Applied Sports Science newsletter – February 2, 2017

In a nondescript strip mall in Alabama, a little-known Division II prospect once trained for his shot at the NFL. He’s a star now, but Butler still comes back to this gym

Entering this season, with Kevin Durant gone, the Oklahoma City Thunder were keenly aware of of their situation and what that would mean for the workload of Russell Westbrook. While fans and the media fantasized about him doing something crazy like, say, averaging a triple-double, the Thunder worked to hatch a plan to make sure their franchise player — singular now — wasn’t run into the ground by April.

Part of what complicated it was the fact that Westbrook packs as much energy into every minute he plays as anyone in the league. His 34 minutes are unlike most players’ 34 minutes. Westbrook’s usage rate is soaring at about 40 percent — on pace to set a league record — and with his relentless rim attacks, the responsibility to engineer an offense and effectively carry a team into the Western Conference playoffs, the Thunder knew they’d have to manage it.

Researchers at Southern Methodist University in Dallas have developed a concise new explanation for the basic mechanics involved in human running.

The approach offers direct insight into the determinants of running performance and injuries, and could enable the use of individualized gait patterns to optimize the design of shoes, orthoses and prostheses according to biomechanics experts Kenneth Clark , Laurence Ryan and Peter Weyand, who authored the new study.

After the Senior Bowl, there’s a gap of about a month where we lose our eyes on even the top draft prospects. It’s safe to say that none of them are resting, but little is known about the secretive and selective NFL Combine training programs that have popped up around the country in the last decade.

There are only a handful of these programs, which originally popped up due to a demand of many athletes to do more than their own school could do for them. They seek specialization by some of the trainers, most notably IMG (Bradenton, FL), Exos (principally in Arizona, but facilities in Los Angeles and Florida), Velocity Sports (New York) and St. Vincents Sports Performance (Indianapolis.) There are a number of other programs, but overall, a majority of athletes that go to the NFL Combine prep in some manner.

And it’s no wonder. This kind of prep works. Athletes who have attended these camps have increased their measurables, in terms of strength, speed, quickness and even the Wonderlic tests. Fractions of a second on the 40 and a couple reps on the bench can mean the difference between picks or even rounds, potentially meaning hundreds of thousands of dollars.

Nolan Ryan joined the Houston Astros prior to the 1980 season as the highest-paid player in MLB, earning $1.1 million a year. I received a phone call from the Astros GM, Tal Smith, inviting me to a reception to introduce him as the Astros’ newest member. Prior to the reception, I had never met Nolan. I had watched him pitch in the 1979 MLB All-Star game on TV and I knew that he lived in Alvin, Texas about 20 minutes away from my home in Clear Lake.

Before we signed Nolan, our starting pitchers were doing one total body workout between each start. I had four of the five starters on board with the program, but the fifth, Ken Forsch, was not as committed at the other four. I was talking to Ken at the reception when Nolan walked up and introduced himself. He asked if I was the strength coach; I said yes. He said, “Well – I’m going to have a problem with your strength program.”

My first reaction was we have a million-dollar prima donna who will undo two years of the things we have worked for. Ken grinned ear-to-ear thinking he had somebody on his side. But Nolan said, “I can only lift three times between starts.” That set the tone for the Astros for the next three decades and Nolan and I became instant friends.

Anyone who has trained in a group knows that people have very different responses to training. Even if you’re doing identical workouts, some people will make far more progress than others. And an unlucky few, according to some researchers, won’t respond at all—they appear not to get fitter.

Back in 2015, I wrote about a study that challenged the idea the some people are genetically wired to be “non-responders” to exercise. By imposing longer and more intense workout regimens, the study found, the number of non-responders could be reduced to zero.

Still, some critics weren’t convinced. Perhaps these longer, harder exercise programs were simply so discouraging that all the non-responders gave up partway through the study. The only way to know for sure would be with a “repeated measures” study: put a bunch of volunteers through a training program to identify the non-responders, and then make the non-responders do another, harder training program to see if they would improve.

… the NBA’s new rules are, generally, good news for the players—they establish, importantly, a presumption that players own all data about themselves, and ban the use of wearable data in contract negotiations. However, they still leave big questions, especially about the future sale of wearables data to third parties, open and unanswered. (VICE Sports reached out to the NBA and the NBPA for comment but did not hear back.)

Let’s break it down. The first six clauses of the wearables language are mostly boring, and relate to the establishment of a “wearables committee,” which will be responsible for approving the use of new devices and storing data that comes from them. That’s all well and good, and a welcome development, but the meat is in the last three clauses.

Clause the seventh establishes that before any team asks any player to wear any wearable, even approved ones, they have to share with the player: 1) what exactly the device measures, 2) what the measurement means, and 3) why wearing the device will be beneficial to the player. Moreover, the player can stop using the wearable at any time. This is a big deal.

… Started by Professor Lionel Tarassenko, head of Oxford University’s Institute of Biomedical Engineering, the company has developed software that can turn bog-standard video cameras into health monitors, capable of detecting a person’s heart- and breathing rates from metres away. The idea is that rather than having to regularly hook patients up to machines to check their vital signs, hospitals can monitor them less intrusively and around the clock.

