Advanced Stats Update

[Bingo]

December 17th

Incredibly, by Corsi stats the Flames have just played 4 of their best 5 games this season in this 6 game losing streak.

Last night against the Rangers is pretty much proof positive that the eye test overrides the advanced stats.

Just like the Flames were not getting out played in games that had them post .45 Corsi, they were not the better team last night despite a .60 Corsi

Corsi
1 BUF Dec 11
2 NYR Dec 16
3 TOR Dec 09
4 NSH Oct 14
5 S.J Dec 06

33 CHI Oct 15
32 EDM Oct 09
31 CHI Dec 14
30 NSH Oct 31
29 ARI Dec 02

Fenwick
1 BUF Dec 11
2 MTL Oct 28
3 OTT Nov 15
4 CAR Oct 23
5 NYR Dec 16

33 CHI Oct 15
32 EDM Oct 09
31 CHI Dec 14
30 CHI Nov 20
29 T.B Nov 06

The Flames have posted an unsustainable .938 PDO in the 6 game losing streak

PDO
1 CAR Oct 23
2 ANA Nov 18
3 EDM Oct 09
4 ARI Nov 13
5 ARI Dec 02

33 BUF Dec 11
32 ANA Nov 25
31 CAR Nov 10
30 N.J Nov 22
29 MTL Oct 28

[Poe969]

Bring back basic stats!

[Tinordi]

Bingo are you using raw corsi or score adjusted corsi?

[shermanator]

I propose a new CP drinking game.

Take 1 drink if you see the word corsi, fenwick, or PDO
Finish your drink if you see the word RGI in a post

[Dienasty]

I thought they were by far the better team last night.

One stupid play by Russell, a crossbar by Granlund which went the other way and resulted in another goal. They easily could have won that game, and probably should have. The difference is Nash is a pure goal scorer and the Flames are lacking one.

[Tinordi]

You didn't need to look at any advanced stats to know that things were not exactly as they seemed during the Flames hot start. All you needed to look at was their league's best shooting percentage. It was simply too high based on the talent of the roster with zero natural goal scorers. We all thought this team was going to have trouble scoring goals. They bucked expectations with some great luck but now that luck has run out. If this team is to make any noise it will have to grind out the wins.

[Bingo]

Originally Posted by Tinordi

You didn't need to look at any advanced stats to know that things were not exactly as they seemed during the Flames hot start. All you needed to look at was their league's best shooting percentage. It was simply too high based on the talent of the roster with zero natural goal scorers. We all thought this team was going to have trouble scoring goals. They bucked expectations with some great luck but now that luck has run out. If this team is to make any noise it will have to grind out the wins.

I think things have gone to the other side of the map now though. People were suggesting the Oilers couldn't stay at .965 because they were unlucky, the Flames have been sub .94 in the 6 game losing streak.

[IamNotKenKing]

Originally Posted by Bingo

I think things have gone to the other side of the map now though. People were suggesting the Oilers couldn't stay at .965 because they were unlucky, the Flames have been sub .94 in the 6 game losing streak.

Oh, but that's because they have no skill...

[heep223]

Originally Posted by Dienasty

I thought they were by far the better team last night.

One stupid play by Russell, a crossbar by Granlund which went the other way and resulted in another goal. They easily could have won that game, and probably should have. The difference is Nash is a pure goal scorer and the Flames are lacking one.

And there was Colborne's attempt going through Lundqvist's legs, hitting his pad and then hitting the post and out....and that one that Klein scraped off the goalline and off the post and out. It's just how things are going right now.

[Red Menace]

I don't know much about these advanced stats, but I am pretty sure they are useless in as small a sample size as 1 game....you have to look at the whole season...and we are probably now pretty close to where we should be...

[Winsor_Pilates]

There you have it; we can now just stop talking about stupid advanced stats.

[ernie]

I'm sure it's been said over and over again but advanced stats aren't the be all and end all. Nor should they be ignored. They do show general trends that better "advanced stats" teams tend to be, well, the better teams when all is said and done. Not always. there are always exceptions and outliers etc. But if one were to build a team you are going to want to be a better advanced team than a lesser one.

Sample size is also important. It may not mean a whole heck of a lot over a game or handful of games. Over 82 regular season games and the playoffs, however, those trends certainly exist. For last night, I'd wager a team that is down 3-0 2 minutes into the second period may come out looking pretty good on Corsi because the opposition is going to pull back and protect the lead.

