2006/07 Statistical Analysis: Toughest Divisions
[Promoted from the Diaries - Joe]
This is not a post about the Colorado Avalanche, rather a statistical analysis on the strength of each NHL division in the 2006-07 season. As I don’t have a hockey blog, I post it here.
I thought: "How do you know if a team has benefited from playing in division X or Y." And then I thought "Well, you could do that if you follow a premise". That is, if each team plays 82 games, 32 within its division, 40 against the other divisions in its conference and 10 games against teams in the other conference, in an "ideal" (let’s call it that) world, each team would do 32/82 of their points in divisional games, 40/82 in conference games and 10/82 in interconference games...
Division games: around 39% of the points. That means a team with 100 points should have 39 from divisional games.
Conference games: around 48.8 % of the points. That means a team with 100 points should have 49 from conference games.
Interconference games: around 12.2% of the points. That means a team with 100 points should have 12 from interconference games.
These values will be referred to as expected.
I thought: "If a team does more points than the "expected" within its division and then when plays against other divisions does bad, that might mean that the team’s division is weak.
We, Avs fan, generally regard the Northwest division as one of the toughest in the NHL. That would mean in my analysis that the Northwest division teams usually have less than the expected in divisional games and more in games against other divisions. There’s a team in the Northwest division who made only a third of its points (33%) in divisional games, 14.6% less than the expected. But made a lot more than you would expect in the other games if you look at its point total: 3.5% more in Western Conference games and 32.8% in interconferece games. That’s Vancouver.
So, by compiling the data for each team, I was hoping I could rank them in order of percentage of points higher or lower than the expected in divisional, conference and interconference games. The fact that a team has a high percentage of its points in divisional games means it made too many points in its division and that division is too weak for them. On the other hand, if a team has less points that the expected in their division and then has a lot of points in games against other divisions, it means that the division is tough: the team is able to make more points against teams in other divisions that in their own.
So, let’s go to my conclusions...
First, division totals:
Division / % Division / % Conf / %Other Conf
NW / -4.2 / +0.1 / +13.2
C / +1.8 / -2.1 / +2.5
P / -1.7 / -1.7 / +12.5
A / +2.7 / -1.6 / -2.5
NE / -5.1 / +4.0 / +0.1
SE / +7.5 / -1.3 / -18.8
So, first quick conclusions:
- The Western Conference is stronger than the Eastern Conference. While every division in the Western Conference made more points than expected in games against East, only Northeast can say the same in the Eastern Conference and not by much: 0.1%. Even teams like Edmonton (+27%), Pheonix (+46.9%) and Los Angeles (+32.6%) did very good (proportionally) against the Eastern Conference.
- Southeast can thank Bettman or whoever for the current division alignment: they made 7.5% more of their points in games between themselves than expected.
- Atlantic Division was the second weakest division: You might find it strange: They put 4 teams in the playoffs last season. Yes, but that’s only because of an abnormally bad Flyers. If the Flyers were just regular bad, you could be sure that either Toronto or Montreal would have been in the playoffs instead of the Islanders.
- Central Division... Oh, Central Division... The easy road for Nashville and Detroit. Both did great in their division, 13.4% and 4.8% better than expected.
- Pacific Division starts the tour for the toughest three divisions.
- Northwest division was not the toughest as we Avs fan though, Northeast was. Both divisions made more points than expected in games against other divisions than in divisional games.
So, my ranking of the toughest divisions to play in:
1- Northeast (East)
2- Northwest (West)
3- Pacific (West)
4- Central (West)
5- Atlantic (East)
6- Southeast (East)
Now, for teams that benefited or not from playing their division:
TEAM %DIV %CONF %O CONF
1 Dallas 17,3% -8,0% -23,4%
2 New Jer 17,3% -0,4% -54,0%
3 Carolin 16,5% -9,1% -16,1%
4 Atlanta 13,6% 3,6% -57,7%
5 Detroit 13,4% -9,3% -5,7%
6 Tampa B 10,2% -7,4% -3,0%
7 NY Isla 5,8% -4,2% -2,0%
8 Colorad 5,2% -9,4% 20,8%
9 Pittsbu 4,9% -4,3% 1,5%
10 Nashvil 4,8% -6,8% 11,8%
11 Chicago 4,7% -1,8% -7,6%
12 Minneso 3,5% -1,4% -5,4%
13 Washing 2,5% -9,2% 28,9%
14 Montrea 2,5% -8,9% 27,6%
15 Anaheim 0,2% -1,2% 4,4%
16 Ottawa 0,1% 7,4% -29,7%
17 Calgary -1,2% 2,5% -6,0%
18 Boston -2,2% -5,6% 29,5%
19 NY Rang -4,6% 2,5% 4,7%
20 Florida -7,6% 14,4% -33,3%
21 Columbu -8,7% 9,5% -10,1%
22 Los Ang -9,6% -0,5% 32,6%
23 San Jos -11,4% 3,5% 22,6%
24 St. Lou -11,4% 3,8% 21,5%
25 Buffalo -11,6% 10,7% -5,7%
26 Phoenix -12,0% -2,1% 46,9%
27 Toronto -12,7% 12,6% -9,9%
28 Vancouv -14,6% 3,5% 32,8%
29 Edmonto -17,0% 6,8% 27,0%
30 Philade -22,2% -1,2% 75,7%
So, conclusions about this:
- Colorado can not blame not making the playoffs on the tough division they play in, as the Avs made 5.2% more than expected in games against Northwest teams.