The technology is based on the tongue-twisting concept of photoplethysmography – measuring tiny changes in the colour of your body, which briefly turns slightly redder each time your heart beats. Its function is similar to that of the ‘pulse oximeter’ clips you might have been given to put on the end of your finger when visiting the hospital.

In this paper we present an on-manifold sequence-to-sequence learning approach to motion estimation using visual and inertial sensors. It is to the best of our knowledge the first end-to-end trainable method for visual-inertial odometry which performs fusion of the data at an intermediate feature-representation level. Our method has numerous advantages over traditional approaches. Specifically, it eliminates the need for tedious manual synchronization of the camera and IMU as well as eliminating the need for manual calibration between the IMU and camera. A further advantage is that our model naturally and elegantly incorporates domain specific information which significantly mitigates drift. We show that our approach is competitive with state-of-the-art traditional methods when accurate calibration data is available and can be trained to outperform them in the presence of calibration and synchronization errors.

Background:

Despite the significance of anterior cruciate ligament (ACL) injuries, these conditions have been under-researched from a population-level perspective. It is important to determine the economic effect of these injuries in order to document the public health burden in the United States.
Purpose:

To describe the cost of outpatient arthroscopic ACL reconstruction and health care utilization among commercially insured beneficiaries in the United States.
Study Design:

Economic and decision analysis; Level of evidence, 3.
Methods:

The study used the Truven Health Analytics MarketScan Commercial Claims and Encounters database, an administrative claims database that contains a large sample (approximately 148 million) of privately insured individuals aged <65 years and enrolled in employer-sponsored plans. All claims with Current Procedural Terminology (CPT) code 29888 (arthroscopically aided ACL reconstruction or augmentation) from 2005 to 2013 were included. “Immediate procedure” cost was computed assuming a 3-day window of care centered on date of surgery. “Total health care utilization” cost was computed using a 9-month window of care (3 months preoperative and 6 months postoperative). Results:

There were 229,446 outpatient arthroscopic ACL reconstructions performed over the 9-year study period. Median immediate procedure cost was $9399.49. Median total health care utilization cost was $13,403.38. Patients who underwent concomitant collateral ligament (medial [MCL], lateral [LCL]) repair or reconstruction had the highest costs for both immediate procedure ($12,473.24) and health care utilization ($17,006.34). For patients who had more than 1 reconstruction captured in the database, total health care utilization costs were higher for the second procedure than the first procedure ($16,238.43 vs $15,000.36), despite the fact that immediate procedure costs were lower for second procedures ($8685.73 vs $9445.26).
Conclusion:

These results provide a foundation for understanding the public health burden of ACL injuries in the United States. Our findings suggest that further research on the prevention and treatment of ACL injuries is necessary to reduce this burden.

Dietary guidelines come under a lot of fire. They have been accused of not being based on evidence, not being environmentally sustainable and being out of touch with nutritional science. They also fail to change people’s eating habits, as shown in Australia and the US.

The time has come for us to rethink the purpose of dietary guidelines, what they contain and how they deliver their message.

As part of this, we need to think about how the public views dietary guidelines (and other health guidelines) so they become relevant. Hopefully, then people will be more likely to follow them.

Injured Premier League players have cost their clubs over £79million in wages during the first half of this season, according to a specialist insurance broker and risk consultant.

JLT Specialty (JLT) collated injury data from every Premier League match during the period of July 1, 2016 and December 31, 2016, with injuries causing players to miss at least one game being counted in the report.

In this article we will discuss the machine learning method called “decision trees”, moving quickly over the usual “how decision trees work” and spending time on “why decision trees work.” We will write from a computational learning theory perspective, and hope this helps make both decision trees and computational learning theory more comprehensible. The goal of this article is to set up terminology so we can state in one or two sentences why decision trees tend to work well in practice.

Yesterday we argued that Federer’s unusually strong backhand anchored his Australian Open win over Rafael Nadal, and today we found a startling statistical basis for that claim. It comes courtesy of the always helpful Jeff Sackmann at TennisAbstract. Relative to other sports, tennis remains fairly data-poor, but Sackmann’s hacking away at the problem with his Match Charting Project, which rallies volunteers to log tennis matches shot-by-shot, producing a granular picture of a sport usually painted in simplistic narrative strokes.

Sackmann used those logs to home-brew a stat he calls backhand potency (BHP), meant to gauge the efficacy of that particular stroke over the course of a match by assigning values to each of the specific outcomes it produces.

… Pardon the political correctness of this term, but it’d be unfair to call some of the players who have had their bodies called into question “fat.” Then again, it’s all relative. Not every baseball player is statuesque; one need only to take a look at the careers of guys like Kruk and Cecil Fielder to know it’s possible, but scouts are always wary of the large human—especially position players.

“It’s always concerning,” an NL Central scout said. “The fact of the matter is, baseball is an everyday game. If you wanna say it’s not exactly a cardiovascular workout, that’s fine, but it’s still a grind. You have to have a certain level of conditioning to play this game at your best for 162 games, close to a thousand innings a year for most of these guys. It’s tough to do that when you’re in tip-top shape. When you’re fat? There’s almost no chance.”

The history of the overweight position player isn’t exactly illustrious, and it hasn’t gotten much better over time, either.