Wednesday 3 September 2014

Blog New World Pt. 2 - I'm Too Fancy For My Stats


Welcome to part two of my of my two part 'new definitions' series. Last week we looked at some newer (at least to me) lingo that's been floating around the hockey world that you may be hearing/reading lately. Today, we continue on with probably the most talked about and somewhat controversial topic of advanced statistics or "fancy stats".

Average to casual hockey fans have probably never come across the majority, if any, of the terms described below. But for those of you who have been searching through the internet for more information or analysis on your favorite team and/or players recently, they are as unavoidable as the Oilers NOT making the playoffs. And, like most people, you probably just skipped over those articles, disregarding them as nonsensical jibber-jabber.


Dont' be distracted by the handsome devil above. He'll break your heart, trust me. Instead pay attention to the stats row printed in black on the right side of the card. These have been on hockey cards forever and memorized by hockey nerds (ask me how many goals Mogilny got in his rookie season, I dare you) to be regurgitated at the appropriate moment. Apparently they are all meaningless now. 


In the old days, hockey statistics were easier than Ke$ha. Number of wins or losses, number of games played, number of goals or assists, average of shots saved or goals allowed. We even gave goons the PIM stat (or Penalties In Minutes if your really a noob) so that they could have a category to thump their chests for. It was so straight forward that if you could master counting, you could participate in any hockey conversation. This simplistic approach to game data is one of the reasons we LOVE hockey. I mean, have you ever tried following baseball stats? Christ, some of those acronyms and numbers are more confusing than an Esa Tikkanen pre-game speech!

Look at the hockey card above. There are maybe four or five categories there we used to care about. Typically the first four because as fans, we love offense and we'll talk about it endlessly. But it seems it's the last one that became a central issue for some folks - good ole plus/minus. It's the Jared Staal of hockey stats. Originally, it was meant as an indicator of how good a player was defensively while still being able to help generate offense - whether or not they were on the ice for more goals for than against and by how many. More often than not though, it was just indicative of how good your team was as a whole and nobody really cared that much about it.

 Example: Pavel Bure once scored 59 goals with the Floroida Panthers. Us hockey fans thought he was awesome! Unfortunately, he was also on the ice for 61 goals against and thus, much to the ire of his teammates, was -2 for the year. But we didn't care!

  In fact, +/- was held in such little regard that the award for highest plus/minus was once called the Alka-Seltzer Plus Award and it was never presented on T.V. with the rest of the prestigious trophies. Nothing says pinnacle of achievement like an antacid! I bet it was probably a lot like getting the "Most Improved Player" award I received in Little League...

Another reason the plus minus stat was disregarded by many was because it really wasn't a good indicator of the player. It really didn't give you any relevant information on the player based on what their skill-set was or their contribution to their respective team. Many players play in different situations. Of course you would have a terrible +/- if the majority of your minutes were on the penalty kill or on a checking line against the leagues best players or with Luke Gazdic as a line-mate. On the flip side, so what if you had a good plus minus if you were playing mostly against the other teams weakest lines or predominately on the power play with your teams best players - a bag of equipment could have had a great +. Seriously, does anybody remember Paul Ysebaert?!

So it was, with these questions in mind, that a handful of folks started to delve further into the number of +/- in order to get a better understanding of a given players abilities or team play; to answer the questions begged by the vague stat. Folks who had graduated from counting and sprung ahead into addition and then division. Folks who are reverently referred to as "Hockey Nerds".

In very simple terms, the new "fancy" stats are numerical expansions of the +/- stat. They provide us with much more information on how that number came to be and what it really means for that individual...or so they say. There are those that disagree, but I'm not here debate the merits of the new stats, only to explain them.-ish. So without further delay....


NEW STATISTICAL TERMINOLOGY


CORSI: I'm starting you right off with the big boy. This is probably the most common advanced stat around. So much so that it even got mentioned a few times on CBC's Hockey Night in Canada towards the end of the year last season. A Canadian institution to be sure, but they're not exactly as heavy on the statistical analysis as they are on statements that start with, "When I used to play....". It's becoming as common as reality television and I predict it will make its way onto the back of trading cards in the not too distant future.

Not named after an Italian deli, but rather an old goalie coach who came up with the idea. essentially it measures the difference between the number of shots you (or your team) generates while your on the ice vs. the shots the other team generates. It also includes missed and blocked shots, not just shots on net and can be expressed as a % of a total or as  a +/- differential (+ being for and - being against).

Example: The Oilers fly into the offensive zone with the puck on the rush and generate one shot which misses the net, three miles to the right. The opposing team picks up the loose rubber and in seconds have flown the zone going the other way. Because all of the Oiler forwards are taking their sweet ass time back-checking (hoping someone else will get the puck back) the opposing team is able to keep control of the puck in the Oiler zone and generate 5 shots on an exhausted and tail-spun defense. They block the first shot accidently because they can't even get out of their own way, the goalie saves the next three,  but eventually lets the fifth one in and wishes he was back in the AHL.

Now let's do our calculations. The Oilers got one shot away (misses and blocked shots count remember) so their Corsi increases by +1. The opposing team got +5 (1 blocked, 3 saves and one goal). Therefore, that 5 man Oiler unit and/or individual on the ice would have a Corsi of either -4 (don't make me show my work) or 20% depending which way it was being expressed. Follow?

Over the course of a season these numbers can show how well you drive offense or how poor you play defense and can show how well you push the direction of play. Typically these Corsi values are expressed in %. A good player or team has a Corsi over 50% meaning that they generate more than half the attempted shots in any given game. As an example, the Oilers had a team Corsi last year of around 44% ( I think. You can go look it up yourself, I can't. It's like looking at photos of a car accident or of the Canucks' jerseys). That means that on average, the Oilers only took 44% of the shots in any given game. Bleccchh!

