Oilers, Leafs, Flames near the bottom, Sens near the top - Analytics report

[Del Preston]

That image pretty much sums up Eric Belanger's career as an Oiler.

 

[Nona Di Giuseppe]

People had the Leafs at 6th to 9th in the East after they finished 26th in the NHL? in what world did that happen? If anything it gave people even more ammo to say the Leafs would suck.

Before spouting off on my comment, how about checking the context ?

You make no sense.

 

[Spring in Fialta]

This thread is gonna go far.

 

[swiftwin]

Yeah most analysts nailed the leafs being a lottery team last year, so I guess we might as well not even try this season

lol, analytics is not the same as analysts.

 

[Morbo]

Pretty hard to argue Florida wasn't the worst. -59?? yeesh.

 

[SmellOfVictory]

Analytics find averages and put them on a massive pedestal. Good teams are good because they defy averages. Especially in a game that is so emotional and energy based, I have trouble with such stats.

They do. They tend to defy averages based on providing a superior shot differential.

In terms of percentages, I don't have any citations on-hand, but there has been a strong correlation found between shot attempts (Fenwick, Corsi) and scoring chances, and there is very obviously a strong correlation between scoring chances and goals.

So yes, as an earlier poster said, goals determine the winner. However, goals are variable based to a notable degree on luck (not majority, but notable) and thus very difficult to predict. The ability to outshoot, however, is based substantially less on luck and thus easier to predict. Just because a team outshoots does not mean it will outscore and thus win, but long term it is the best way to analyze hockey. It's certainly better than "lol but this guy hits hard and scares other guys" or "look at the heart this team plays with".

 

[Ishdul]

Oh what wonderful responses we have had so far in this topic.

 

[SmellOfVictory]

The Leafs adding a top 5 league-wide volume shooter in clarkson is most definitely going to help their stupid shot differentials. As someone that hates the leafs I have to play they had a lot of skilled players scoring off the rush instead of dumping pucks on net and hoping for rebounds. The ridiculous advanced stats narrative (hurr durr) will tell you this means the leafs were unsustainably lucky because all it looks at is shot differential and league averages lol. So stupid.

You realize the reason people use advanced stats is to limit the faults of human observation as much as possible, right? The argument your making is that a single non-expert observer (you) has done such a good job of viewing every single Leafs play of the past x number of games that this single non-expert observer provides better analysis of the games than indicators (shots and percentages) collected from every game by multiple people who are paid to do so.

Did you watch every second of the game intently? Did you never stop to talk to your buddy, eat some food, or drink some beer? Were you 100% sober for all of the games? Did you never let your mind wander to any other topic? Did you always follow the puck from end-to-end without watching what was happening on the bench or in the crowd? Did you keep track of the number of shots off the rush vs the number of dump-ins, or was it just a "feeling" you got from watching the games? Were you entirely unemotional while watching? Did you track the distance of each shot from the net?

Because those are just a chunk of the human factors that lead to faulty, garbage analysis of sports, especially fast-moving ones like hockey, and the reason that some people prefer numbers that, while imperfect, are better than a single non-expert observer (or a single observer of any kind).

In conclusion: hurr durr.

 

[The Legend]

Oh well, numbers are always right, we shouldn't even play hockey this season.

You do realize that the point of statistics isn't absolute certainty right?

 

[Pyrophorus]

Before spouting off on my comment, how about checking the context ?

You make no sense.

Frankly, I hardly saw 6-9, I would say on average people placed us at 12th. There were quite a few who did say 15.

I can see 6-9 this year as "credible" predictions.

 

[Pyrophorus]

lol, analytics is not the same as analysts.

I should ask Kyper and MacLean, to at least say "ok, we were wrong",
like many others.

 

[AUAIOMRN]

A perfect example of someone who doesn't get it.

Analytics are used to determine the most likely outcomes. Teams that tend to get heavily outshot tend to lose the most games.

Hockey analytics are funny. Instead of finding a model that fits reality, they create a model and say "this is what reality should look like". Except I've never heard of any model that is better at making predictions than a typical TV analyst.

 

[Fourier]

This is "regression to the mean" on steroids. I would not give this much weight.

As an Oiler fans we know very well where the Oilers have issues. Being lucky on the pp is not one of them.

 

[Worraps]

Ottawa is a gutsy squad and I like them as a perennial playoff team for the foreseeable future but seeing them ranked ahead of teams like Pittsburgh, Boston, and Chicago just goes to show how awful Vollman's brand of "analytics" are.

Unfortunately for the Sens, players like Crosby, Malkin, Chara, Bergeron, Toews, Hossa, and Kane aren't so easily regressed down to the mean.

 

[Worraps]

Hockey analytics are funny. Instead of finding a model that fits reality, they create a model and say "this is what reality should look like". Except I've never heard of any model that is better at making predictions than a typical TV analyst.

Pretty much this.

How anyone can take this stuff seriously is beyond me.

