decline in ejections related to instant replay due to a steep learning curve. Those who look at instant replays from abroad will make better decisions, while coaches and players will get more assimilated with the new rule. Here are some additional fun facts: Observation 1: A player is about 25 percent more likely to be ejected if they are loosing the game. Observation 2: A majority of ejections (52 percent) are from balls and strikes. The introduction of replay review decreased safe/out ejections by 19 percent. However, ejections from replay review are about 12 percent of all ejections. Observation 3: Ejections increase by 20 percent during the second half of the season. Observation 4: Ejections are most likely to occur in the 7th and 8th inning, 17 and 16 percent respectively. Observation 5: Out of all ejections, mangers get thrown out 48 percent of the time, pitchers and catchers at 11 and 5 percent respectively. However, pitchers were ejected at a record high last season, which can be attributed to more hit by pitches. Observation 6: When evaluating whether an umpire makes the correct call on a play that ends up leading to an ejection, umpires are correct 53 percent of the time, incorrect 21 percent of the time while the remaining 26 percent is indeterminable. Observation 7: Looking at data since 2008, there was not enough evidence to determine a relationship between a team’s win percentage and the number of player ejections (p value of .48).
At this point, we’ve reached the halfway point of the second round of the NBA playoffs, with 14 of a potential 28 games completed. If not for Paul Pierce’s buzzer beating bank/”I called GAME” shot, the Hawks-Wizards series would be arguably the most boring in recent memory. Meanwhile, after the Clippers’ complete domination in Games 3 and 4 (both wins by over 25 points), the Rockets look all but defeated. Memphis/Golden State still holds a lot of intrigue but they haven’t completed Game 4 yet, so I’ll hold off on looking at that series. For now, we’re going to look at what has been going on in the most compelling series of the Conference Semifinals thus far: the Cleveland Cavaliers versus the Chicago Bulls.
Cleveland Cavaliers (2) vs. Chicago Bulls (3) Series: Tied 2-2 Next game: Tuesday May 12th, 7:00 PM
We researched the correlation between EA Sports video games FIFA 13 and FIFA 14 and the results of the top 5 leagues in Europe over the last 2 years. The videogame FIFA gives each team ratings for offense, midfield, defense, and gives an overall star rating and we compared these to the final statistics in each league. The leagues that we gathered data for were the English Premier League, Bundesliga, La Liga, Serie A, and Ligue 1. By plotting this data and fitting it with a best-fit line and an r2 value, we found a few key correlations between the video game and real life. The strongest correlation existed between points for all of the leagues besides ligue 1, as a function of defense and midfield. By excluding ligue 1 and offense in our model, we ended up with an r2 value of .658. This is surprising as it suggests that having better rated forwards means little to how well a team does over the course of a season. Using weights assigned to defense and midfield we project that the overall point total for a team playing a 38 game schedule (Bundesliga plays 34 games a season so we weighted their point total to reflect 38 games) follows the form Points = -244.8511 + 2.0671*Midfield + 1.9066*Defense. We took these projections and compared them to the current season even though the seasons aren’t over. We compared current place in standings to the projected standings of each league given their projected point totals. Of the four leagues, La Liga and the Premier league are closer to the predicted table averaging within 2.35 and 2.5 places respectively of their projected finish. The Bundesliga and Serie A are not as close to predictions as they average within 3.6 and 3.75 places respectively.
Results: In soccer, the most common method of determining the value of a forward (or striker) is by comparing a single statistic, goals scored. While this metric certainly constitutes the majority of a forward’s worth, that player contributes far more to their team than just goals. The purpose of this project was to consider multiple components of a forward’s performance, and then compress those statistics into one number. Finally, the score is adjusted for the number of minutes played by the forward. The score is intended to reflect both the raw contribution of a forward and the efficiency with which they perform. We recognize that the coefficients for calculating a forward’s raw total score have been subjectively determined; however, those coefficients were selected so that the average player’s value derived from each category reflected a desired distribution. The coefficients ultimately produce the following distribution of average score derived from the specified category: - Goals: 50% - Assists: 17% - Passing: 8% - Key Passes: 11% - Dribbling: (8%) à Negative due to the formula (Successful Dribbles – Dispossessed) - Offside: 3% - Yellow Cards: 1% - Red Cards: < 1% - Fouls: 0.5% - Shots on Target: 12% - Shot Conversion: 13% These percentages do not directly add up to 100% since some were deducted and others were added. Rather the percentages represent the average addition/deduction relative to the total score. Comments: The highlighted names indicate forwards that we believe, personally and subjectively, did either surprisingly well (green) or surprisingly poorly (red) in our analysis. - Marco Reus: An incredibly talented player, but a name most would not put in a top 5 list with Messi and Ronaldo. He led the list in key passes, ultimately making up nearly 20% of his value. Additionally, his assists (13) made up more than a quarter of his score. - Daniel Sturridge: His score accurately reflects an incredible season for this young forward, finishing second in the scoring charts in the 2013/14 EPL season. He also benefited from a relatively high shot conversion of over 20%. - Karim Benzema: While none of his numbers were noticeably poor, his dribbling reduced his score by more than 8%. His low ranking is primarily attributable to fairly low involvement in gameplay. - Diego Costa: Though a prolific scorer and target man for Atletico Madrid in the 2013/14 La Liga season, he did not do much to set up his team mates. Additionally, he suffered tremendously from a very poor dispossession record, losing more than 20% of his value to poor dribbling. - Edinson Cavani: Similar to Benzema, Cavani did not score too impressive of a number of goals in the 2013/15 Ligue 1 season. The Uruguayan also did very little for his teammates, recording both low numbers of assists (2) and key passes (23). - Radamel Falcao: Perhaps a difficult assessment of his ability given his injuries last season. However, the Colombian forward did little in the way of setting up opportunities, failing to record even a single assist in his 17 appearances. - Olivier Giroud: Though he recorded a respectable number of goals and assists, Giroud had one of the worst dispossessed to successful dribbles ratio. His low shot conversion rating and poor passing accuracy both hurt his score.