 NOTE: There are many extensions of Corsi values. Its a lot like the nomenclature of techno music from the early 2000's. You just add a word/words to the existing one and its a whole new thing (House - Deep House or Hard House -  Deep Acid House or Hard Acid House etc. It was endless). Like Corsi Rel is the Corsi value relative to the rest of  team or Corsi For includes only the number of shots generated and excludes the shots against portion. There are also values for 5x5, 5x4 and 4x5. Corsi For 5x5 or Corsi Rel 5x5 etc. There are as many different ways to name Corsi as there are John Tortorella tirades.

FENWICK: Not the name of the in-dressing room butler, although that would be really cool if teams had one!

Taylor Hall: FENWICK!!
Fenwick: Yes, Master Hall?
Taylor Hall: Did you remember to bring my soda water?
Fenwick: Yes, Master Hall
Taylor Hall: Good. Poor me a glass, would ya? And don't forget the lime this time!
Fenwick; Yes, Master Hall

Actually, Fenwick is just lazy man's Corsi. and is also named after the dude that came up with it. It's exactly the same except you don't count blocked shots when doing the calculations. All other rules and variations are same as above.

VOLLMAN SLEDGEHAMMER:  Surprise! Another analytic named after its inventor! These are player usage charts that show how teams use the players on their rosters and how well they preform in those roles. I first came across them at Lowetide.ca ( probably my favourite Oilers blog site - check it out here). The guy has a major hard-on for these things.

See the graph below? The y-axis is QualComp (remember that from Pt.1???) and the x-axis is what percentage of zone starts are taken in the offensive zone. The size of the circle is the number of minutes that particular player played and the color indicates Corsi (bluer is better)

This looks more like a balloon launch than a graph


Now, let's simplify things a bit and put it in an average persons terms. Below is a similar style chart except that here we are measuring number of confrontations or disagreements won in the Maier household. QualComp is still on the Y-axis, but this time number of total confrontations is on the x-axis. Size of the circle is equal to number of disagreements going that individuals way and the color is blue because my daughter said I had to - it's my favourite color.


My circle is high on the QualComp because, of the group, I am the only man and I fold like a house of cards. My wife has the highest # of confrontations since she is home all day with the kids while I'm at work. She also wins her fair share of the battles. Rowyn, our eldest is low on the QualComp faced because although she is only 5 and a half, she is already smarter than us but is lower on wins as we are still bigger than her. Syd is only 21 months, but wins pretty much every confrontation she has as they usually only entail, "Feed me", "Pay attention to me", and "I'm poopy."


NOTE: As a former science major, I have noticed how mathematicians and scientist in general are of the largest category to name discoveries and/or inventions after themselves. I think it might have to do with low self esteem and an insecurity towards recognition. Conversly, athletes typically ooze self-confidence and never have anything as their name-sake. Is the slap shot called the 'Geoffrion'? Is the goalie mask called the 'Plante'? Is to stomp on your girlfriends' chest the 'Varlamov'?

WOWY: Stands for With or Without You. You've heard the saying, " He's a player that makes everyone around him better."? Well, this is a method of plotting of various stats like shot differential, goal differential, zone start differential etc. while a particular player is on the ice with particular team-mates and how those stats change when he's on the bench (or in the press box). It actually singles players out with hard numerical data to demonstrate how it is they can elevate the play of the team around them (ex: Sidney Crosby) or how badly they bring everyone else down - like Vancouvers' Zach Kassian.

Could you imagine what it would be like if we had charts like this for real world situations? Man, it could make dating really hard for some fellas....

Lucy: I'm sorry Steve, I just don't think it's working out between us anymore.
Steve: Aw, c'mon baby! How can you say that? We're so good together!
Lucy: Actually, no Steve, we aren't. My chart right here shows I have 25% more fun with John and               I'm at least 33% more sexually satisfied with Tom. In fact, you finish dead last in almost every           category compared to any other guy I've been with.
Steve: Ah, crap.

NHLE: Or National Hockey League Equivalency. Yup, some uber-nerd(s) out there compiled all of the data for all of the players who have played in any other league other than the NHL - KHL, European, the minors, junior etc. and tracked what they achieved points-wise once making the jump to the NHL. He/she/they then calculated averages for all of those results and came up with mathematical equivalencies for any given player from any one of those leagues at different points in their careers and/or ages.

Example: Joe Smith is a 19 year old in his third year with his WHL team. His box-cars (I'm testing you) are 65GP 38-45-83. Based on his NHLE, we can expect him to score in the neighbourhood 30-35pts in his first full year in the NHL. *Note: I don't know if this is the actual NHLE for these stats, I made them up for demonstrations' sake. How accurate are these predictions you may ask? I don't know, I don't practice witchcraft.

There you have it. A beginners guide to some of the new "fancy" stat work. As for the controversy surrounding it, whether or not they have any value, I'm pretty m'eh on the whole thing. I'm certain it has merit, but I'll leave it for GM's, coaches and their staff to use accordingly in their team decisions. They get  paid for that sort of thing. I don't intend to delve into it too deeply here and bore anybody with paragraphs of data, that's for sure. I'll let the bloggers with WAY more time on their hands do that.

 I guess all I'm saying is that I'm a fan of watching the game - of taking in all of the things that make me enjoy the game - like sniped goals, bone-crushing hits and sweet sauce passes. So that's what you'll find here. Discussion on the happenings of hockey games, player transactions and general news with a little less numbers.

I hope I never get to the  point where I leap out of my chair during a game and scream at the top of my lungs, " Did you see that, baby!!! Eberle just raised his Corsi by 2!! Woot!! Woot!!"
















No comments:

Post a Comment