Friedman's 30 thoughts - analytics say Oilers' luck should turn

[djsFlames]

http://www.sportsnet.ca/hockey/nhl/e...eal-canadiens/

Claims that PDO indicates Oilers have been very "unlucky" and could indicate a turnaround, if the law of averages are to play in.

Anyway, the theory behind PDO is the more you go above 1.000, the luckier you are. The lower you go, the unluckier you are. The critical part is this: eventually, your luck is supposed to even out.

You know who is dead last in this category right now? Edmonton. The Oilers’ total is .961, 11 points below Columbus. The last team to finish that low for a full season? Atlanta, who posted a .957 back in 1999-2000.

Someone in the Edmonton organization did something awful to a black cat.

Now, here’s the biggest question: Is there any way on earth the Oilers can sell that?

Edmonton’s taken a deep dive into the analytic wading pool this season, and, when things were rough early in the year, the organization’s mood was positive, because the underlying numbers weren’t bad. And the PDO was low — very unlucky.

Thoughts on Flames:

21. After the 2014 Calder Trophy voting was announced, one exec I regularly talk to was apoplectic (and no, he doesn’t work for Calgary, have unruly hair and refuse to tie his tie). “No problem with Nathan MacKinnon winning,” he said, “But how does a 19-year-old centre with 22 goals finish eighth?” Sean Monahan is on pace for 26 this season, but more impressively, he’s doing it without veteran support down the middle.

Asked if Monahan is part of the leadership group, GM Brad Treliving paused before answering, “Yes. You don’t just want your young players to be part of the team, you want them to feel they can be more than that.” As injuries ravaged their forward units, Treliving said Monahan was nudged even further. “He’s younger than some of the guys we called up, but he’s played more games. Could he show them how to prepare and what it takes?” It appears the answer is yes.

22. The Flames are the analytics target right now, with their underlying numbers worrying the pocket protectors. Treliving, who dabbles in this, says he’s not too concerned.

“If we were still winning games the way we were earlier in the season, that would be a problem.” He’s referring to nights like Oct. 15, when Calgary beat Chicago 2-1 despite being outshot 50-18.

[PeteMoss]

Teams end up around 1.000 but being above or below 1.000 isn't luck or unluckiness. Good teams have high PDOs year after year.

[burn_baby_burn]

You need to be good to be lucky. You can't try to get away with backup goalies and AHL level defenseman (among other problems) and expect to have good luck on a consistent basis.

[_Q_]

I'm convinced that Hartley's system is his way of trolling the analytics folks.

I hope he continues with this for 5 years, wins a couple of Stanley Cups and makes a bunch of heads explode in the process.

"Advanced stats" are no good.

[kevman]

Yup, no reason to apologize for good goaltending. Or in the case of Edmonton, teribble goaltending.

[Dion]

This made me laugh....

30. During my third-year of university—while getting into the late rounds of our annual playoff pool—one of the guys picked Jock Callander, a 31-year-old forward who had 112 points for Muskegon in the old International Hockey League.

We thought he was insane. He said he read somewhere the Penguins might use him, so why not take a chance? Callander was indeed called up, and he grabbed four points in 12 games as the Penguins retained the Stanley Cup. The guy who picked him won the pool (by two points, I think).

The rest of us swore about Callander for years, but he had a real good career, playing almost two decades. It was nice to finally meet him this past Sunday while he was working for Colorado’s AHL team (Lake Erie) in Toronto.

[djsFlames]

Originally Posted by _Q_

I'm convinced that Hartley's system is his way of trolling the analytics folks.

I hope he continues with this for 5 years, wins a couple of Stanley Cups and makes a bunch of heads explode in the process.

"Advanced stats" are no good.

That these stats can argue FOR the Oilers (claiming 'bad luck') and against the Flames, says enough about how much weight they should carry.

Yes, things average out. And the Flames unsustainably high percentage goaltending did fall back to earth, but it did not affect the consistency of them winning games.

I know our shooting percentage must be well above average, but if you look at where the majority of our shots are being taken game after game, versus our opposition, we're often shooting from much higher probability locations on the ice. We're averaging less shots per game than most, but making our shots count because the majority of them aren't from the perimeter.

[devo22]

I look forward to the day when people run out of both the sad excuses for the Oilers ("oooh, they are sooooo unlucky right now!") and the tedious forecasts of how the Flames are going to crash and burn ("just wait, their advanced stats are sooooo unsustainable!").

I'm not demonizing advanced stats, I still think they are helpful. But there's a point when it becomes just laughable ... one team is winning and another team is losing, and that has nothing to do with Corsi, Fenwick or PDO. Advanced stats can never be more important than the win column.

[Locke]

Well it cant possibly stay as bad as it is.

Needless to say, whatever you read out of the stats sheets, the Oilers havent lost 10 in a row because they're 'unlucky,' they suck out loud.

[CaptainCrunch]

The Oilers record has little to nothing to do with bad luck.

for every one goal allowed from the goal line they are allowing multiple times more due to lazy play, bad coverage, giving up on the play or just plain inability to execute.

There is too much reliance on advanced Stats, they should go hand in hand with actually watching a team play and understanding how they play. Instead too many people are using it as a substitution to watching a game.

Its a lazy mans way to make themselves look smart.