Introduction: Now that there is only about one quarter left of the 2015 NBA season, analysts are trying to predict the next MVP winner. ESPN and Basketball References created models to rank the top MVP candidates. However, credible news sources like USA Today argue that it is impossible to not only create a model that can accurately predict the next MVP winner, but also generate a correlation significantly greater than 50%. Analysts believe that there are too many unquantifiable variables that influence voters. Each player has their own narrative that appeals to voters differently. Over the past few years, one can accumulate a total of 1,250 points. However, this number changes depending on how many people vote from the media. The media ranks players from 1-6 and a greater weight is assigned to those who rank higher. For example, if a voter ranks Lebron James as number one and Kevin Love as six, Lebron James will receive 10 points while Love will only receive 1 point. My research started by collecting data on NBA players going back to the 1994 season when Hakeem Olajuwon won the MVP. Michael Jordan and Karl Malone battled for the first place for three years until Shaquille O’Neal won in 2000. He almost became the first and only player to win the MVP unanimously, but one voter voted for Allen Iverson who finished 7th. The only other player to only lose one first place vote was Lebron James thirteen years later. Results: We predict that Stephen Curry will win the MVP with 844 votes with James Harden in second with 696 votes. Russell Westbrook, Lebron James, and Chris Paul will finish 3rd, 4th and 5th, respectively. While Chris Paul had a better season than Lebron James, we don’t believe that the media will vote in accordance with the statistics. To run the model, I used SPSS to create an optimal model with backwards elimination as well as StatPlus in excel to generate a multiple linear regression model and residual plots of our independent variables. These variables consist of USG%, WS/48, BPM, VORP, Win-Loss Record and Competition. The model as a whole is statistically significant with a p-level near zero. Also, we have an R-value of .83, an adjusted R Square of .67, and a Se/ybar of .92. Thus, it is accurate to conclude that we have a credible model that shows a moderately strong correlation between our independent variables and the number of MVP votes received. Methodology: I will explain my thought process when constructing this model starting at its inception to the final product. Initially, I collected data from basketball references and ESPN going back to 1994. I started experimenting with linear regressions but couldn’t find anything significant. There were two reasons for this. First, the population size wasn’t a normal distribution since it included every player instead of only players who received at least one vote. However, when I just looked at players with at least one vote, the model couldn’t accurately predict MVP votes due to high variations among players. Additionally, those who receive only one vote are mostly due to chance. It could be that a voter ranked a player sixth to create noise or perhaps they went to the same college. Second, the variables I was using were basic statistics ranging from points to rebounds to blocks. However, I soon realized that more advanced metrics had better results. Instead of going back to the 1994 season, I only used data going back to the 2008 season and excluded the 2011 season. I wanted to go back to the 2008 season because voting and rules have changed prior to that year. I also had to change my data in the 2012 season because of the lock out. I also excluded the 2011 season because there were too many outliers in the data so the must fair way to handle it was to categorically remove the year. Lebron James finished third behind Derrick Rose and Dwight Howard even though his statistics were better. The reason he finished third was because everyone hated him after he left Cleveland. This is where the narrative can sometimes come into play. However, I do believe that a credible model can sufficiently predict MVP votes. Now to talk specifically about the model I created. Each independent variable that I use is straightforward. I looked at a vast array of complicated and advanced metrics, but found the combination of USG%, WS per 48, BMP and VORP yield the best results. Also, I included a win-loss record and a competition variable. The competition variable rewards those who have a VORP AND WS per 48 that are 1.5 standard deviations greater than your average player who receives at least one MVP vote. I chose these variables because they had the greatest impact on the dependent variable of MVP votes, which I determined before including the competition variable. Players get a 1 for having either a VORP or WS per 48 1.5 standard deviations greater than the mean or 2 points for having both. There are two things to note when creating the model. First, VORP and BMP have an extremely high multicollinearity. When I remove those variables, the model is slightly worse. I was able to conclude that by adding the two together, it makes the model more accurate. This can be due to over fitting or the advance metrics working together to show a player’s skills. Second, I had to create a multiplier to adjust the number of votes. The data might be slightly skewed because Lebron James wins almost every MVP with Durant typically at a close second. However, this year, there isn’t one player putting up similar numbers to MVP winners in the past six years. Adding a multiplier that looks at voting trends of the past six years, I was able to adjust the number of votes to accurately reflect historical media voters. ***The excel document can be manually edited using a Mac. If you have a PC, you can alter the tabs using data references to pull from online sources and set up macros to update automatically. If you have a Mac, all you have to do is update tabs that are highlighted in red. The document runs from right to left. Follow the instructions below: 1) Go to tab labeled “UPDATE Advance Stats”. Cell A1 has the link to the website which you will use to copy and paste data 2) Follow the same steps on “UPDATE Regular Stats” and “UPDATE Team Record” 3) Go to “Real Time Filtered” a. Click Cell G1 on the filter icon. b. Click Clear Filter c. Click on Cell L1 on the filter icon. d. Click Clear Filter e. Click Cell G1 on the filter icon f. Set greater than -.5 g. Click Cell L1 on the filter icon h. Set greater than 0 i. Finally, copy and paste values into the “MVP Votes” tab 4) Go to “MVP Votes” a. Click on the filter to have the MVP votes descend to see the final ranking
