Team chemistry study conclusion: Let's get defensive

What do you do when you're stuck on a shitty team and you can't leave? 

Comparing yourself to a better team's performance won't help your self-esteem; in fact, it might make you feel worse about your team and its prospects. But, what if, instead of focusing on wins and losses, you emphasize nonperformance aspects like academic performance, good looks, or general awesomeness? Research suggests that these comparisons should engender team-wide feelings of positivity, leading to a greater sense of identity, plentiful team chemistry -- and, ideally, increased performance.


In our last article, we gave a midseason report on the intervention that we were in the midst of conducting aimed at implementing some of these ideas. With the overall goal of increasing these teams' performance, we built a team chemistry intervention, and administered it to three Southern California college football teams via a series of surveys throughout the season. 

Our initial glance at the data showed promising returns -- a slight trend on offense, with an increase in points scored per gameand a stronger trend on defense, with a decrease in points allowed per game. Now that all the results are in, we thought an update was in order.

A little of this, a little of that

After 5 rounds of treatment, the final results are a mixed bag of the study are a mixed bag. First, the offensive trend seems to have washed out: all three teams' points scored per game were on par with their four year averages. 

More encouragingly, the defensive trend became a bit less dramatic, but remained consistent: Two of the three teams that participated in the study showed four year lows in points allowed per game. 

Some is good, more is better

These results represent an encouraging start. But to really increase our confidence in claims about team chemistry leading to better performance, we'll need much more evidence. Fortunately, that's exactly what's in the works -- just like Kahneman and Tversky, we're looking to be sports analytics disrupters

We're using team chemistry to increase performance for a few college football teams

There’s an old adage in psychology that says, “the best predictor for future behavior is past behavior.” This formula pretty much sums up how sports analytics works: analytical minds compile players' previous performances, and use them to build models to predict how they’ll perform in the future. Though this method seems to be pretty good, it’s far from the whole story.

Rob Arthur's recent piece on 538 was an amalgamation of some of the more reliable baseball prediction outlets (PECOTA, Vegas lines, and Fangraphs projections).
Together, the predicted win totals from these sites correlate well with the actual win totals of most baseball teams (r = .60). If we do a little math, we find that previous performance explains about 36% of the reason why baseball teams win as much or as little as they do. Now, if we knew 100% of the reason, we could flawlessly predict the number of wins for each baseball team; unfortunately, we've only got 36%. The result is that something other than previous performance accounts for around 64% of baseball team's performance1.

So the question becomes: Can we measure other stuff to add to the already awesome previous performance prediction models to increase our predictive power? If we said "Hell yes, you can!" would you call us crazy? You wouldn’t be the first. 

The future is now

First off, psychology as a discipline is quite large. There are plenty of sub-topics within psychology that can be brought to the professional sports landscape to help predict future performance. That said, research has time and time again shown that chemistry and performance are related. So, for this reason, we've decided to focus on this often talked-about (but not-so-often accurately measured) component of a team’s ability to stick together. 

Utilizing over 30 years of research, we designed a survey to measure and increase team chemistry; we are currently implementing it with a few community college football teams in Southern California, with the aim of increasing their performance. After 3 rounds of the survey, here’s a midseason report of how it’s going:
Each team has played 6 games thus far with an average opponents winning percentage at 47%
Each team has played 6 games thus far with an average opponents winning percentage at 47%
Looking at the 4 year averages, the overall trend indicates that increasing team chemistry has played some role in improving these teams' offenses and defenses. Before you get your scientist hat on, we know these findings are preliminary: It's a small sample of schools -- perhaps these teams would be performing this well without our survey2. There's also another 4 to 5 games left to play for each team, so things could change by the end of the year. However, it's a nice enough trend to get us excited about the intervention. Stay tuned for an end-of-season report to see if the results hold.

The million dollar pitch

Previous research in team chemistry has often occurred within college teams, so the mid-season results of this study are by no means revolutionary. What could be revolutionary is if professional teams begin to use this research: monitor players, build a team with players high in team chemistry, and design interventions to increase overall team chemistry. The idea is to chip away at that missing 64% of future performance that previous performance doesn't capture.

1 This assumes our predictive models built using previous performances can’t improve, right? We're capturing about 36% of the reason (or variance) of future performance right now in 2016, what's to say we can't get better in the future? True, the correlation coefficients have ranged from just over r = .8 to just below r = .3. So, at most we can unreliably predict upwards of 64% of the variance on a good year, which still leaves 36% unexplained. Predictive power has varied over the last 20 years, but the average guess is that previous performance only eats up around 36% of future performance.

2 Each team was randomly chosen and had to be a low-performing team in past.

Call for Articles

Dear Colleagues,

My name is Dr. Benjamin Rosenberg and I am the managing editor of Burkmont.com. Our aspirations are to bring sports psychology research to the applied realm. If you have published sports psychology literature in a peer-reviewed journal, we would like to feature it on our website. We feel sharing your work in this way would help your work to reach people beyond the scholarly community – namely those for whom your work has practical application. If you have not published your work, but are looking to promote your research to an applied audience, we'd love to hear from you as well. 

