Francis: Treliving filled with disappointment, anger after team's collapse

[Jay Random]

Originally Posted by Gaskal

Case in point - corsi is quite awful while used as a predictive model. It is much better seen as a performance measure.

Precisely. Goal differential is not a predictive measure of expected outcomes; it is a post-hoc measure of actual outcomes.

Shot-differential stats are, at most, useful as a diagnostic tool. What they tell us about this year's Flames is that the team was good at getting opportunities to put the puck on the opponent's net, and piss-poor at converting them into goals. What they tell us about several individual players on the team is that their shooting percentage declined dramatically from their career norms. Some of those players are quite young: if Sam Bennett can't put the puck in the net, it isn't because old age is catching up with him. The question is what went wrong with him and several other guys.

So we have to look at second-order diagnostics. This, in theory, is where the eye test comes in. But unless you're a professional video coach, the chances are extremely remote that you have the time to watch all 1,271 games in an NHL regular season and break them down. And that, approximately, is what you would have to do not only to assess your own players, but to have a full understanding of the competition. There are two teams on the ice in every game, and if you're content to analyse just one of them, you are at most getting half the picture.

So we're back at metrics – trying to apply uniform standards to many different people's eye tests so that the results can be aggregated in a reliable way. It's a hell of a problem, and no general solution has been found.

[VilleN]

Originally Posted by zamler

You can micro analyze things to a fault. Take for example audio equipment, a given piece of gear can check every possible mark but still sound mediocre. There is a point where science is not enough the art of making something great is what separates average from the best.

That’s likely because the science isn’t fully understood. There isn’t any mysticism going on here.


Sent from my iPhone using Tapatalk

[Resolute 14]

Originally Posted by stone hands

Advanced stats get a bad rap around here ever since they accurately predicted the miracle season was a mirage and extremely statistically unlikely to be repeated

Using anything to back your opinion up outside of the ubiquitous eye test gets you quite a ribbing around these parts

Actually, advanced stats got a bad rap when their proponents spent six months predicting the 2014-15 Flames would fall off a cliff any moment now and it never happened.

Then, the following season, with a whole new set of factors beyond corsi in play, those same people came back and yelled "I TOLD YOU SO!", as if a new season with new players, new factors, mental let downs, changes in other teams, etc., didn't actually matter. All that mattered to these people was that they were so desperate to be right that they leaped at their first opportunity to do so, even if they looked foolish in the end.

[The Boy Wonder]

Until analytics can cover every “why” question or to at least 5 levels of questioning they truly aren’t explaining or finding the root cause.

Flames have our shot opponents for the past 30 games give or take.

Flames have lost the majority of their games this season...

The depth that current analytics can go is to say “well they have a terribly unlucky shooting percentage”

Until analytics can show us why GGs systems produce shots that lead to this lower shooting % or show some other reason we have to just chalk it up to “luck”

Until the analytics gets to a baseball level which would require understanding each coaches system, how it is deployed in various game situations etc we can’t truly rely on analytics to predict anything

[Cube Inmate]

Originally Posted by Jay Random

Precisely. Goal differential is not a predictive measure of expected outcomes; it is a post-hoc measure of actual outcomes.

This is true of every statistic. It only becomes predictive and useful when (a) you can find a positive correlation to winning that's based on a variable you can control, and (b) you can figure out how to make that variable change in the way that correlates with winning.

The problem is that most of these "event-based" stats are not things you can control directly even IF you understand the correlation. Essentially, the stronger the correlation to winning, the less you can control the variable directly. Goal differential has a strong correlation...but we have very little control over that directly. CF% (or whatever)? Has a weak correlation with winning, but it's easy to design a system around the idea that throwing pucks on net from anywhere is a solution, so we do it.

The Flames can pay someone whatever they want to crunch these numbers. But if there were truly any value in these stats, then a lot of people advocating for their use would be better off making a living by gambling on sports.

[Enoch Root]

Originally Posted by stone hands

Advanced stats get a bad rap around here ever since they accurately predicted the miracle season was a mirage and extremely statistically unlikely to be repeated

Using anything to back your opinion up outside of the ubiquitous eye test gets you quite a ribbing around these parts

No, they have long gotten a bad rap around here because there are a large number of posters that understand their inherent flaws.

And your completely subjective and biased defense of them here is deliciously ironic.

[Shazam]

Originally Posted by The Boy Wonder

Until analytics can cover every “why” question or to at least 5 levels of questioning they truly aren’t explaining or finding the root cause.

Flames have our shot opponents for the past 30 games give or take.

Flames have lost the majority of their games this season...

The depth that current analytics can go is to say “well they have a terribly unlucky shooting percentage”

Until analytics can show us why GGs systems produce shots that lead to this lower shooting % or show some other reason we have to just chalk it up to “luck”

Until the analytics gets to a baseball level which would require understanding each coaches system, how it is deployed in various game situations etc we can’t truly rely on analytics to predict anything

It works well for baseball because it is a turn-based game, with player positions reset before each event (pitch). Football, same sort of thing. Even soccer, because it's so damned slow.

