Maple Leafs Vollman Chart 11-12

Wonderful, high-explanatory chart. If you know what it means. See below for explanation.

Ever wondered why a player you thought was great was considered horseshit by others?

Ever believed a player to a complete lazy, pancake-eating bum but then found out he was one of the most effective players on the team?

Me too.  And then, thanks to the simply fantastic Lowetide, who blogs both at his own blog and at OilersNation – and not to mention has his own radio show – posted one day about Robert Vollman’s Player Usage Charts.  Great, I thought, more stats stuff.


I’ve always had a love/hate thing with stats, in hockey and elsewhere.  I sucked, simply sucked, at maths in school, although I managed to pull through my GCSEs with the help of one genius teacher who used to wear a big matrix-style leather coat and play around with his MP3 player at his desk – it’s often the most unlikely!  I just never got on well with numbers.  Words and music and art were more my thing, not to mention sport.  Not an academic, I suppose you could say.  I never saw what the point in it all was, but then I got to uni.  Now my course was based around the live event industry – lighting design for concerts and theatres and corporate stuff, sound design, rigging, all pretty cool behind-the-scenes stuff that makes things like hockey games and the Olympics look so awesome a spectacle – but there was actually a significant amount of maths involved.  The physics of sound, light, load-bearing weights, all that sort of stuff was vital to the industry not just for health and safety reasons but to make it look awesome aswell.  And so begrudgingly I worked away at the numbers crap, but it eventually dawned on me that, yes, this stuff is pretty damned important.  You basically can’t go through life without it, like it or not.


A lot, and I mean a lot of hockey fans don’t like the use of stats.  But without them, you couldn’t add goals to assists and come up with points.  You couldn’t have plus/minus (flawed as it is).  How about goalies with their goals against average and save percentages?  Nope.

Now over the past decade or so, various writers and bloggers have been working on what are called Advanced Stats.  As I’ve mentioned before, if you’ve seen or read Moneyball, you’ll know basically what they are.  They’re basically a way of adding in context to the goals being scored on the ice.  I said above that plus/minus is a flawed stat.  This is because it doesn’t provide any context as to what the players are actually doing on the ice when the goal is scored – were they the primary culprit that allowed the other team to score, or were they the guy doing everything right by covering his man but someone else screwed up and he has a minus against his name as a result?  Was he the guy that scored the goal or the guy that had just the luck to step onto the ice on a line-change the moment his team scored and was awarded a plus?  No context.


Robert Vollman, the former ESPN columnist and founder of Hockey Abstract, has utilised the available Advanced Stats – summarised in a glossary-of-sorts below – to come up with Player Usage Charts.  These charts are wonderful for people like myself  who struggle with just reading data off a table, and would like something more visual to show us the “aaaaah, ok, I see what your saying” point of the whole thing.  Usually, I tend to skip tables and graphs and go straight on to the summary at the end to gather the main points, but with these charts you can see for yourself how every single player is used over the season in relation to the rest of their team.

Vollman Player Usage Chart Explanation

The maths equivalent of a 70s disco floor.

To summarise:

  • The horizontal axis shows Offensive Zone Start % – the percentage of times a player started their shift in the offensive zone.  Those with a percentage over 50% will include rookies being sheltered and one-dimensional scorers who lack defensive skills.  Those with a % below 50, will be defensively responsible players being used to shutdown opposition scorers.
  • The vertical axis shows Relative Corsi Quality of Competition – essentially the higher you are up on this axis, the more difficult competition you are facing as a player.  Players high on this axis will again be the defensively responsible ones, OR the offensively gifted players who are so good they have to face tougher opposition.  Players lower on this axis will be the ones who aren’t capable of handling tougher assignments, scorers preying on weaker/softer competition or 4th line/fringe NHLers who get little ice time.
  • If a player has a blue bubble, it means that their Relative Corsi (essentially plus-minus but based on shots, not goals, adjusted for how a player’s team performs when he is not on the ice) is positive and that they are usually moving play in the right direction and in all probability creating more scoring chances for and less chances against as a result.  The bigger the bubble, the more positive the effect he is having for his team when on the ice.
  • If a player has a white bubble, it means that their Relative Corsi is negative, and that the play is generally moving back towards their own net when they are on the ice and allowing more chances against.  The bigger the bubble, the more negative the effect he is having on his team when on the ice.
  • CONTEXT!!!!!!!!!!  Players such as the unfairly-much-maligned Shawn Horcoff, Manny Malhotra and Jay Bouwmeester all have fairly large white bubbles, so you might think they are really bad NHLers.  However, as the chart shows, they are also facing really tough competition (i.e. the opposition’s best players) whilst starting in the defensive zone more often than not, meaning they don’t really have a chance!  They are doing their jobs so others don’t have to.  Likewise, players who have large blue bubbles but are facing soft competition with easy zone starts are also doing their job, but are being put in favourable circumstances.  It is the players with blue bubbles in the upper left quadrant of this graph that I would describe as the elite NHLers, the ones who can score whilst facing difficult competition and tough zonestarts.
  • Mikael Backlund of the Flames is an absolute gem, for starters.  Ales Hemsky, Taylor Hall, Getzlaf and Perry, Patrice Bergeron, Daniel Winnik, most of the Detroit Red Wings (!), Marcel Goc, PK Subban, Shea Weber and Ryan Suter, Frans Nielsen, Jordan Staal, Victor Hedman, Dan Hamhuis, etc..  THOSE are the guys you build a team around.