If you’re interested, please follow the link below and complete the brief (4 question) online form. From there, we’ll create a short, colloquial version of your article to share with a new audience. Additionally, we will include a link to your original journal article (assuming it's already published), and can also create a link to your website or any other supplementary material. We could also feature you in a profile, or in any other way that seems relevant to you.

If you have any questions, please feel free to contact us!

Thank you for your time,
Ben Rosenberg, Ph.D.

Sloan 2016: What's Next?

The conference was advertised as bigger and better, with the former being certainly true. There were more talks, more attendees (roughly 1,000 dorks and 3,000 wannabe dorks), and even longer lines for the men’s restroom than last year. Naturally at the tenth anniversary of the Sloan Sports Analytics Conference, you start things off with a Moneyball reunion: author of the book Michael Lewis, Godfather of sabrmetrics Bill James, right-hand man Paul DePodesta, and of course Billy Bea……hey, where’s Billy?

Other highlights include more tennis analytics, more blabbering about machine learning from people who can’t describe it, Nate Silver talking about politics instead of sports, and more of the same from previous conferences. Even NBA analyst Zach Lowe said, “What’s next?” Sports analytics are a nice reminder of how a novel idea can blossom into a beautiful epidemic, but when does the analytics movement become the very thing it sought to defeat in the first place – dogmatic tendencies, homogeneity, and close-mindedness? There’s something that I think needs to be mentioned: it’s not the teams or organizations that drive the industry, it’s the people who can change the game; it’s the Sloan attendees who can uniquely brand themselves and melt this frozen paradigm.

The following list is a humble, tongue in cheek, not-sure-I-really-endorse-some-of-these-things guide for standing out at future conferences. All of the following are direct observations that I made at the 2016 Sloan conference, and were neither embellished nor fabricated. Each suggestion is rated on a scale from 1 (Don’t Actually Do This) to 7 (Definitely Do This).

Wear something ridiculous (5). What better way to get attention while marching through the corridors of a business school inspired jcrew ad? I counted 1 “Make America Great Again” trump hat, and 1 Canadian tuxedo. Both were on the same person. I’d suggest something less aggressive.

Get surly at the toaster (6). Some of the best networking I witnessed was while standing in line for the coffee, water, and food. If you sit near the toaster, the director of analytics for your favorite team can’t exactly just walk away from you, lest they relinquish their breakfast. Shmear their bagels with satire. Sip coffee. Eat bananas.

Impersonate someone (2). Hide your badge and introduce yourself as someone you’re not. Conduct an ad-hoc interview. Create a scouting report of the people you interview for the team and/or individual you’ve impersonated and email that team and/or individual your detailed report. They’ll be confused but delighted that you’ve done the work for them. Attach your resume.

Play Sloan bingo (7). (link)

Brand yourself with a unique concept and bitch about it endlessly (6). Arm your bitchiness with a bitchy product, it doesn’t have to necessarily be an instrument, but it has to demonstrate your bitchiness in a way that anyone can understand it. Make it something unique, do your research.

Network for other people and they will network for you (7). Meet people. Genuinely meet people and show interest in their work. Digest it. Spitball with them. Then meet more people with similar ideas and connect the dots. They’ll reciprocate for you.

Build lasting relationships (7). All in all, Sloan attendees are some wonderful, intelligent people – with equally impressive ideas and insights. Here are a few I came across:
        Mark Conway (MVP) - Film your tennis match with your phone, send Cizr Tennis your video, and they’ll send it back fully annotated and analyzed. Certainly a better option than buying expensive equipment that does the same damn thing. Mark is not only paving the way for analytics in tennis, but he’s showing the sports analytics community how to innovate.

        Dave Thomas (Most Unique) - Sure, you could outsource for technology that records data on kinesthetic output of your athletes – or, you could have Dave on your #squad. Dave works with five or six Olympic teams (Summer and Winter), trying to find coachable performance levers in both granular sensor data, and higher level results data for the English Institute of Sport.

        Sam Sorkin (rookie of the conference) - Sam is a high school senior at Junipero Serra in San Mateo, CA, and a staff writer for SB Nation's Golden State of Mind. He’s already founded his own business, Sorkin Sports Intelligence, where he serves as a private game analyst for companies taking its clients to sporting events. He’s currently planning to attend the University of Michigan and major in business and sports management. Keep this kid on your radar.

       Vikram Maran (most likely to succeed) - Vik is a software engineer with an information science background. He enjoys studying the beautiful game from a strategy and performance perspective, and has Ben Pugsley to blame for pulling him into the space. Vik is a chimera of critical scientific thought and technical expertise. Hire this man.

Sloan is the time and place in which it exists. It’s Boston, a hotbed of scholarly insight, science, and business acumen. But it’s also in late winter, and it’s blanketed in chilly weather (not as cold as things have been, but still cold to this LA nerd). And it feels like it’s going to warm up soon, as evidence by a few tiny green buds that are peeking through the cool air. Stats are great. Previous performance is great too, but there’s a whole world of other research ideas and disciplines to which the analytics movement can shift its attention – (ahem, psychology). There’s still time to thaw out and diversify.