 

[SmellOfVictory]

Hockey analytics are funny. Instead of finding a model that fits reality, they create a model and say "this is what reality should look like". Except I've never heard of any model that is better at making predictions than a typical TV analyst.

It predicted the Wild dropping like a stone after a hot start two seasons ago. A style of analytics, the stuff EA does, has done a very good job of predicting Stanley Cup contenders (IIRC they predicted the Hawks' first post-lockout win).

I like advanced stats a lot more for individual players than teams, as there are so many variables with teams, but even on the team level where they're less good at predicting, variation from the predicted result is generally fairly well-explained (PDO) in a systematic manner.

 

[Ishdul]

Hockey analytics are funny. Instead of finding a model that fits reality, they create a model and say "this is what reality should look like". Except I've never heard of any model that is better at making predictions than a typical TV analyst.

What, exactly, are you basing this on? Simple looks at possession stats run counter to your example and that's what most people consider to be hockey analytics. Every sensible analytical prediction is meant to be an average of all the possibilities but that is not them eschewing variance.

 

[smackdaddy]

Advanced Stats Gone Wild.

 

[I am Canadian]

Can the former underdogs Ottawa perform now that they are a favourite favourite for the playoffs?

Probably, who knows.

 

[Erik Alfredsson]

Yeah most analysts nailed the leafs being a lottery team last year, so I guess we might as well not even try this season

I'm pretty sure that was Ottawa last season and the season before.

 

[Ishdul]

I'm pretty sure that was Ottawa last season and the season before.

They were projected to be a bubble playoff team last year by most analytical models. Plus every reasonable person was down on the Senators for 2011-2012.

 

[AUAIOMRN]

What, exactly, are you basing this on? Simple looks at possession stats run counter to your example and that's what most people consider to be hockey analytics. Every sensible analytical prediction is meant to be an average of all the possibilities but that is not them eschewing variance.

It's not about making predictions, it's about making better predictions than a human can make. More importantly it's about making non-obvious predictions. Anyone can say that the good teams from this season will be good next year. Picking Toronto to come in second last and Ottawa to come in first are certainly non-obvious predictions, we'll see how accurate they are.

If there is a comparison out there showing how various analytics do against "common knowledge" I'd be interested in seeing it.

 

[Warden of the North]

Good thing luck has nothing to do with playing sports.

 

[Aceboogie]

I support analytics and advanced stats. They are helpgul in many areas. However in a dynamic game like hockey with so many variables in the way its played, in the dynamics of the team and the dynamic of the players themselves, there is not single model that can accurately predict a result id be comfortable with.

I agree that there is luck, and outleir on ice s% can regress, and I also agree that injuries play a big role (which is why I am high on the Sens next year). I disagree with some aspects of this article.

First for the Oilers, it doesnt take into account subjective data such as injuries players played with during the season and how that can affect them. Such as RNH and Eberle. I firmyl believe their offensive output was largely effected by respective injuries this year. To what degree they will bounce back is pure speculation, but a point to be taken into account.

Second is development of players. A young team such as the Leafs and Oilers will have their players progress year over year. Much more so than older teams. By that own merit, these teams will get better. Kadri,Gardiner,Kessel,JVR,Hall,RNH,Eberle, Schultz etc are all another year better. History shows that players can progress leaps and bounds in their prime development years. See Stamkos and other young elite talents. This model only seems to take a snap shot of the current stage of the players, but doesnt factor in their growth.

Lastly is the offseason moves made. I wont try to go full homer and pump my GMs tires because he didnt have some miraculous offseason, but he did address some key issues, many of which prevalent in Vollmans model. MacT has stated he uses advanced stats to make decisions, they are not the sole resource used. But when he makes move he takes possession #'s and shot differentials into account. His acquisition of Perron was made to address possession numbers and well as shot differentials as he is a player with good possession numbers in STL (upgrade over MPS). Ference is a an addition to play a shut down game and prevent chances against (upgrade over N. Schultz). Jonessu can play a two way game to and should also help with possession and shot differentials, so should Gordon (upgrades over Horcoff and Smyth). His model uses the stats of Horcoff, Smyth, Potter, Whitney, Belanger, Fistric, MPS versus the upgrades in Perron, Ference, Gordon, Belov, Larsen, Klefbom, Jonessu etc.

He also brought in puck moving D for the bottom pairing. Grebsekov, Belov, Larsen and Klefbom can all move the puck very well and should help with possession numbers. Gone are the days of bottom pairing face punchers who cant move the puck. The better you can move the puck to the Oiler forwards, the more chances your getting.

These moves were made to address shot differntials and possession #s, if Vollman is simply using 2013 stats, then these improvements wont show up in the rankings

Tl;DR I think the Oilers improved in the advanced catorgories. These stats dont show the development of either the Leafs or Oilers. I have the Oilers fighting for the playoffs next year and the Leafs regressing a touch but still making it

 

[Asher]

I wonder how he counts a player who is playing injured and puts up fewer points because of it.