Howe and Why: 2010 Golden Elbow Awards, Part Two
What an amazing year 2010 was for statistical hockey analysis! We had a bookalmost twoand more sites devoted to the hobby than ever before. The inaugural Golden Elbow Awards are our way of celebrating the remarkable achievements of our diverse community.
In Part One, we took a look back at the best analysis of a skater, goalie, team, and league, the best analysis of goaltending in general, and a fun look at the best historical analysis. This week, in Part Two, we saved the very best for last.
Best Analysis Analysis
Best Gabe Desjardins
Best New Innovation
Best New Insight
Once again, I'd like to thank everyone who submitted their favorite articles, and most especially our panel of advisors and judges who helped us identify the work that was the most complete, thorough, innovative and insightful. The panel included: Alan Ryder, Gabe Desjardins, Kent Wilson, Tom Awad, Phil Myrland, Ben Wendorf and Corey Pronman. We apologize in advance for any omissions, and please enjoy these awards in the spirit with which they were meantencouragement, promotion and, most of all, fun!
It's impossible to study anything objectively without occasionally challenging the established conventional wisdom. The Golden Elbow for Best Controversy is awarded to the article that hits the mainstream and has the most meaningful positive impact to the game.
2010 Winner: Who Did Colin Campbell Call a Little Fake Artist? by Tyler Dellow
The saga dubbed Colin Campbell's Nightmare began with Tyler's detailed analysis of NHL front office emails, and found its way into the mainstream like nothing before. Though it lacked the detailed level of analysis found in his other work, it drew attention to our field, and it was without question the most common submission for article of the year.
Corsi analysis found its way into the mainstream when Don Cherry fought back on Coach's Corner, slamming the approach in defense of Ryan Johnson. Accepting the award on behalf of the larger community is David Staples with Keep Your Head Up, Corsi Gang! Not only is it a great summary of the Corsi/Cherry controversy, but it was written by a Corsi skeptic working in the very center of the Corsi worldEdmonton.
Note: Even as the year winds down, the Corsi controversy remains as heated as ever, as Gabe Desjardins recently took it from what seemed like every hockey fan in Colorado in Retro NHL and Anger at Corsi.
Best Analysis Analysis
Many of the best articles we write aren't about any particular part of the game, but about our approaches to the analysis themselves. The Golden Elbow for the Best Analysis Analysis is awarded to the most insightful perspective on why we look at the sport the way we do.
2010 Winner: Likelihood and the Way Humans Think by Vic Ferrari
In this fantastic article, Vic divided us into two rough pools of thinkers, the frequentists and the Bayesians, how they would each look at a hypothetical rookie who scores 16 goals in 103 shots, what conclusions they would each draw, and ultimately, the colorful debate they would have. You've probably been in debates like these countless times, and never had the experience described so accurately and explained so well before.
In The Limits of Observation, easily one of my favorite articles of the year, Kent Wilson took an extensive look at the nature and limits of subjective, observation-type analysis.
An important property of an analysis is that it has good predictive value, so the Golden Elbow for Best Prediction is awarded to the article that contains the analysis with the best predictive value.
2010 Winner: Washington Capitals vs. Montreal Canadiens Preview by Olivier Bouchard
Not only did Olivier predict the amazing upset, but clearly explained himself with a careful examination of line combinations, Fenwick and Zone starts. In fairness, there were others who saw the upset as a possibility (like Timo Seppawho even predicted the 2nd round upsetand Tom Awad), but no one else took the actual plunge. I know I sure didn't.
To paraphrase one of the judges, Jonathan Willis made a clear, relevant and timely presentation on variation, luck and randomness while Projecting Ryan Miller. In this piece, he looked at results by month, and by situation, and predicted that Ryan Miller, who posted a .929 save percentage in 2009-10, would find himself at .915 in 2010-11. Guess what his save percentage is today?
Note: Keep your eyes on the Vancouver Canucks, because yours truly predicted that Henrik Sedin Wont Score 100 Points. I looked at even strength scoring rate and made a comparison with his two closest historical comparables Adam Oates and Doug Gilmour. Sedin's on pace for 98
Best Gabe Desjardins
Gabriel Desjardins spits out quality insightful analyses faster than watermelon seeds at a picnic. The Golden Elbow for Best Gabe Desjardins is awarded to whichever article our community was least likely to have managed to produce without him, or at the very least would have been unable to explain as clearly and eloquently. This category also stops him from flooding all the other awards.