[Textcritic]

Originally Posted by Cube Inmate

This is true of every statistic. It only becomes predictive and useful when (a) you can find a positive correlation to winning that's based on a variable you can control, and (b) you can figure out how to make that variable change in the way that correlates with winning.

The problem is that most of these "event-based" stats are not things you can control directly even IF you understand the correlation. Essentially, the stronger the correlation to winning, the less you can control the variable directly. Goal differential has a strong correlation...but we have very little control over that directly. CF% (or whatever)? Has a weak correlation with winning, but it's easy to design a system around the idea that throwing pucks on net from anywhere is a solution, so we do it...

I tend to agree with the first part, but I don't think this accurately describes the outcome of the Flames' season. More to the opposite, actually—it seemed as though more often than not players were overthinking and overplaying their setups and shots, and the biggest problem with offensive zone chances was that they took too much time when generating them.

[Enoch Root]

So we are now at the point where watching the games, and analyzing them from what we see, is too subjective, and we have only sufficiently analyzed the play if we employ (weakly correlated) counting stats to verify our analysis?

Wow.

[edslunch]

Originally Posted by Textcritic

I tend to agree with the first part, but I don't think this accurately describes the outcome of the Flames' season. More to the opposite, actually—it seemed as though more often than not players were overthinking and overplaying their setups and shots, and the biggest problem with offensive zone chances was that they took too much time when generating them.

That seems to be one of the prevailing theories about why the Flame's high danger shots aren't going in - they have the distance but not the speed of execution to be dangerous.

[Pellanor]

Originally Posted by squiggs96

You wouldn't be my favourite teacher if this was your method. Let's say this is stats, algebra, or something complex. You're wanting to find x to five significant digits. The answer is 14.387. I put down 14.387. I clearly understand how everything works as I got the exact answer. You cannot guess the answer 14.387 when it's not multiple choice. If you could, you should be playing lottery numbers, the stock market, or setting up a booth in the mall to tell people their fortunes, since you have an incredible gift of guessing. If you think I copied the answer off of someone else, then give me another question and I'll do the same. It would also be odd how I copied the answer and still finished ahead of the rest of the class.

You giving me a zero for having the correct answer because I didn't write down what was going on in my head is wrong. As a student, I wasn't there to explain it to you or help you understand. I wasn't your tutor. If I got the right answer, then you can pat yourself on the back that you did a great job teaching me. If I get the wrong answer, I'll take the zero. That's a risk I'm willing to assume. There are problems I need to write down to work out. There are others I don't. I've often heard not everyone learns the same way, so it seems silly to punish someone for getting the right answer, because they didn't copy down every thought. If I see a sprinter get to the finish line first I give him the gold medal. I don't ask him what work he put in to get there.

How does this relate to the Flames? Many of us arrived at the conclusion that GG needed to be fired. Others contended throughout the year that because of advanced metrics that they weren't sure. After coming around to the same conclusion, we are told that our analysis wasn't done properly. Just because it too some longer to get to the conclusion, because they used some more data that got to the same point, doesn't make their conclusion better, when it's the same. Obviously this isn't getting to the answer of 14.387 on a math test, but it's not much different.

First off let me say that I totally agree with you on the math thing. It annoyed me to no end when I got the right answer but lost marks for not showing my work.

However this is a discussion board, not a math test. Being able to predict that GG is not the right coach for the team is all well and good, but it doesn't exactly help the discussion. When you show your work others can learn from your analysis. When the outcome doesn't align with the underlying stats, we can ask what other factors may be contributing, or if there's better stats to track. When it's time to look at a new coach we can analyze their previous teams and try to use that to predict how they may coach the Flames.

Now that's not to say you have to use fancy stats to have a good discussion. The eye test is still useful, but you still need to explain why certain things are failing the eye test. You may hate how much time GG spends looking at his iPad, but does that actually matter? If you base your predictions on your gut instincts, or don't bother explaining your analysis, how are we supposed to talk about it, and how are we supposed to give you any more credibility than a coin flip?

Gaskal and other posters have done great jobs in the past of breaking down plays, showing those of us with less real hockey knowledge insight into what flaws there may be in GGs system, or the play style of the team. We don't need fancy stats for great analysis and discussion. Just people willing to take the time to explain why they believe what they believe.

If you just want to say "I told you so", be my guest, but don't expect the same level of respect as those who take the time to show their work.

[Steve Bozek]

After we’re finished with all the statistical jiggery-pokery, it’s become clear to me that the 2017-18 Calgary Flames are simply “the gang that couldn’t shoot straight” (apologies to Jimmy Breslin). I think, after a full season, lack of skill (other than Monahan Gaudreau and Tkachuk) provides a better explanation of the low goals/shot ratio than bad luck.

[Strange Brew]

Originally Posted by Bingo

It's more than that, but really hard to pin down definitively.