In the modern NFL, the offense is ruled by the passing game. There have been more quarterbacks to throw for over 5,000 yards in the 2000’s than the history of the NFL before that. It seems as if the passing game, not the running game or the defense, is the focus of NFL teams. But do they lead to wins? In our investigation, we wanted to determine if the passing game actually has the biggest impact on wins, or if another aspect of the team contributes more. In order to determine the factor which contributed most to wins in the regular season, we created an Excel model which used linear regression to analyze the influence of multiple team factors on their regular season record that year. Our dependent variable was the number of wins for the team, and our independent variables included ESPN’s Quarterback Rating and Football Outsider’s Rushing, Passing, Rushing Defense, and Passing Defense rating for the previous six years. Our R-Squared value was 0.746 and our Multiple R value was 0.864 which means that the data is statistically relevant. Additionally our Standard Error was 1.582 which means that the data is reliable. Since our analysis had these error and R-values, it meant that there was significant correlation between our independent variables and wins.
First, the two statistics which were important for our data was the p-value and the absolute value of the coefficient of each of the independent variables. We wanted the independent variable to have a low p-value and a high coefficient. The independent variable which fit this criteria the best is passing defense rating, with rushing defense next, then passing rating, rush rating, and then QBR. This means that if a football team is interested in maximizing their wins, they should invest in their passing defense the most.
Although this may seem surprising, given the direction that the NFL is heading, this assumption that passing defense is the most important factor for a team’s wins makes sense because even though quarterbacks are able to produce at an all-time high, the best teams are the ones who are able to stop these offenses. The key for teams is not to passing for the most yards, or pass for the most touchdowns, but to be the team which is best at stopping those plays. Evidence for this is even apparent in this year’s Superbowl. The Seattle Seahawks are known for their Legion of Boom, perhaps the best secondary in the league, and the Patriots, coming off a loss to the Broncos in the 2013 AFC Championship, invested heavily in their defensive backs with the signing of Darrelle Revis. The teams with the best secondaries are playing for the best prize in football. As a result, NFL teams should use this information to help them decide how to best build and invest in their teams. The Philadelphia Eagles, although mired by controversy over their decisions this offseason, have made decisions which improve their defensive passing game. They signed Byron Maxwell of the Seattle Seahawks, the top corner in free agency, to a six year, $63 million contract with $25.5 million guaranteed, and even though many analyst have called this contract overzealous, these analysts may be underestimating the importance of the passing defensive and the larger impact it has on wins compared to other team positions. The Eagles and Chip Kelly know what their team needs to win, and in a tumultuous NFC East, they should be the favorites to win the division. Finally, in less than one month, the NFL draft will be held in Chicago, and all 32 NFL teams will have opportunities to build their team for the future. There are star quarterbacks, star defensive linemen, and star wide receivers, but according to our study, the teams who end up being the most successful in the next few years will probably be the ones who shy away from the flair of the offense, and find their Earl Thomas or Richard Sherman. To read more about our adventures at the SSAC head to:
http://olinblog.wustl.edu/2015/03/sports-stats-club-sizes-up-the-pros/ WU STATS Analysts (Juan Espinosa, Michael Aronson and Robbie Steirn) researched the correlation between the management practices of Ecuadorian soccer clubs and team success on the field in 2013 and 2014. To perform the study, we accumulated data to evaluate several factors controlled by a soccer club's management including team infrastructure, youth development programs, on-time payment to players, transfers to Europe, fan engagement and the leadership behind the board of directors. We performed a hypothesis test using the simple linear regression model for a small sample. The STATS Analysts concluded that there is a strong positive linear relationship between our ranking system and team rankings in 2013 and 2014 for Ecuadorian soccer clubs. http://www.larepublica.ec/blog/deportes/2014/07/17/ecuador-futbol-ranking-administracion-gestion/ http://independientedelvalle.com/2014/07/idv-el-mejor-equipo-del-ranking-de-administracion-y-gestion/ http://news.wustl.edu/news/Pages/27143.aspx
|
Analysts
All
SportsArchives
December 2017
|