Predictive Value

So, here we are moving into the second round of the playoffs with, in my opinion very few surprises.

The Calgary Flames have to be the David to Anaheim’s Goliath but I’ve been telling you all season that the Flames should NOT be discounted.

I suspect they will have a tough time advancing to the 3rd round just based on Anaheim’s size and skill but, then again, folks who have been trashing Calgary for their inability to show well in the “advanced” stats don’t understand what Calgary is doing, how and why.

I thought it would be interesting to go back and take a look at what the Statzis were predicting this season based on “puck possession” and hold their feet to the fire a bit.

The following graph is lifted from an article on SportingCharts.com and you’ll see pretty quickly how poorly the author performed in predicting how the teams in the WC finished in league standings based on their possession metrics.

bob-nhl-5_650x393

The first thing you’ll notice is that almost half of the teams predicted to finish in the playoffs didn’t even make the post season.

The other striking errors are the Kings finishing 1st in the west and Minnesota finishing 13th.

And therein lies the major problem of using past performance to predict future success or failure using nothing more than a statistical approach without raising your eyes from your spreadsheet to rationally assess how the ever shifting lineups of teams are affected by new additions, injuries, progression by young players, coaching changes and a host of other things that are reflected in the “advanced” stats.

If you’ve been following along, you will know that I have been predicting big things for the Wild all the way back to last season and, despite an early season scare caused by dreadful, goaltending, the Wild have been the best team in the NHL for months.

My being bullish on the Wild has been based on roster makeup, systems execution and coaching.I could go on at great length about those things but I’m going to address them a little later in this post when we look at some predictions for the second round.

Before leaving the Corsi/Fenwick debate behind, I’d like to remind you that there are many devotees of “advanced” stats that will tell you that those metrics can accurately predict who will win in the playoffs and win a championship. Really?

Here are the Top 16 CF% teams in the league from the 2014/15 season…let’s see how predictive of post season success that metric has been:

LAK 55.4

CHI 53.6

DET 53.5

TBL 53.0

PIT 52.8

NYI 52.7

NSH 52.7

WPG 52.5

CAR 52.5

DAL 52.1

STL 51.8

BOS 51.7

SJS 51.5

WSH 51.4

FLA 51.3

MIN 51.0

I think you would have to agree that using “possession ” stats to predict playoff success is pretty much useless and while any fool would agree that having the puck is better than not having the puck, there are so many other moving parts in play, that ignoring them leads to embarrassing moments like the article linked to above.

I’m sure the author was all agog that he was on to something brilliantly insightful but Maggie the Monkey could not have done a poorer job.

So, what about the Flames?

Why have they been so successful despite being 28th this season in CF%?

If you watched the Flames/Canucks series with a critical eye, you will have seen exactly the factors that can overcome or ignore the current “possession” fad a lead a team to win more games that they lose.

Bob Hartley (who should win the Jack Adams trophy hands down) has designed his systems play around the personnel he has and has and has received total buy in from his team.

There are several facets to what he is preaching:

1) Relentless forechecking and puck retrieval.

2) Pounding the opposition D in their own end to wear them down.

3) Keeping opposition shots to the outside

4) Blocking shots

5) A mobile D that turns pucks the other way with accuracy and speed.

5) Speed through the neutral zone

6) Crash the net.

There’s nothing magical about his formula and it only works if executed relentlessly but it doesn’t lead to Corsi success…it overcomes that metric.

Interestingly enough, I think you’ll see that Anaheim plays a very similar style although their personnel are bigger and more experienced.

That likely gives the Ducks the advantage in the series between the 17th and the 28th ranked CF% teams but I expect the Flames won’t go quietly and could upset the Ducks if they outwork them and execute better.

I’ve likely watched about 50-55 Flames games this season and I can honestly say that I have never seen the team take even ONE shift off.

We’re about to find out if they want it more than the Ducks.

Advertisements

3 thoughts on “Predictive Value

  1. If Corsi was important, teams would be lining up to hire Dallas Eakins. One thing is important, scoring more goals than the other team.

    If your strategies lead to generating more good shots, great. The key is, they need to be good shots. As Gretzky said, “No shot that misses the net is a goal.”

    There is an hilarious line of reasoning at Lowetide tonight. Evidently, if you look at Corsi Against, Corsi Rel. (but not Corsi for), Zone Starts and Quality of Competition, you can tell if a defenceman is shutdown, puck moving or offensive! You might even be able to tell if he is any good! Obviously, watching him actually play is a complete waste of time.

    I am beginning to see why Brian Burke is so dismissive of the Statzis.

    Like

  2. Back in the seventies (I know, I’m showing my age) Clare Drake had us do several team and player evaluations as part of the Advanced Hockey class. The process was exhaustive, but I learned infinitely more filling out one form than I ever have spending hours looking at stats.

    After all, Ladi Smid looked pretty good to the fancy stats crowd when with Petry, but “good skater, blocks shots but handles the puck like a grenade, can’t make a 15 ft. outlet pass and panics under pressure” was always the real picture.

    Sometimes it seems like a curse, not being able to accept the “Taylor Hall is the second best player on the planet” (after McDavid) story.

    Like

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