Lies, Damned Lies, and Statistics

[Mike F]

I'll say off the top that I've never been a big follower of Analytics (i.e. "Advanced Statistics") other that having it pounded in all last year that the Analytics said Calgary's success was unsustainable. However, I dove into War on Ice the last 24 hours because what I was viewing in the recent games wasn't corresponding to the results on the ice (terrible losses) or the takes of many posters on this forum (that the players have stopped working and the team has sucked royally, deserving the losses).

If I've understood the numbers correctly (which is a big if, I'll admit), they show that in the past 5 games the Flames have played quite well despite the scoreboard results. In fact, by some statistical measures, it's been their best stretch in the last year +. Specifically...

The Flames have had a >50% CF% in 4 of their last 5 games. All of last season, they only managed that once - between Jan 17 and Jan 31 they had a >50% CF% in 5 of 6 games (yes, that's technically 2 5 game segments overlapping, but that would obscure the point).

They're also doing a much, much better job when it comes to balancing scoring chances for vs. against. Over the last 5 games they are +26 in SCF/SCA (they were +14 vs. Montreal, +12 vs. Ottawa, -3 vs. the Islanders, -4 vs. the Rangers, and +7 vs. Detroit). That blows away the two best 5 game segments last season - Dec 09 to 15, where they were +2, and January 17 to 31 (actually a 6 game segment (I was stretching to be generous)) where they were +13. In fact, their last 2 games, where they were +12 and +14, represent two of the team's three most dominant games by that measure in this regular season and the last, with only the December 9th game vs Toronto seeing them match this season's high of +14.

What to take from this? Either:

A. Analytics aren't actually useful;
B. The Flames are actually playing well, but are being undermined by terrible goaltending and snake-bitten shooters; or
C. I've misunderstood the stats on War on Ice and have wasted everyone's time.

[BurningSteel]

Im more confused than when I came in here. Were just not very good this year for unknown reasons.

BUT a new month starts tomorrow

[transplant99]

Originally Posted by Mike F

I'll say off the top that I've never been a big follower of Analytics (i.e. "Advanced Statistics") other that having it pounded in all last year that the Analytics said Calgary's success was unsustainable. However, I dove into War on Ice the last 24 hours because what I was viewing in the recent games wasn't corresponding to the results on the ice (terrible losses) or the takes of many posters on this forum (that the players have stopped working and the team has sucked royally, deserving the losses).

If I've understood the numbers correctly (which is a big if, I'll admit), they show that in the past 5 games the Flames have played quite well despite the scoreboard results. In fact, by some statistical measures, it's been their best stretch in the last year +. Specifically...

The Flames have had a >50% CF% in 4 of their last 5 games. All of last season, they only managed that once - between Jan 17 and Jan 31 they had a >50% CF% in 5 of 6 games (yes, that's technically 2 5 game segments overlapping, but that would obscure the point).

They're also doing a much, much better job when it comes to balancing scoring chances for vs. against. Over the last 5 games they are +26 in SCF/SCA (they were +14 vs. Montreal, +12 vs. Ottawa, -3 vs. the Islanders, -4 vs. the Rangers, and +7 vs. Detroit). That blows away the two best 5 game segments last season - Dec 09 to 15, where they were +2, and January 17 to 31 (actually a 6 game segment (I was stretching to be generous)) where they were +13. In fact, their last 2 games, where they were +12 and +14, represent two of the team's three most dominant games by that measure in this regular season and the last, with only the December 9th game vs Toronto seeing them match this season's high of +14.

What to take from this? Either:

A. Analytics aren't actually useful;
B. The Flames are actually playing well, but are being undermined by terrible goaltending and snake-bitten shooters; or
C. I've misunderstood the stats on War on Ice and have wasted everyone's time.

Ding ding ding!!

[GirlySports]

statistics are like mini-skirts.......

[GranteedEV]

Originally Posted by Mike F

B. The Flames are actually playing well, but are being undermined by terrible goaltending and snake-bitten shooters;

Most certainly.

That said there's one Flame who hasn't been positive in the analytics department over this stretch... Russell. He's been in the last 5 games:

49.09 (0) (Russell-Hamilton on the comeback attempt.. A New Hope)
----------
41.46 (-2) (Russell-Engelland) (The Possession Menace)
41.18 (-3) (Russell-Engelland) (The Corsi-Clone Wars)
--------------
39.39 (-1) (The Return of the Wideman-Russell)
36.36 (-3) (Wideman-Russell Strike Back)

[cofias]

B. The Flames have been playing pretty good hockey of late but can't convert on their chances and have had abysmal goaltending coupled with poor puck luck. It happens sometimes. Unfortunately it happened after a stretch of poor play to start the season.

Some players still aren't contributing (perhaps most notably Backlund, Russell, Wideman, the goaltenders), but as a whole, the team is playing well enough to win games at this point. The wins will come. But it may be too late to salvage the season.

[Manhattanboy]

Originally Posted by GirlySports

statistics are like mini-skirts.......

I don't get it....

[Badgers Nose]

Originally Posted by GirlySports

statistics are like mini-skirts.......

Don't ruin miniskirts for me.

