Book Review – The Long Tail

The Long Tail: How Endless Choice is Creating Unlimited Demand
by Chris Anderson, 2006

A bit outdated on its examples, but it’s still fun to read about Myspace and early Amazon. Documents the early ability of the internet to offer a scalable solution to long-tail demands. Many more products can be in inventory for an online retailer than a brick-and-mortar store, as sales are aggregated across space. Digital goods take this to the extreme by having no significant inventory costs, allowing for the “stocking” of every possible item. Some fun graphs from the book:

Paper Published – Incentiveā€Compatible Prehospital Triage in Emergency Medical Services

Joint work with Alex Mills. Forthcoming in Production and Operations Management: https://onlinelibrary.wiley.com/doi/abs/10.1111/poms.13036

Abstract: The Emergency Medical Services (EMS) system is designed to handle life-threatening emergencies, but a large and growing number of non-emergency patients are accessing hospital-based healthcare through EMS. A national survey estimated that 17% of ambulance trips to hospital Emergency Departments (EDs) were medically unnecessary, and that these unnecessary trips make up an increasing proportion of all EMS trips. These non-emergency patients are a controllable arrival stream that can be re-directed to an appropriate care provider, reducing congestion in EDs, reducing costs to patients and healthcare payers, and improving patient health, but prehospital triage to identify these patients is almost never implemented by EMS providers in the United States. Using a decision model, we show that prehospital triage is unlikely to occur under the current structure of fee-for-service reimbursements, regardless of how effective the triage process might be, unless low-acuity patients are unprofitable and a hospital is willing to coordinate with EMS. We demonstrate several mechanisms a payer such as Medicare could use to promote prehospital triage: reforming fee-for-service reimbursements or offering a value-based payment, such as bundled payments or shared savings contracts. Using data from a national survey and levels of triage effectiveness demonstrated in the literature, we conservatively estimate that Medicare alone could save between $3 and $70 million per year (depending on triage effectiveness) by providing incentives for prehospital triage. Between 26,500 and 628,000 non-emergency patients could be diverted to more appropriate care options, making prehospital triage a practical step to address hospital emergency department crowding.

Book Review – Robot-Proof

Robot-Proof: Higher Education in the Age of Artificial Intelligence
by Joseph Aoun, 2017

Written by the president of Northeastern University, this short book makes numerous suggestions for crafting successful universities in the near-future. As more jobs become automated and skill upgrades become more frequent for workers, universities need to adapt. The whole book is worthwhile and suggested. Here are two suggestions I particularly like:

-Include experiential learning in classes, and focus on technological literacy, data literacy, and human literacy. Core cognitive capacities should include critical thinking, systems thinking, cultural agility, and entrepreneurship.

-In focusing on lifelong learning, allow students to build custom degrees/certificates, taking only the classes most relevant to their personal situation and goals. One possibility to promote lifelong learning would be to create a subscription model in which students can take X classes per semester/year as they move forward in their career.

NFL Picks – Super Bowl 2018

Overall Against the Spread: 123-117
Week 2: 7-9
Week 3: 10-6
Week 4: 6-8
Week 5: 7-7
Week 6: 9-6
Week 7: 3-10
Week 8: 5-8
Week 9: 3-10
Week 10: 7-6
Week 11: 11-2
Week 12: 8-7
Week 13: 9-5
Week 14: 12-4
Week 15: 10-5
Week 16: 9-6
Week 17: 6-9
Wild Card: 0-4
Divisional: 1-3
Conference: 0-2

NFL Picks – Divisional Round of 2018

Last week’s picks should have said wild card, not divisional round.

Overall Against the Spread: 122-112
Week 2: 7-9
Week 3: 10-6
Week 4: 6-8
Week 5: 7-7
Week 6: 9-6
Week 7: 3-10
Week 8: 5-8
Week 9: 3-10
Week 10: 7-6
Week 11: 11-2
Week 12: 8-7
Week 13: 9-5
Week 14: 12-4
Week 15: 10-5
Week 16: 9-6
Week 17: 6-9
Wild Card: 0-4


NFL Picks – Week 17 of 2018

Three games without lines at this point.


Overall Against the Spread: 116-99
Week 2: 7-9
Week 3: 10-6
Week 4: 6-8
Week 5: 7-7
Week 6: 9-6
Week 7: 3-10
Week 8: 5-8
Week 9: 3-10
Week 10: 7-6
Week 11: 11-2
Week 12: 8-7
Week 13: 9-5
Week 14: 12-4
Week 15: 10-5
Week 16: 9-6


Kaggle Punt Analytics Competition

Kaggle has an interesting new competition, available here.

We’re excited to let you know about a non traditional competition launch. Unlike our typical supervised Kaggle machine learning competition. Instead, this challenge asks you to use data to propose specific rule modifications for the NFL that will reduce the occurrence of concussions during punt plays. Click the button below to learn more about this format.
About the host:
The National Football League is America’s most popular sports league, comprised of 32 franchises that compete each year to win the Super Bowl, the world’s biggest annual sporting event. They’re committed to advancing progress in the diagnosis, prevention and treatment of sports-related injuries.  Since 2002 alone, the NFL has made 50 rules changes intended to eliminate potentially dangerous tactics and reduce the risk of injuries.
Competition Description:
Your challenge is to propose specific rule modifications (e.g. changes to the initial formation, tacking techniques, blocking rules etc.), supported by data, that may reduce the occurrence of concussions during punt plays. Submissions will be evaluated based on the efficacy of the solution while maintaining the integrity of the game. Winning teams will be invited to present their submissions in a pitch-style event the day before Super Bowl LIII in Atlanta, GA and the top two presenting teams will each be awarded two Super Bowl LIII tickets.
Prizes:
$80,000 Total prize pool (and more!) awarded to top teams
Next important deadline:
January 9: Final submission deadline