Wild Card Race (Western Conference)

[Oling_Roachinen]

Originally Posted by polak

I'll bet $50 that at least one of LA and Winnipeg will finish ahead of Calgary.

Any takers?

But, based on the chart, you certainly shouldn't be expecting to get anywhere close to even money on that bet. What are the odds you are giving?

[polak]

Originally Posted by EldrickOnIce

One doesn't fly.
According to those odds, both should.
That would have to be your bet, imo.

Nope. I'll only do it if one's enough. If WPG and LA finish below Calgary, I owe $50.

I don't care about the Corsi chart. I just saw an opportunity for a bet.

[heep223]

Originally Posted by Street Pharmacist

Yes, but not of you exclude "score effects". It's well established that teams trailing by more than one goal really rack up the shot attempts in order to get closer and leading teams sit back and prevent quality chances

The chart that CorsiHockeyLeague posts uses Fenwick Close which is unblocked shot attempts while the game is within one goal. There won't be much data for that game included because Calgary was up so early. Maybe 5-10 minutes of data. Interestingly, Fenwick Close had stronger predictor power than corsi

Originally Posted by CorsiHockeyLeague

Correct. The chart shows that Toronto came out strong in the first 5 minutes, followed by a flurry from the Flames leading to a goal. The Flames then continued to press, but shortly after there is a bunch of PP time, and by the time that sequence was done the game was effectively over.

Everything after is what we refer to as "garbage time" - i.e., a bunch of game time during which the data isn't particularly useful because the score is so lopsided, which impacts how teams play in a significant way.

That makes sense. So when I see for example the TSN power rankings, and they list the team's Corsi% and reference it (ie. The Flames have a Corsi of 43%), wouldn't that be kind of misleading then, based on what you're describing above?

[Street Pharmacist]

Originally Posted by heep223

That makes sense. So when I see for example the TSN power rankings, and they list the team's Corsi% and reference it (ie. The Flames have a Corsi of 43%), wouldn't that be kind of misleading then, based on what you're describing above?

They use corsi 5v5, but flames are 4th last instead of 3rd last with Fenwick close 5v5 of 44.6. Doesn't really alter things much

[heep223]

Originally Posted by Street Pharmacist

They use corsi 5v5, but flames are 4th last instead of 3rd last with Fenwick close 5v5 of 44.6. Doesn't really alter things much

Sorry I was pointing to the fact that you and CHL referred to the Corsi 5v5 in the Flames/Leafs game the other night as being useless because the game wasn't close.

But don't the team Corsi 5v5 stats incorporate this (useless) Corsi number? Ie. wouldn't the Flames team Corsi that TSN references in their power rankings include that game. I feel like I'm misunderstanding something, because the number that critics point to and saying "unsustainable" include stats that you're saying are garbage?

[CorsiHockeyLeague]

Originally Posted by heep223

That makes sense. So when I see for example the TSN power rankings, and they list the team's Corsi% and reference it (ie. The Flames have a Corsi of 43%), wouldn't that be kind of misleading then, based on what you're describing above?

Depends. Straight up 5v5 corsi (if that's what TSN is listing) is still a good predictor of future success, partly because the score effects wash out to some extent over time. In one game, the Flames might be up 3 goals for 35 minutes, in another, they're down 2 for 40 minutes, etc. For what it's worth, Calgary's overall FF% is 45.6 and CF% is 44.4 (3340 minutes). Their corsi close is 44.5% and fen close is 45.6% (2033 minutes). Almost identical whether you use the score close stats or not.

So by this point in the season, it's almost leveled out. For other teams there is a more significant gap - it's usually not huge, though.

The issue is sample size. Using corsi close or fenwick close when you've only got 20 games of data is less worthwhile than overall corsi, because it'll be far fewer total minutes of data. Close stats become more useful when you've got the full season (or at least half a season) of data. But as you can see, in the long run it usually ends up leveling out quite a bit anyway.

That being said, when you look at the corsi chart for an individual game you can pretty easily explain what happens looking at where the lines go. There, it's pretty clear that once Calgary was up a few goals they took their foot off the pedal, which a lot of teams do (and this usually lines up with the "eye test").

[Street Pharmacist]

Originally Posted by heep223

Sorry I was pointing to the fact that you and CHL referred to the Corsi 5v5 in the Flames/Leafs game the other night as being useless because the game wasn't close.

But don't the team Corsi 5v5 stats incorporate this (useless) Corsi number? Ie. wouldn't the Flames team Corsi that TSN references in their power rankings include that game. I feel like I'm misunderstanding something, because the number that critics point to and saying "unsustainable" include stats that you're saying are garbage?