From what I've read every shot within home plate qualifies as a scoring chance, but there are additions and subtractions with points that qualify only 3s as high danger and 2s as regular scoring chances.

Thanks - I truly didn’t know this. Guess there is some subjectivity/murkiness here but this paints the team in a slightly more flattering light, atleast for me. High danger might really mean high danger.

I am still left with the fact, or should I say observation, that Flames simply don’t play as well in bigger moments. Which quite possibly means they are not as good as the competition. Over a long season, teams and players coast and ratchet it up when needed. Maybe Flames are more consistent than we give them credit for, and they simply get outclassed when the competition elevates their game?

[Textcritic]

Originally Posted by Steve Bozek

After we’re finished with all the statistical jiggery-pokery, it’s become clear to me that the 2017-18 Calgary Flames are simply “the gang that couldn’t shoot straight” (apologies to Jimmy Breslin). I think, after a full season, lack of skill (other than Monahan Gaudreau and Tkachuk) provides a better explanation of the low goals/shot ratio than bad luck.

There is more to it. Several Flames players suffered historically bad shooting percentages, which I think points to a problem beyond absent skill: Backlund is quite consistently a 10–12% shooter, but was less than 7% this year; the same with Bennett; Frolik's SP was the lowest it has been in five years, and Brodie's less than half of where he has been shooting the past three years. Even though Giordano had a solid year his SP was lower by a bit than it has been the past few seasons.

There is clearly some evidence there to suggest that the key players are good enough, but execution was a serious, perplexing problem for too many of them.

[4X4]

Originally Posted by Textcritic

There is more to it. Several Flames players suffered historically bad shooting percentages, which I think points to a problem beyond a skill-issue: Backlund is quite consistently a 10–12% shooter, but was less than 7% this year; the same with Bennett; Frolik's SP was the lowest it has been in five years, and Brodie's less than half of where he has been shooting the past three years. Even though Giordano had a solid year his SP was lower by a bit than it has been the past few seasons.

There is clearly some evidence there to suggest that the key players are good enough, but execution was a serious, perplexing problem for too many of them.

Do you think it has anything to do with Gulutzan's stupid system, where he has the D pass it backward a couple times before engaging the attack? By the time the Flames are in shooting position, the opposing D are also in position, and so is the goalie?

It's been painfully obvious for a really, really long time by actually watching the games instead of counting stats.

[Textcritic]

Originally Posted by 4X4

Do you think it has anything to do with Gulutzan's stupid system, where he has the D pass it backward a couple times before engaging the attack? By the time the Flames are in shooting position, the opposing D are also in position, and so is the goalie?...

I think it is system-related, but don't see this specifically as the problem. Like I said above, it is more about overthinking and overplaying on the part of the players while in the zone, and not necessarily tied to a blueprint by coaches for moving the puck.

[4X4]

Originally Posted by Textcritic

I think it is system-related, but don't see this specifically as the problem. Like I said above, it is more about overthinking and overplaying on the part of the players while in the zone, and not necessarily tied to a blueprint by coaches for moving the puck.

I'm seriously starting to wonder if you, Jiri, and especially Bingo are getting paid by the Flames. Good for you, if so. They're getting their money's worth.

[the2bears]

Originally Posted by 4x4

i'm seriously starting to wonder if you, jiri, and especially bingo are getting paid by the flames. Good for you, if so. They're getting their money's worth.

lol...

[Cali Panthers Fan]

Originally Posted by 4X4

I'm seriously starting to wonder if you, Jiri, and especially Bingo are getting paid by the Flames. Good for you, if so. They're getting their money's worth.

Ok, seriously. These kind of comments need to end around here.


And why is it so hard to have a discussion where the merits of statistical analysis are included somewhere in the thought process? It doesn't mean that we should only look at the stats and form conclusions based on them alone, but why should we not include them at all? Why not use as much data as possible to aid our ability to understand what we're seeing on the ice? Perhaps it could show us the specific areas to improve on the team going forward. No? You say that it's entirely useless and we should focus on our gut instincts only to improve the team. Sounds like a poor idea to me. That's how we end up with the idea that this team needs more truculence.

What this team needs is more skill, all the way through the lineup. Do that and a lot of the problems go away. Give just about any coach enough skill and they will win their share of games, and then hopefully we can stop worrying if the coach is good enough.

[nik-]

Originally Posted by Cali Panthers Fan

Ok, seriously. These kind of comments need to end around here.

Why?

There was a pretty noticeable turn in some of the posts from the team around the time the Arena fiasco was really heating up. I thought the timing was curious, and it seems to have continued into excessive defense of the coaching staff well beyond what would seem normal.

The owners of this site don't owe anything to anyone here except to keep the site running (and even that they don't). And if they want it to be profitable, then they should take income from whoever they want. They're not journalists, they've taken no oath to impartiality. It's just curious.

As far as stats, how many times does it need to be pointed out that the results are disconnected from the stats that keep getting parroted? It's borderline zealotry at this point.