Follow Colin on Twitter!: @ColinDJD
I have found, of late, that there continues to be a war of such when it comes to discussing #FancyStats, or in laymen’s terms, advanced stats.
I get the feeling that this blog has already been closed by a handful of readers after seeing the above, and for those who decided to continue reading, I’d like to provide an “advanced stats for dummies… angle toward all of this.
What I keep seeing or reading when the notion of advanced stats are brought up is a backlash from certain fans who will resort to comments such as “watch the game… or “this isn’t Moneyball…. Many, if not most of fans who seem unwilling to accept these “stats… into their everyday hockey discussions, to me, either do not understand or don’t care to understand the conversation around them.
To start things off, I will present you with a few of the main stats used and what they directly translate to, courtesy of BroadStreetHockey.com. Following this, I will explain how you have likely been using these stats all along, without necessarily knowing it or quantifying them.
TOI/60: Time on ice per 60 minutes. This calculates even-strength time on ice for the player per 60 minutes of team play. Two things to notice: 1) It isn't even-strength time on ice per GAME, as it adjusts for over-time periods; and 2) It is in decimal form, so 11.5 is eleven minutes and thirty seconds. - QoC: Quality of Competition. The average on/off-ice +/- of opposing players faced by a player. - QoT: Quality of Teammates. The average on/off-ice +/- of players - Corsi: a simple plus/minus-style rating of the total number of shots on goal, missed shots and blocked shots directed at the opposing net while a player is on the ice at even strength. - Fenwick: the same as Corsi, but excluding blocked shots. - Relative Corsi: measures the difference in Corsi between a player's on-ice performance and his team's performance when he's on the bench. - Corsi Rel QoC: The average relative Corsi of opposing players, weighted by head-to-head ice time. - Corsi Rel QoT: The average relative Corsi of teammates, weighted by head-to-head ice time.
The above may still sound like jargon or nonsense to many of you, but here’s the thing - all that advanced stats are doing is taking every day aspects of hockey, and providing them with quantitative labels in order to track them individually.
What I’m trying to say is that we've always discussed the importance of shots for/against when comparing teams and players. It is common knowledge that the more a certain player or team shoots at the opposing goalie, the more likely said player or team is to score and win.
That said, we take this discussion and break it down into a value which we can measure and label it as Corsi/Fenwick. What we are interested in seeing here are shot totals from individuals and teams, including or excluding blocked shots. This allows someone to provide a value to teams/players in saying that they may increase/decrease their odds of scoring and or winning said games.
What’s more is that most who analyze or discuss the sport (some who have played and others who have not), seem to think that shots and goals/points alone are the most integral part of team or player assessments. Well, here’s where advanced stats offer a litle more than just looking at shot totals or point totals; possession rate. What Corsi/Fenwick also provides is an ability to measure possession for teams and players, given that teams who shoot more often than others will generally “possess… the puck more.
If you can take anything from the above is that these two particular stats should be, and can be used to help “predict… games or individual efforts.
And this is the kicker, or the point where most get hung up on. Let me make this clear for folks, these advanced stats do nothing but provide supportive data to teams or even scouts which allow them to make better judgment calls. Examples could be who to ice in certain situations, be it 5v5 or 5v4, or in other words which player may improve your “odds… of succeeding against your opponent.
I have played hockey my entire life up to the point where you start getting paid, and I have watched hockey my entire life up to the point where I do get paid. Both of these factors, for an individual who writes about hockey, provides me with the ability to write about and assess games or players in a subjective and objective manner.
How many of you, for instance, would have offered advice to “pull your goalie with 3 minutes left in a game…? Do you think the success of this tactic which we’ve seen a lot of from Colorado’s coach Patrick Roy came from a hunch? Methinks not. By analyzing the numbers it would have been apparent that the success rate is greater when a goalie is pulled when there are X minutes remaining in the game, providing a greater chance of success.
Point is, there are little things that coaches/managers can point out or address to improve their odds at winning that may not be apparent simply by “watching the game….
By looking at a certain player I can say things like “Boy he moves the puck very well and seems to care for his on-ice assignments without much fault. This player is defensively responsible…. While such statements could likely be accurate from you are able to see on the ice, what you may not see are times in which said player is most effective. In other words, how does the player handle himself when the game is close, when the game is tied, and what opposition lines/players does he succeed while playing against?
Knowing this allows to effectively assess that a player seemingly finds greater success when playing in full strength situations vs. shorthanded, or against the oppositions second line vs. their first. I could take it a step further and with that player’s personal data, assess that his team appears to create greater offensive opportunities when he is on the ice, versus when he is not.
Sure I could make a mental note that he had a great couple of games of late and he’s been doing well offensively given his point totals, but perhaps this leads a coach to start said player in offensive situations more than defensive situations as other players appear to be stronger in that regard.
What I am really trying to say is, advanced stats are really no different than most of the conversations we’ve all been having for years now. The only real difference is that people have given certain parts of the conversation labels and have allocated a certain measurable value to it, which can be used as a predictive tool, if you will.
These stats, however, are not the bible and there is always chance or “luck… that comes into play, but using this tool in addition to your ability to watch games, creates a much clearer picture than would otherwise be presented to you.
Whether or not this discussion was helpful to you, I strongly consider that you look over all these “advanced… stats and break them down to what they really are. All these fancy terms and abbreviations really mean nothing more than certain parts of the game you are likely already focusing on, the only difference again is that you likely have no interested in measuring them and using them to try and better assess a team or player, because frankly, you’re likely not an NHL coach or manager anyways.
Thanks for reading!