Jump to content

Claimed:Statistically Advanced - Week One: Intro to Advanced Stats


Recommended Posts

Statistically Advanced - Week One: Intro to Advanced Stats

 

Statistically Advanced is a weekly feature which will give an in-depth look at a VHL team, discussing said team's weaknesses and strengths, using fancy "advanced" stats to predict trends. Each week, I'll analyze a different VHL team. Warning: If you hate numbers, this post will likely not be for you.

 

This, being the first post, will be a brief intro to the world of hockey's advanced statistics. If you have any intention of reading my future posts, and finding out what I think (and what math thinks) about your team, I'd strongly advise you to read this post, and read through the supplementary articles I've provided to become acquainted with the important stats.

 

Calculator-Pencil-300x199.jpg

Woo, math!

 

Bear with me, this probably won't be the most interesting post I'll do, but once I start analyzing teams, you'll want to have read this one.

 

Probably the most commonly used advanced stat is called Corsi. This is a basic measure of how often and how well a team possesses the puck. Also known as a team's "Shot attempt differential", it's calculated by subtracting a team's shot attempts against (i.e. the opponent's shots on goal, shots missed and shots blocked) from the team's shot attempts for. Easy, right?

 

A team with a positive Corsi number tends to do well, since they have more scoring chances, and allow the other team to have fewer. Corsi can be calculated for an entire team, or an individual player.

 

You'll see here that Leafs player Joffrey Lupul doesn't really like Corsi, but he's wrong. Love you Joffrey, but still.

 

Now, on to Fenwick. This stat is basically the same thing as Corsi, except that it excludes blocked shots. There are several reasons for this exclusion, and you should feel free to read all about them and make your own decision as to whether you prefer Corsi or Fenwick. I personally prefer Corsi, because of its larger sample size.

 

The final stat I'm going to look at today is PDO. This is, effectively, a rough estimate of how lucky a team has been recently. Probably one of the simplest "advanced" stats, PDO is calculated by adding a team's save percentage to their shooting percentage. If the number comes out to anywhere below 98%, that team's probably had some bad bounces, and are likely to turn it around. Above 102%, and they've probably been pretty lucky, and aren't likely to maintain their current success.

 

This helpful chart can give you a rough idea of how it works. Green is normal, red is deviation:

702vt0z.jpg

(Disclaimer: I didn't make this chart.)

 

Shooting percentage is a relatively good marker for how lucky an individual player is. The average NHL shooting percentage is around 8%. If a player's shooting well above that (and I'm talking up into the 17% or higher area), then you can probably bet that they'll come back down soon enough.

 

So, although it's impossible to actually measure luck, history has shown that teams which are doing well, and also have an absurdly high PDO, are very likely to stop doing so well in the near future. If they're doing well and have an average PDO around 100, then congratulations! That team is actually very good at hockey, and all its players are absolute beauties.

 

Before I finish this first installment of Statistically Advanced, I think you should all be aware of the hockey stats I dislike, because I'm a conceited bastard and I think everyone cares what I think.

  • Hits: I've always been amazed at peoples' fascination with the hits stat. Guys! Wake up! What's the one thing that needs to happen in order for your team to make a hit? The other team has to have the puck! And, what can you not do when the other team has the puck? SCORE! 
    Historically, teams with high hits totals have not been a league's best teams. For example, take this year's VHL Hits per Game table:

    Pos Teams                    Value
    1   Wranglers                20.30      
    2   Legion                   19.70      
    3   Express                  19.30      
    4   Bears                    19.10      
    5   HC Dynamo                19.00      
    6   Americans                18.30      
    7   Iron Eagles              17.10      
    8   Reign                    16.30      
    9   Meute                    15.50      
    10  Titans                   15.20  
    

    There really isn't much correlation between this table and the current league standings. The Wranglers, leading the league in hits, are (with all due respect to the players and GM) not doing terribly well this year. Meute, on the other hand, are. And where do they rank in hits? 9th. That's second last, people! 

  • Goals-Against-Average: I'm sure most people know why GAA is not often the best measure of a goalie's talent. You could have a great goalie on a terrible team, who faces 40 shots a night and lets in 2 or 3 of them. He'd have a decent save percentage for one game, but 3 goals per game is not a great GAA. His team is to blame for giving the opposition so many shots (Cough Toronto Maple Leafs Cough). So, unless he absolutely stands on his head every night, the goalie might have a less-than-stellar GAA, and a good save %, which, in my opinion, is a more accurate measure of a goalie's performance.
    Don't get me wrong, GAA is often a decent measure of an entire team's defensive performance. But for just the goalie? Seems a bit unfair. (Disclaimer: I used to be a goalie, I may be extremely biased. Also, don't even get me started on considering wins as a measurement of goalie performance)

  • I'm absolutely certain there are other stats that I hate, but it's late and I'm tired and slightly depressed because the Maple Leafs are shit. (If you want a really good example of advanced stats nerds being able to say "I told you so", look at one of the many posts about the Leafs.) So, I might remember one in a future post. Stay tuned.

 

The VHL very handily has the available resources for the calculation of all these stats. So, over the next few weeks, I'm going to be making a statistical analysis of every VHL team. Each week I'll cover a different team, highlighting certain players, and giving you a general idea of how well that team is actually playing, and how likely they are to continue at that pace.

 

I'm going to start next week with the Toronto Legion. Keep an eye out sometime next week for my post about the Legion! Also, please let me know if anything is unclear to you, if you'd like to see something in particular looked at, or if you think that this whole advanced stats thing is stupid and I should just screw off. Please, feel free to call me out if anything I write is utter bullshit!

 

Seriously, I promise it won't be this boring in the future. You'll get actual information about your team that might actually be interesting and useful!

Can we just take a step back and realize that we want to analyze advanced statistics for a hockey simulator that is nothing but statistics itself? The statistics of the statistics will reveal deeper statistics.

 

This website is beautiful.

Can we just take a step back and realize that we want to analyze advanced statistics for a hockey simulator that is nothing but statistics itself? The statistics of the statistics will reveal deeper statistics.

This website is beautiful.

Haha hadn't thought of it that way. I guess I will actually be calculating statistics used to analyze other statistics

I love you.

 

I planned on doing something related to advanced stats with the VHL during the summer, but I was busier than anticipated at my internship (who would have though :P). 

Can we just take a step back and realize that we want to analyze advanced statistics for a hockey simulator that is nothing but statistics itself? The statistics of the statistics will reveal deeper statistics.

This website is beautiful.

statistiception

Content: 3/3 - I hate numbers. There's like, too many of 'em and stuff. But through this statistiception, you're gotten yourself quite a nice legion of fans. Keep up the good work!
 
Grammar: 2/2 - I didn't see a single thing. Either your math hides grammatical errors or you're a master of both, which frightens me.
 
Appearance: 1/1 - Yup, but where's the pie chart?
 
Overall: 6/6 - You done good.  (Y)
  • 8 months later...
Guest
This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...