First and foremost the advanced stats are a piece of information that can cause you to look at things make you ask why? The stats don't look good but a team is winning over a pretty significant stretch of games...well why? For example, the advanced stats crowd weren't saying the Leafs would come back down to earth last year because the advanced stats were bad. It was because the advanced stats were bad AND after asking why it was thought the goaltending and shooting percentage were unsustainable. Far Far Far above career averages. Now if it was a situation where the goaltending was simply outstanding but they had Vezina performing Hasek between the pipes you might say it is sustainable because you have the proven guy between the pipes that more than compensates for those possession troubles.

[Bingo]

Funny to see the Hockey News suggesting that Leaf Corsi is a result of perhaps skill and not unsustainable shooting percentage.

http://www.thehockeynews.com/blog/ma...ced-stats-how/

[Street Pharmacist]

Originally Posted by Bingo

Funny to see the Hockey News suggesting that Leaf Corsi is a result of perhaps skill and not unsustainable shooting percentage.

http://www.thehockeynews.com/blog/ma...ced-stats-how/

Except that the "we keep them to the outside" is hogwash when you look at where the shots actually come from:



Sure they get more of their own shots from in tight, but their goalies face 30% more shots than league average from in tight

[ricardodw]

Originally Posted by Street Pharmacist

Except that the "we keep them to the outside" is hogwash when you look at where the shots actually come from:



Sure they get more of their own shots from in tight, but their goalies face 30% more shots than league average from in tight

When you these stats to try prove a point you need to go back to the base stats and see how well they reflect what actually happens on the ice,

On the last Game winning goal by the Flames Monahan was listed in NHL.com play-by-play as scoring on an 11 ft wrist shot when the replay shows it being a 3 footer from just off the crease.

When the official stats are that wrong on a game winning goal you have to think that they might be out by more than a couple feet on a nondescript SOG.

I took a lot of flack about the variability and lack of consistency of Hits stats.... I think that Hits stats would be at least as reliable as SOG and save stats.

[Bingo]

Flames Best and Worst lists for Corsi, Fenwick and PDO update ...

Corsi
1 BUF Dec 11
2 NYR Dec 16
3 TOR Dec 09
4 NSH Oct 14
5 S.J Dec 06

43 CHI Oct 15
42 DAL Dec 19
41 EDM Oct 09
40 VAN Jan 10
39 CHI Dec 14

Ortio game stealer makes the worst five possession games this season list. There is little argument about the better team on that night, but scoring 5 minutes in would have the Canucks pushing for 55 minutes.

Fenwick
1 BUF Dec 11
2 OTT Nov 15
3 MTL Oct 28
4 CAR Oct 23
5 NYR Dec 16

43 CHI Oct 15
42 VAN Jan 10
41 L.A Dec 29
40 FLA Jan 09
39 EDM Oct 09

Fenwick worst has Vancouver and Florida games. Flames not playing well.

PDO
1 CAR Oct 23
2 ANA Nov 18
3 EDM Oct 09
4 EDM Dec 27
5 ARI Nov 13

Oilers are no good.

43 BUF Dec 11
42 ANA Nov 25
41 CAR Nov 10
40 N.J Nov 22
39 MTL Oct 28

No top 5 worst games posted since December 11th.

[the_only_turek_fan]

Where do you find these stats?

[darthma]

I just don't understand why these are considered 'advanced stats'. They are just different proxies that people have come up with, much like the +/- number, and suffer the same problems as +/-. Every time I see 'advanced stats', I expect some deeper analysis, like a linearized model or principal components analysis.

[Bingo]

Originally Posted by the_only_turek_fan

Where do you find these stats?

the core data is from War on Ice, but the tables and ranking I've built custom.

[Bingo]

Update again ...

the past three games haven't registered at all in good or bad lists

Games 46
Corsi
1 BUF Dec 11
2 NYR Dec 16
3 TOR Dec 09
4 NSH Oct 14
5 S.J Dec 06

46 CHI Oct 15
45 DAL Dec 19
44 EDM Oct 09
43 VAN Jan 10
42 CHI Dec 14

Fenwick
1 BUF Dec 11
2 OTT Nov 15
3 MTL Oct 28
4 CAR Oct 23
5 NYR Dec 16

46 CHI Oct 15
45 VAN Jan 10
44 L.A Dec 29
43 FLA Jan 09
42 EDM Oct 09

PDO
1 CAR Oct 23
2 ANA Nov 18
3 EDM Oct 09
4 EDM Dec 27
5 ARI Nov 13

46 BUF Dec 11
45 ANA Nov 25
44 CAR Nov 10
43 N.J Nov 22
42 MTL Oct 28