Please, if you enjoy NHL hockey and want to know how effective your favourite players are, or want some ammunition to reinforce how bad your buddy’s favourite player is,  I urge you to read the latest Player Usage Charts (linked to above).  It is stunning stuff, and whilst it is just putting something we already had available in a different format, THAT is what makes it so groundbreaking – ease of use, openness, perhaps a chance for advanced stats to be useful to the masses (i.e. the dumb-asses like me!).


David Staples of the Edmonton Journal’s Cult of Hockey blog started tracking what he referred to as True Plus/Minus, by going over each and every goal for and against the Oilers and determining who was either at fault on a goal against, or who contributed to a scoring chance (even if they didn’t touch the puck, e.g. screening the goalie).  Tough work, but it resulted in a much clearer picture of who was to blame for bad defensive plays or not.

Corsi is one of the most used stats used by bloggers.  This is a very simple stat which basically adds together ALL shots taken by a player at even strength – shots on target, shots that miss the net, and shots that are blocked.  It is more stable than Plus/Minus due to it tracking more events on the ice than just goals, and is a useful indicator of puck possession.  If a player is generating a high Corsi number, he is most likely spending more time in the offensive zone generating shots.  However, it still lacks some context, and many people prefer to use Relative Corsi (Corsi-Rel), which is a player’s Corsi number minus  the entire team’s Corsi number when that player is off the ice.  If you have a player with a middling Corsi number, you can adjust the number to Relative Corsi to see his actual value to the team by showing how the team performs when is not on the ice.  However, Corsi and Corsi-Rel do not take into account quality of competition (see below) or how productive the player ultimately is, but they are more accurate indicators of which players steer the play in the right direction (i.e. towards the opposition net and away from yours).  For example, Daniel Sedin was top of the league with a Corsi-Rel of 22.5, whilst Manny Malhotra was worst with -32.6 (don’t worry Canucks fans, I’ll get to how Malhotra isn’t as bad as he looks – all about context!).

Quality of competition (Corse-Rel QoC) – this is vitally important, and one of my favourite stats.  This is calculated by averaging a player’s opponents’ Relative Corsi rates across their head-to-head ice time.  This way you can see which players played the harder competition – for example, Nick Lidstrom has the highest QoC number over the last 5 years at 1.43, and Kris Russell had the worst at -1.07.  This is the view also supported by watching the game – we all know that players like Lidstrom (Bolland, Bouwmeester, Zetterberg etc.) play tough competition and now we can quantify it.

Finally, Zone-Starts and Zone-Finishes are also brilliant ways of seeing how players were used over a season.  This tracks where every players shift starts and finishes their shifts.  To calculate Zone Starts, simply take the number of shifts started in the offensive zone, and divide by the total number of shifts started in both the offensive and defensive zones.  Sub-in number of shifts ended in the offensive zone to calculate Zone Finishes.  These numbers help us see whether a player was given huge defensive responsibilities, or a massive help by starting constantly in the offensive zone (where they are more likely to score).  The former number helps identify players leaned on heavily by coaches to prevent goals (such as Shawn Horcoff, who started just 43.9% of his shifts in the Offensive Zone), and the latter number helps identify which players perhaps aren’t considered defensively responsible enough by their coach – for example rookie forwards (such as Jordan Eberle who started 60.7% in the offensive zone).  The league extremes were the Sedin Twins, at around 79% O-zone starts (perhaps the largest differential in NHL history, you can see why they get so many points!), and Manny Malhotra at around just 13% – a miniscule number, perhaps the tiniest in NHL history and showing why his Corsi-Rel is so bad – he was having to face a LOT of shots against because he rarely saw O-zone time!



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s