On Luck, Talent, Parity and Predicting the Upper Bounds with Machine Learning

Josh W
October 01 2013 09:32AM


This last week I spent at the European Conference on Machine Learning where there was a workshop on Sports Analytics.  I happen to be presenting on some of my previous #fancystats machine learning in predicting a single game. The more I look into this the more I realize just how much random chance (or "luck") plays a role in determining the outcome of a game in the NHL. But when I talk to people about this they just cannot wrap their head around the idea. 

Read Article | 4 Comments


Brian Sutherby
September 30 2013 05:57PM

I have been thrust in charge of most things Streakcred because I’m the person around here that likes to win every useless thing in the world and I’m quite proud of it. With a mantle full of fishing titles and fantasy sports trophies I set out to dominate this year’s Streakcred. I’ve trained long and hard for this.

Read Article | 32 Comments

Reviewing the 2013-2014 Hockey Prospectus Guide

Cam Charron
September 30 2013 08:53AM

Cover of the .pdf version

The 2013-2014 Hockey Prospectus Annual is a book much different in style and content than the 2013 Hockey Abstract written by Robert Vollman. The Abstract was Vollman's general look at where statistical hockey analysis is in the year 2013, while Prospectus focuses on teams and players, and less on theory.

The writing is primarily done by Vollman, but there are many familiar names that pop up if you read often about hockey metrics online, such as Adam Gretz, Ryan Wagman and Corey Sznajder. With Vollman writing up many of the essays introducing teams and players, it's no surprise that the tome is heavily stocked with player usage and performance charts.

Read Article | 1 Comments


September 25 2013 03:19PM

As the preseason comes to a close and the NHL regular season is squarely on the horizon StreakCred is ready to go with an early bird - $15 for the entire year.

Read Article | 22 Comments

Six-Year FenwickClose%

Travis Yost
September 20 2013 11:58AM




Behind the Net has always been an invaluable resource for checking in on team data, particularly for metrics like Fenwick Close -- a shot-attempt differential that mitigates the damage often inflicted by score effects in hockey games. Unfortunately, the data is usually tabled by season, and it's sometimes difficult to jump back and forth between seasons, particularly when looking at improvement or decline in performance. 

Below, I've compiled the FenwickClose% data for each NHL team, sorted by conference, and dropped into a graph that's not particularly friendly to the color-blind, like myself. I've also grabbed a hold of the biggest one-year risers/fallers, something I think speaks a bit about coaching in today's game. I think it's a nice little tool to gaze at, especially if you want to get a quick grasp of the true best/worst teams of the last six years. 

Read Article | 8 Comments