Yes for this game it didn't do a good job explaining who was better because the game was out of hand early and score effects changed things. Wouldn't you say even visibly the Leafs weren't getting their arises handed to them after Calgary was up 4 zip? Calgary dominated early on and coasted to victory.

That doesn't explain the season as a whole though. It's like weather vs climate. The world may be getting hotter, but we still have cold snaps.

Calgary has a 44.3% Corsi% 5v5. When the score is close and we remove blocked shots they're still only marginally better at 44.6%. So yes in this game the Corsi 5v5 didn't tell the story, but on the season it did.

In this particular game Calgary had a corsi% 5v5 of 43.4%, but a Fenwick% Close of 85.5%

Clearly Fenwick close described more accurately who was the better team, but in the season the b numbers aren't that disparate

[Mattman]

Originally Posted by polak

I'll bet $50 that at least one of LA and Winnipeg will finish ahead of Calgary.

Any takers?

Eat
a) a broadcasters pubes
b) your pubes

Then we have a deal

[polak]

Originally Posted by Mattman

Eat
a) a broadcasters pubes
b) your pubes

Then we have a deal

I'd like to keep my mouth free of pubes.

[Igster]

Originally Posted by polak

I'd like to keep my mouth free of pubes.

Wouldn't we all.

[hwy19man]

After Monday's games:

1. Minnesota (38-23-7) 35 ROW, 83 pts
Vancouver (39-25-4) 35 ROW, 82 pts--second in the Pacific Division
Calgary (38-26-5) 34 ROW, 81 pts--third in Pacific Division
2. Los Angeles (34-22-13) 32 ROW, 81 pts--and fourth in Pacific Division
3. Winnipeg (34-23-12) 28 ROW, 80 pts
4. San Jose (34-27-8) 31 ROW, 76 pts--and fifth in the Pacific Division
5. Colorado (32-26-11) 23 ROW, 75 pts
6. Dallas (32-28-10) 29 ROW, 74 pts

Kings won in regulation

[Tsawwassen]

Originally Posted by normtwofinger

I liked this thread when it was about the playoff race.

I don't know how much more I can fasten my seatbelt before I squish my guts out. This is nuts, where 3 points separates 5 teams and it's just bizarre that the Wild could still finish 9th. I want to know what voodoo that Dubnyk does or whatever he does. I still think the Canucks could be the odd team out. Tuesday is going to be must see TV.

[polak]

Originally Posted by Tsawwassen

I don't know how much more I can fasten my seatbelt before I squish my guts out. This is nuts, where 3 points separates 5 teams and it's just bizarre that the Wild could still finish 9th. I want to know what voodoo that Dubnyk does or whatever he does. I still think the Canucks could be the odd team out. Tuesday is going to be must see TV.

Again, this is why I wouldn't get my hopes up that any set number of points is enough.

I would not be surprised if this was a record setting season.

[Robbob]

The problem is everyone plays each other so if one team wins it takes away points another team can get. For the teams in the mix (VAN, CGY, MIN, LA, WPG, SJ, COL, DAL) this is how many games they play one of those teams:
VAN: 6
CGY: 6
MIN: 3
LA: 6
WPG: 6
SJ:4
COL: 5
DAL: 4

[fire_starter]

So what's our chances at now?

[hwy19man]

After Tuesday's games:

1. Minnesota (39-23-7) 36 ROW, 85 pts
Vancouver (40-25-4) 36 ROW, 84 pts--second in the Pacific Division
Los Angeles (34-22-13) 32 ROW, 81 pts--third in the Pacific Division
2. Winnipeg (35-23-12) 29 ROW, 82 pts
3. Calgary (38-27-5) 34 ROW, 81 pts--and fourth in the Pacific Division
4. San Jose (34-28-8) 31 ROW, 76 pts--and fifth in the Pacific Division
5. Colorado (32-26-11) 23 ROW, 75 pts
6. Dallas (32-28-10) 29 ROW, 74 pts

Jets and Canucks won in regulation
Sharks and Flames lost in regulation
Wild won in overtime

[Manhattanboy]

Sigh.

[hockey.modern]

Calgary will still make it. The Blues game was probably our last hard game to play for a while. We can get domes points while L.A has to go through their tough road trip for the rest of the month.

[Darkwater]

Believe.

[Itse]

Originally Posted by fire_starter

So what's our chances at now?

78% says playoffstatus.com
73.6% says sportclubstats.com
69.9% says hockey-reference.com

Honestly I'd say the lower end is probably more realistic. Not surprising our chances took a big hit after two losses and essentially everyone else winning everything.

Still more likely to make it than not. But time to start talking about "must win games" now.