2010 Winner: Tale of Two Talents (and part two) by Gabriel Desjardins
Noting how shooting percentage regresses so strongly to the mean, Gabe managed to break down the components of shooting percentage to determine exactly what constitutes a goal, in what ratio, and how much of it is talent, and how much is transient (luck). While some may consider it impossible to actually select Gabe's best work, this analysis was certainly as good as any other.
In Shooting Percentage vs. Age, Gabe conducted a careful analysis of shooting percentage and age, showing that it does not peak in the mid 20's as commonly believed, but is rather roughly constant until age 26 before a long, slow decline. The peak in goals in the mid 20's is actually based on an increase in shots. The Observed Distribution of Shooting Talent was another popular finalist, as was his work on Corsi and Score Effects (here's part two), and I was partial to his look at Jacques Lemaire's impact on the Minnesota Wild, using Corsi and PDO, something you Devils fans might want to consult.
Best New Innovation
Building up to our two most important categories, the Golden Elbow for the Best New Innovation is awarded to the new statistic and/or analytical approach, or major refinement on an existing statistic or approach.
2010 Winner: Regression-based Adjusted Plus-Minus by Brian MacDonald
Fixing the plus-minus statistic is almost a rite of passage for all hockey analysts, which makes it that much more impressive that someone would come along with an analysis that covers new ground. Brian's comprehensive work is, to paraphrase one of the judges, the most effective way to make plus-minus more meaningful, without the bias of teammates and quality of competition.
In a year that saw Gabe Desjardins refine QualComp by adding Corsi, the introduction of Tom Awad's DeltaSOT: Plus-Minus and Corsi Have a Baby, and of course Iain Fyffe's Hall of Fame Inductinator (all-stars, defensemen, forwards and goalies), it is an amazing accomplishment merely to be among the finalists. So I shall mention them all, in no particular order:
This year kudos go to all of those who collected and analyzed scoring change dataan approach pioneered by the great Roger Neilson 30 years ago. In his postseason piece Blackhawks 7 Flyers 4Game 5 Scoring Chances, Scott Reynolds can accept the award on behalf of all those using Vic Ferrari's tools, including Derek Zona, Kent Wilson, Dennis King, Slava Durvis, Olivier Bouchard, and others.
The man for whom this category should probably be named, Tom Awad's multi-part series What Makes Good Players Good was the first analysis of its kind, defeating issues related to sample size by grouping players based on ice time, looking at what traits they share and discovering several interesting links, including those related to finishing ability, special teams and preventing opposition scoring.
In keeping with 2010's theme of luck-based analysis, Adam Kubaryk's bold new approach of the topic in a two part series Odds-Based Approach to Luck in 2009-10 brought us a lot of insight by running simulations based on the closing betting odds on games. The accumulated data was excellent, and will no doubt be used by others in future studies.
Best New Insight
The whole point of our work is to dispel myths and improve our understanding of the game. As such, the most important Golden Elbow is awarded to the piece that best helps correct a misperception.
2010 Winner: Blocked Shots: Luck or Skill? by Sunny Mehta
When someone asks you how excluding blocked shots (Fenwick) could possibly be a better proxy of territorial advantage and puck possession than Corsi, send them Sunny's way. Leading the switch from Corsi to Fenwick was certainly one of the most important contributions of 2010.
The two leading finalists was JLikens work on the NHL's Shot Recording Bias, and Vic Ferrari hammering home the Importance of Quality of Competition.
Note: As above, I'd like to mention the two other finalists, Does Shot Quality Exist by Tom Awad, who looked at the differential between expected goals (using the Krzywicki model) and actual goals, and while the arbitrary randomness of assigning an extra point based on the shootout is already well known by Gabe, Alan and most people in our community, the case has never been made so clearly and comprehensively than by the latest entrant to the award, Michael Schuckers in NHL Shootout as a Crapshoot.
Thanks everyone for a fantastic year! As someone who has been involved in statistical hockey analysis for years, I feel so fortunate to have such a wealth of insightful research available. It was certainly a lot of fun to look back at all the great work, and my deepest apologies for anyone that was left out. With so many fantastic articles written by many great writers, your favorites could easily have been unintentionally missed, so please leave comments to refresh everyone's memory.
Happy New Year!
Robert Vollman is an author of Hockey Prospectus.
You can contact Robert by clicking here or click here to see Robert's other articles.