[formulate]

I would say most people watching the games would say these statistics pass the eye test. Specifically the Ottawa and Montreal games where it seemed the Flames controlled the play for the most part, and generated many legitimate scoring chances.

The problem has been when the puck does enter their own zone, it's defensive break down after defensive break down (led by Russell and Wideman, who I've defended in the past but who have both been a train-wreck defensively this year) and seemingly every puck ends up behind our goalie.

Call it a nice combination of poor shooting/nerves, being snakebitten (luck), truly awful defending at times (and not only defensemen are guilty), truly horrific goaltending, and some more poor luck just for fun.

[ripTDR]

My big take on this is, last year the flames were excellent at the big comeback later in the game. Dangerous for sure, but they did it. Made for excellent and entertaining hockey, but ultimately not so good in the long term. You can't always rely on the comeback.

[trublmaker]

Originally Posted by Manhattanboy

I don't get it....

They're nice to look at but that's about it

[getbak]

Originally Posted by transplant99

Ding ding ding!!

I disagree. Analytics aren't useless. The problem is that "Advanced Stats" aren't particularly advanced, nor are they really analytics.

Originally Posted by GirlySports

statistics are like mini-skirts.......

Something something drunks. Something something lampposts.

[Drak]

Originally Posted by Mike F

I'll say off the top that I've never been a big follower of Analytics (i.e. "Advanced Statistics") other that having it pounded in all last year that the Analytics said Calgary's success was unsustainable. However, I dove into War on Ice the last 24 hours because what I was viewing in the recent games wasn't corresponding to the results on the ice (terrible losses) or the takes of many posters on this forum (that the players have stopped working and the team has sucked royally, deserving the losses).

If I've understood the numbers correctly (which is a big if, I'll admit), they show that in the past 5 games the Flames have played quite well despite the scoreboard results. In fact, by some statistical measures, it's been their best stretch in the last year +. Specifically...

The Flames have had a >50% CF% in 4 of their last 5 games. All of last season, they only managed that once - between Jan 17 and Jan 31 they had a >50% CF% in 5 of 6 games (yes, that's technically 2 5 game segments overlapping, but that would obscure the point).

They're also doing a much, much better job when it comes to balancing scoring chances for vs. against. Over the last 5 games they are +26 in SCF/SCA (they were +14 vs. Montreal, +12 vs. Ottawa, -3 vs. the Islanders, -4 vs. the Rangers, and +7 vs. Detroit). That blows away the two best 5 game segments last season - Dec 09 to 15, where they were +2, and January 17 to 31 (actually a 6 game segment (I was stretching to be generous)) where they were +13. In fact, their last 2 games, where they were +12 and +14, represent two of the team's three most dominant games by that measure in this regular season and the last, with only the December 9th game vs Toronto seeing them match this season's high of +14.

What to take from this? Either:

A. Analytics aren't actually useful;
B. The Flames are actually playing well, but are being undermined by terrible goaltending and snake-bitten shooters; or
C. I've misunderstood the stats on War on Ice and have wasted everyone's time.

Ding ding ding!!!!















Completely joking of course

[GirlySports]

Originally Posted by trublmaker

They're nice to look at but that's about it

yup

statistics are like miniskirts, what they show is revealing but what the hide is critical.

[calumniate]

Puck luck has returned to the statistical mean :-\

[sharkov]

It tears my *** apart to see that the same stats people who we proved wrong last year getting the last laugh at the end of this season if the flames continue playing the way they have

[Vulcan]

Originally Posted by sharkov

It tears my *** apart to see that the same stats people who we proved wrong last year getting the last laugh at the end of this season if the flames continue playing the way they have

I don't buy that. Last season we had bad stats with good results. This season we have decent stats with bad results. It's the opposite of what the stats guys are saying.

[opendoor]

Originally Posted by Vulcan

I don't buy that. Last season we had bad stats with good results. This season we have decent stats with bad results. It's the opposite of what the stats guys are saying.

Only if you completely ignore the effect that trailing as much as the Flames have has on possession stats. The only thing propping up the Flames' numbers are the fact that they've trailed for 3 times as long as they've led so far this season. Trailing teams virtually always generate more shot attempts than leading ones, and that's a big reason why the Flames' overall numbers aren't too bad. But, when the game is tied the Flames have the worst possession numbers in the entire league.

If you adjust for score effects to account for that disparity, they're currently 27th in the NHL in Score Adjusted Corsi which is right in line with their 27th best points %.

[mdubz]

Cuz goaltending.

[EldrickOnIce]

Originally Posted by opendoor

Only if you completely ignore the effect that trailing as much as the Flames have has on possession stats. The only thing propping up the Flames' numbers are the fact that they've trailed for 3 times as long as they've led so far this season. Trailing teams virtually always generate more shot attempts than leading ones, and that's a big reason why the Flames' overall numbers aren't too bad. But, when the game is tied the Flames have the worst possession numbers in the entire league.

If you adjust for score effects to account for that disparity, they're currently 27th in the NHL in Score Adjusted Corsi which is right in line with their 27th best points %.

He was talking last 5 games.
Last 5 games, they have been ok score adjusted, probably 50% and not much good to show for it.