HighSchoolHockeyForums.com

The Girls' Forum has had a fascinating discussion going for a couple of days, in which a poster has attempted to scale players' stats using an strength of schedule metric derived from KRACH. For those of you who may not venture over there often and have any sort of interest in this, I highly recommend it:

http://www.ushsho.com/forums/viewtopic.php?t=30271

It would be great if someone could try this with the boys as well.

I believe this is fundamentally flawed for several reasons. The main one is that with so many non-conference games in boys hockey, the strength of schedule is largely a by-product of how talented a team/program is.. For example, a team like Sibley which is in the same conference as Hill Murray has a dramatically different schedule. Where Hill plays 10 top 15 teams and enters tournaments like the Schwann's Cup Gold, Sibley schedules weaker opponents, as they should, to make for more competitive games. So the Strength of Schedule is correlated with team talent to a certain degree.

Thus a player on a team with a strong strength of schedule like Hill or STA, Edina etc is also more likely to be surrounded by higher caliber players which aid his scoring opportunities. Same for Goalies of these teams who tend to have stronger defenders in front of them then a team such as Como Park etc (don't want to just pick on Sibley.) So, really, there is already a coefficient that adjusts for this factor somewhat.. Though still room for more..

Just wanted to point out that in dynamic systems, which a hockey league is, there are a multitude of interdependencies and cross-correlational effects. That it's next to impossible to separate out one factor as if we have isolated and purified it with a single adjustment like strength of schedule.. Granted when BSM puts up between 15-21 goals in 5 games (last year) it begs for some normalization of the stats.. But it's hard to pick out individual games for adjustments and broad ones are subject to complexities and interdependencies that are extremely hard to untangle

However, I applaud the effort. I posted a while back about the need for more robust statistics for hockey and there was some lively discussion - which is great as long as we are aware of the biases and limitations that can distort the results..

-

test

Why change anything? We all know there is way more to a player than simply their stats. The best goalies in the state year in and year out are not the best statistically, same goes for the forwards/defenders, or at least with the most common statistics recorded.

Most years the top point scorers are on teams from smaller schools with a low SOS or a good player on a good team with few other scorers. There's no telling for sure what either player would do in a different situation by the numbers alone.

Would Besse have more goals playing for Morris-Benson?
Would Picht have more points playing for BSM?
Neither of these questions can be answered by stats alone.

Any new look statistically is always interesting to look at, but this seems like it'd be nothing more than a guess that's hard to know what it means for sure.