- Kind of weird Dallas had made 49 of its 107 points (48.8%, 17.3% higher than the expected 39%), specially because San Jose and Anaheim did not do so great. Maybe that’s why Dallas is so high. Not weird to find New Jersey there, as the Flyers gave them a little help.
- 3 Southeast teams in the top 6 of the teams that most benefited from playing in their division. No comments...
- NY Islanders did great in their division, which might explain why they made the playoffs when it was probably fairer to see Toronto or Montreal there.
- Anaheim and Ottawa, both Stanley Cup finalists are in the middle of the table and can neither thank or curse their luck as they did the expected percentage of points in games in their division.
- Vancouver and Edmonton in 28th and 29th: the top and bottom team of the Northwest division did a lot less than the expected in divisional games and it was on games against others they made the most points.
- Philadelphia were just bad, but did great against the Western Conference: You’d expect to see a 56 point team to do around 7 points against the other conference, but they got 12 and had a positive record (6-4-0).
And if we give each team the points corresponding to their position in this table, we could see that the strength of divisions is equal to the one posted above, where NE beats NW.
Well, take your conclusions.
Notes:
1- This probably has errors and I can not deduct things this way, but I’m no stats expert, and I’m sorry if it has any errors.
2- There is more to it than what is here, I just did not make a thorough analysis.
3- I’ll gladly explain anything that is not clear from what has been written, I know my English is far from being perfect.
MileHighHockey.com is a fan community, allowing members to post their own thoughts and opinions on the Colorado Avalanche and hockey in general. These views and thoughts may not be shared by the editors of MileHighHockey.com.
4 comments
|
0 recs |
Do you like this story?
Comments
Stats
Luis, this is an excellent diary entry! Considering that English is not your first language, you do an excellent job of getting your point across.
And the stats are fascinating. Good work!
Don't read too much into these stats...
Hey Luis
Very interesting analysis. Unfortunately, I don't think we can read too much into these statistics.
For example, we CANNOT say that the Avs can't blame being in the NW division for not making the playoffs. The relevant question is not "how well did the Avs do in their division games" (which is essentially what this stat tells us - we would expect an "average" team to have a 0 in this statistic). Rather, the relevant question is "What would this percentage have been if the Avalanche were in a different (read: easier) division?" It is clear to me that the Avalanche would've had more points at the end of the season had they been able to play Columbus, Chicago and St Louis 8 games a year. I'm not complaining - although I hate the NHL scheduling now and think Gary Bettman is doing his best to sink the NHL into obscurity.
Still, very interesting stuff here...thanks for the effort! Just be careful what conclusions you try and draw from these statistics.
by tdo on Oct 7, 2007 11:28 AM MDT reply actions
Re: Don't read too much...
tdo, good points as well.
I think we must also be careful with any kind of statistical assumptions. Just as Luis' analysis has weaknesses, asking questions like, "how would they have done in another division?" has all kinds of weaknesses as well. Any kind of hypothetical question is, at its root, impossible to prove or disprove.
But it's still fun.
reply
Yes, you're right. Of course the Avs would have more points if they were playing 8 times vs Columbus and Chicago but what I mean was that it wasn't on divisional games they underperformed: it was on West Conference games they should have been doing better. Actually, although Colorado finished 4th in the division, they were the second best in Northwest in divisional games, 3 OTL behind Minnesota. When you do that good in a division so tough, you have to blame yourself for not going as well against easier divisions...
And of course, when we analyzing statistics, we say what we interpret and see in it and it may not exactly be the truth or what other people see. Thanks for the feedback.

by 


















