As the summer comes to a close, I find it difficult to fully reflect upon my time in the DSSRF17 program. Each week, I’ve reflected on particular hardships and decisions made. For the final week, I’d like to promote the program that has consumed my tim over the last 8 weeks.

First, the program is an excellent opportunity to expose oneself to undergraduate research. In particular, the program allows students to investigate questions that on which faculty may not be focused. For example, my research was in hockey analytics, and my peers investigated topics in art history, Chinese economics, and LGBTQ+ representation in film. Individuals, with the help of our facilitators, were able to tackle research questions in all of these areas, no matter how unique the area of research may have been.

Another great aspect of the DSSRF program is the availability of professional help. Although individuals conduct the research, Bucknell faculty made themselves available, along with Bertrand Library and IT staff. The program facilitators dedicated their busy summers to providing any and all support we needed. From putting us in touch with industry professionals to giving tutorials in a variety of digital tools, Carrie and Courtney went out of their ways to provide us with the help necessary to conduct well document research. Research professionals and mathematics professors also contributed to the summer research, and I found everyone involved to be very willing to help. This program put me in touch with individuals who I plan to stay in touch with long beyond the DSSRF17 program.

Lastly, spending a summer in Lewisburg is great. The campus is quiet, and great camaraderie results. The same amenities available during the school year are available to students throughout the summer, and the weather is fantastic. A summer on campus is a great chance to explore the local area. I’ve floated down the Susquehanna River, sightseen both Harrisburg and Scranton, experienced minor league baseball games, played recreational sports, and tried many local restaurants. The weekends are free, so I had plenty of time to adventure through Central Pennsylvania.

I am envious of my peers, all three of whom are younger than I. Every other member of the DSSRF17 program have at least one chance to spend another summer conducting research in Lewisburg. And though they might not be involved in the same program in the future, the DSSRF program has allowed us all a chance to see how great the experience can be. I encourage anyone interested in research to apply, even if you have not identified a research question. The program facilitators will help you along the way, from identifying this question, to presenting findings to a plethora of Bucknellians eight weeks later. I am so thankful to have found this program before my time at Bucknell has ended, and I hope future Bucknellians have an opportunity to do the same.

As we wrap up week six of the DSSRF program, I finally feel I have made a unique contribution to the rapidly changing world of sports analytics. Though my model is far from Earth-shattering, I am confident such a model could be used as a jumping-off point to continue immersing oneself in sports analytics. Such a simple model created by a hockey enthusiast begs the question: Is sports analytics entirely beneficial?

Critics of sports analytics are quick to highlight a few examples of blatantly wrong analysis. For example, ESPN published a “state of the industry” site in 2015, examining every team in the major four sports leagues in North America: MLB, NFL, NBA, and NHL. Critics pointed out the “Bottom 10” and “Top 10.” These sections highlighted the franchises that embraced sports analytics the least (nonbelievers) and the most (all in) [1]. In the “Bottom 10,” ESPN shames teams for resisting the analytics movement. Though I am a supporter of the integration of analytics in sports, this language further widens the gap between “nonbelievers” and those who are “all-in.” If those who are against sports analytics justify their position with old-school methods, then shunning them for these beliefs is only going to drive them further away from number-based methods.

I think “The Great Analytics Rankings” is a great leaning opportunity for digital scholars such as myself. The material ESPN based their writing on was factual, included statistics (go figure) and incorporated direct quotes. However, the text between stats and quotes was dismissive and condescending. This language ought to be informative and possibly serve as an education platform rather than an opinionated jab. For example, ESPN objectively discussed the Colorado Avalanche and their lack of acceptance when it comes to advanced analytics: “Last season, the Avalanche were Central [division] champions, however their Corsi for percentage, which indicates productive puck possession, was 25^th in the league, indicating their success would not last” [1]. However when discussing the NFL’s New York Jets, ESPN wrote: “New GM Mike Maccagnan and new coach Todd Bowles likewise sports old-school credentials and were not hired to spearhead a stats awakening” [1]. ESPN proceeds to mention the “good ol’ eye test,” as if there is no way to scout players outside of advanced biological and performance analysis. Putting down the Jets for neglecting advanced analytical methods must be justified, but hiring individuals with different opinions is not reason enough.

When conducting any sort of work in the Digital Humanities, communication is key. Though I spent the majority of my time this summer working with numbers, the language used to describe my final project must be informative and open-minded. Though analytics has barged its way into the sports world over the past 15 years, other outlooks regarding decision making still exist, and each has its own reason to be appreciated. The next two weeks will be spent fine-tuning my final site, and I must keep in mind the importance of providing information clearly, not forcing research findings on readers.