I mean the whole possession thing is kind of ridiculous. If you watch the Flames they don't possess for long periods of time in the offensive zone because they score more often on the rush or with a quick transition, or whatever. So some team comes down to the Flames end and grinds in the corners and does some perimeter stuff for 20 seconds, and the Flames turn it over and a defenseman creates a 2 on one and Calgary scores in about 8 seconds. Well the team with possession didn't get a chance but got a punch of offensive zone time and the Flames get punished statistically for being efficient.

If you want possession stats to be properly analyzed maybe they should be combined with average time to score on possession or chances generated per possession cycle.

But the advanced stats are goofy when they say that the Oilers are merely unlucky. I don't see anything unlucky about their game but I do see a ton of deserved results.

[dino7c]

lol bad luck...I hope the Oilers "pick it up" and get themselves out of a top pick


its December 3rd and the Oilers have not beat a team in their own conference, luck these analytics guys are something else

[Coach]

I would believe the Oilers "luck" may turn around if I didn't also believe that luck, in the context of sports, is created by hard work. You want the opposing defenseman to cough up pucks randomly in their own zone like yours do night in and out? You have to put pressure on them constently, and then eventually they don't feel like they have enough time and make mistakes. You want a puck that takes a crazy hop off the glass to the front of the net? Well that implies that you actually try to put the puck behind the D and skate for it rather than trying to beat them 1 on 1,2 or 3. Need the other team to take penalties? How about skating? Yeah, that might work to create some hooks and holds.

I also don't consider constantly leaving guys unchecked in front of your net or in the high slot as "unlucky".

Blame a game or two on some tough luck and bad bounces? Sure. A 10 game losing streak (or 10 year, depending on how you want to look at it)? Not so much

[automaton 3]

The Oilers are terrible so their PDO fits - there is nothing unlucky about it.

[GioforPM]

I'm not an anti-analytics person by any means. That said, the Oiler's defiance of the stats is explainable to viewers of their actual games (which I am quite often because I like a good laugh). It's because the Oilers defensive miscues are not just mistakes, but rather huge blunders. Combine that with sub-par defence and goaltending and opponents' chances are going to be gold-plated much of the time.

This sounds like the reverse of the Leafs' corsi excuses last year (we allow lots of shots but they are poor quality). But in this case it's true. Arizona didn't play exceptionally well against Edmonton. But the blunders by Hall and Schultz, combined with Scrivens not being able to make a big stop, meant goals against.

[sharkov]

I hope their luck turns in a way.... I don't want them to get Eicchel or mcdavid

[Erick Estrada]

PDO here is simply doing a disservice to the Oilers and their fans. Team is bad. Terrible actually. There is no doubt the losing streak will eventually end and quite possibly be followed up with a short stretch of success. That doesn't change the fact that the team simply isn't getting better after missing the playoffs for the last 8 years.

[Resolute 14]

The argument that the Oilers PDO is unsustainable and should increase at some point is fair opinion based on history. Friedman says this should happen because the Oilers current PDO would be historic in its awfulness. However, he fails to consider one thing: the Oilers themselves might be historic in their awfulness.

[nik-]

When all you shoot is muffins from the perimeter and you give up five 2 on 1's a game, your PDO is going to suck.

Sorry, that's not luck.

[Locke]

Originally Posted by nik-

When all you shoot is muffins from the perimeter and you give up five 2 on 1's a game, your PDO is going to suck.

Sorry, that's not luck.

Agreed.

Watch the 2 shorties against Arizona. Those were not 'bad luck,' those were bad players. From front to back.

[ernie]

I think advanced statistics are useful. There are certainly general trends that can be "relied" on and if building a team you are going to feel more comfortable going forward with a strong possession team. However, they are just one piece of information. The reason people looked at the Leafs last year and expected a fall was because in addition to the poor possession numbers the goaltending was likely playing out of their minds. In short the advanced statistics suggested the team was overperforming in the standings and then looking at the reasons it may have been a result of unsustainable goaltending (and other things).

So I can see how the same thing can be said for the Flames...a month ago. I'm certainly not a Flames fans but Treliving is correct. They are not winning the same way. While good goaltending is certainly a major part of things the team itself is playing extremely well in other areas. So while I'm not sure the goaltending is going to maintain the level they are at and some drop may occur I'm not convinced it will be a precipitous drop.

The Oilers on the other hand are going with inexperienced goaltenders. The numbers appear lower than typical for these goaltenders but can typical even be defined for them as starters or as 1a/1b type guys? They have under 140 games combined. So while the advanced statistics suggests a rise of some sort that rise is also dependent on goaltending getting better (for one instance). Will it? No idea. And in that case mid-range possession numbers coupled with sub par goaltending is likely to land you at the bottom of the standings or close to. And then you have another issue that is coming to the surface with the Oilers if you listen to the opposition: you can beat htem by pressing early and getting ahead by one or two OR you can beat them by playing passive and waiting for the prime scoring chances they will give up. Either one of those scenarios is likely to lead to somewhat misleading possession stats - teams with leads sit back and passive teams looking to capitalize on mistakes sit back. And of course, with the advanced stats there are trends but not everything is going to fit that trend all the time. (this also ignores scoring chances. I think if you looked at scoring chances the Oilers are getting absolutely trounced in that department. Often more shots means more scoring chances but again anomalies happen).

And as mentioned above...to disregard something because it would be some historically low number and assume it come back because of that ignores that some team has to be that historical example.


I guess in the end it aggravates me when I see things like "The Corsi suggests they should be better!". No the Corsi suggests they should be better if in addition they have competent goaltending, are creating equal chances, are getting shots through and on net (I don't think the Oilers actually do) etc etc etc. It's an indicator, not a law.