Rank Team SOS
1 St. Thomas Academy 0.666
2 Edina 0.650
3 Moorhead 0.643
4 Hill-Murray 0.638
5 Benilde 0.627
6 Minnetonka 0.615
7 Duluth Marshall 0.614
8 Holy Family 0.610
9 Wayzata 0.605
10 Burnsville 0.597
11 Duluth East 0.594
12 Grand Rapids 0.591
13 Rochester Lourdes 0.591
14 Breck 0.588
15 St. Cloud Cathedral 0.587
16 Bemidji 0.584
17 Hibbing 0.582
18 Thief River Falls 0.578
19 South St. Paul 0.578
20 Eden Prairie 0.574
21 Eagan 0.568
22 Virginia 0.568
23 Cloquet 0.566
24 Roseau 0.562
25 Hermantown 0.561
26 Lakeville South 0.561
27 East Grand Forks 0.560
28 Warroad 0.559
29 White Bear Lake 0.558
30 Mahtomedi 0.550
31 Jefferson 0.549
32 North St. Paul 0.548
33 Totino-Grace 0.546
34 Alexandria 0.545
35 Providence Academy 0.542
36 Blake 0.542
37 Orono 0.540
38 Mound-Westonka 0.540
39 Little Falls 0.538
40 Mounds View 0.536
41 Duluth Denfeld 0.535
42 Hopkins 0.535
43 Delano 0.534
44 Sauk Rapids 0.533
45 Maple Grove 0.532
46 Forest Lake 0.531
47 Lakeville North 0.530
48 Elk River 0.529
49 St. Cloud Tech 0.527
50 Apple Valley 0.527
51 Prior Lake 0.524
52 Centennial 0.523
53 Blaine 0.522
54 St. Francis 0.522
55 Crookston 0.522
56 New Prague 0.521
57 Holy Angels 0.520
58 Brainerd 0.519
59 Cretin-Derham Hall 0.518
60 Hastings 0.518
61 Princeton 0.515
62 Rogers 0.515
63 Red Wing 0.514
64 Fergus Falls 0.513
65 St. Cloud Apollo 0.511
66 Sartell 0.511
67 Eastview 0.510
68 River Lakes 0.508
69 Stillwater 0.507
70 Tartan 0.506
71 Rosemount 0.506
72 East Ridge 0.504
73 Andover 0.503
74 Cambridge 0.497
75 Monticello 0.495
76 Park of Cottage Grove 0.493
77 Rochester Century 0.492
78 Chanhassen 0.491
79 Osseo 0.491
80 International Falls 0.490
81 Roseville 0.490
82 Lake of the Woods 0.489
83 Woodbury 0.489
84 Proctor 0.486
85 Hutchinson 0.486
86 Kennedy 0.483
87 Albert Lea 0.483
88 Anoka 0.483
89 Detroit Lakes 0.482
90 Willmar 0.480
91 Winona 0.480
92 Litchfield 0.478
93 Rochester John Marshall 0.477
94 Spring Lake Park 0.476
95 Mankato West 0.475
96 Park Rapids 0.474
97 Buffalo 0.474
98 Northfield 0.473
99 St. Louis Park 0.473
100 Irondale 0.472
101 Sibley 0.471
102 St. Michael-Albertville 0.469
103 New Ulm 0.465
104 Windom 0.463
105 Achiever NE Academy 0.462
106 Chisago Lakes 0.461
107 Chaska 0.461
108 Champlin Park 0.461
109 Simley 0.461
110 Rochester Mayo 0.460
111 Marshall 0.460
112 Minnehaha Academy 0.458
113 Coon Rapids 0.457
114 St. Paul Academy 0.454
115 Farmington 0.453
116 Shakopee 0.453
117 Owatonna 0.449
118 Northern Lakes 0.449
119 Richfield 0.449
120 Kittson Central 0.448
121 Worthington 0.446
122 Red Lake Falls 0.445
123 Waconia 0.444
124 Armstrong 0.444
125 Breckenridge 0.440
126 Dodge County 0.438
127 Morris-Benson 0.437
128 Faribault 0.436
129 Austin 0.434
130 Luverne 0.431
131 Becker-Big Lake 0.429
132 Redwood Valley 0.429
133 St. Paul Como Park 0.425
134 LeSueur-St. Peter 0.425
135 Fairmont 0.425
136 Minneapolis 0.421
137 St. Paul Johnson 0.418
138 Cooper 0.417
139 Wadena 0.416
140 Prairie Centre 0.415
141 Eveleth 0.407
142 North Branch 0.407
143 Mankato East 0.404
144 Moose Lake 0.402
145 North Shore 0.395
146 Greenway 0.395
147 Ely 0.394
148 Mora 0.391
149 Pine City 0.389
150 Waseca 0.377
151 St. Paul Highland Park 0.373
152 Legacy Christian 0.371
153 Bagley 0.357
154 LaCrescent 0.350

Through Saturday. Does not calculate in out of state games or Shattuck.

On the surface this seems like a good idea but after thinking the idea through, I don't think it would be a an accurate reflection of the numbers. For instance, a top points player on his/her teams 1st line plays a weak SOS team, so presumably any points scored would be diminished. But that doesn't take into account that "top points player' is matched with the weak SOS teams top line the whole game.

This is the stupidest article ever posted.

Strength of schedule is a biased opinion of how one side of the town or state is better at hockey than the other. The Metro believe's they are better than outstate...the east side believes their 8th place team could beat the first place team from the lake (cough, 2-1, cough)...etc...etc.. it's all a bunch of crap and personal vantage opinion, of course that is my opinion on an opinion while offering an opinion from this opinion.