[1] “The Great Analytics Rankings.” ESPN.com. Accessed July 6, 2017. http://www.espn.com/espn/feature/story/_/id/12331388/the-great-analytics-rankings.

This week our facilitators asked us to write about some of the challenges we’ve faced through the first 5 weeks of the DSSRF program. I will focus specifically on the struggles encountered in the fifth week. That is not to say I didn’t have issues through the first four weeks, but this week has been particularly challenging, and the issues arising this week are still being dealt with.

The largest concern this week involved the model I have created. The model, though fairly accurate in my trials, seems to be accurate by luck. An advisor to the project has informed me that he can identify at least two crucial facets of the sports that are left out of my model. In other words, my model doesn’t account for certain teams’ strengths and weaknesses, and thus will prove to be not accurate when using past or future data. Though serious, I’m optimistic that this issue will be resolved. I have worked to incorporate the missing portions of the game of hockey into my model by adding to my data. This has forced me to return to the very primitive steps of the project: testing relationships between individual metrics and team success. As tedious as this may be, the process is crucial in establishing a sound model. Simply creating a model based on observed causes of team success is insufficient to say the least. This process has slightly derailed the week’s already busy schedule, but by the end of the upcoming holiday weekend, I fully expect to be back on track.

A large portion of the time spent as a group this week was dedicated to learning Scalar, a site creation tool tailored to the needs of Digital Scholarship projects. Scalar allows me, the creator, great freedom in design, while consistently producing a high quality site. The Scalar projects we observed last week at the DSSF Meet-Up at Bryn Mawr were professional in appearance, and rich with information. This week, I struggled to find this balance. “Not too wordy.” I told myself. Yet I felt I needed the entirety of 4 pages to contains the methodology of my project alone. The help of others was desperately needed, but it took me hours of struggle to realize this. The keen eyes of my peers were the keys I needed. No matter how many times I reread my work, I didn’t catch the numerous repeats of information, and the poorly worded explanations scattered throughout the text. After being identified by my peers, these flaws were corrected, and my methodology was condensed to a mere one and a half pages. Though I anticipated a struggle to mold Scalar to the form I wanted, the hardest challenge was keying in on the material to include on the Scalar page.

As in any project, I’ve faced challenges throughout the process. This was expected. The solutions however, have not been. Each has involved the minds of others: peers, facilitators, advisors, and even family members. This week was another stark reminder that although I tell people I’m conducting independent research this summer, the final project will be credited to many people beyond myself.

Jarome Iginla for Joe Nieuwendyk. Wayne Gretzky for Jimmy Carson and others. Trades like these permanently altered the NHL landscape in the late 20^th century, and today, hockey professionals would unanimously agree that one team faired better than the other in both of these trades.

Subban for Weber, Hall for Larsson, and Johansen for Jones. These noteworthy NHL trades have all taken place in the past 18 months. Though immediate results can be cited, the “winners” of these trades are hard to name, as all six of the players listed above are still playing at a high level. Until recently, only time told which teams made out better in personnel exchanges, but the ongoing development of advanced hockey analytics is changing this.

Made famous by Billy Beane and the improbable success of his Oakland Athletics, number-supported personnel decisions are changing the way sports franchises are run. Though only recently introduced to the hockey world, respected voices in hockey have already voice their support for hockey’s newest asset. At an end of season press conference in 2016 Bill Peters, head coach of the NHL’s Carolina Hurricanes and former coach of Team Canada’s Under 18 team said about the use of analytics “It’s all about finding ways to be more successful. I embrace it” (Smith). Though analytics is now respected in the NHL, very few predictive tools have been created.

Hockey analysts such as Emmanuel Perry have created single-value metrics that determine a player’s worth, while others have focused on offensive and defensive efficiency metrics. Though both of these tools can be helpful in assessing the potential outcome of personnel decisions, I believe creating a model to accurately predict team performance given personnel decisions is the next step in hockey analytics. This project is intended to allow fans and hockey professionals alike the ability to assess future team success given front office decisions.

Creating a successful model will have a multitude of uses, including but not limited to helping guide personnel decisions, providing comparative evaluations of individual player contributions, and gambling applications.

Will the Montreal Canadians regret losing PK Subban? Was acquiring Adam Larsson really worth losing Taylor Hall? We may find out sooner than expected which of these teams faired better, and which teams might be haunted by these trades for years to come.

Smith, M. (2016, April 12). Francis, Peters Digest 2015-16 Season. Retrieved June 10, 2017, from https://www.nhl.com/hurricanes/news/francis-peters-digest-2015-16-season/c-878539