Goal prediction and how I manage to come to the Canuck placing anywhere between 2-6 in conference

[Outsiders]

Left Wingers

D.Sedin = 19

Higgins =13

Kenins=6

Prust =4

Centers

H.Sedin = 16

Sutter = 22

Horvat = 20

Vey = 12

Right Wingers

Vrbata = 28

Burrows =17

Hansen =16

Virtanen = 13

Dorsett = 7

[Crabcakes]

Assuming this is the line up. Make your Goal predictions.

Right wingers

G 1014/15

Vrbata 30 31

Burrows 22 18

Hansen 15 16

Dorsett 9 7

Virtanen 14 21 (Cal Jr) If he is ready for NHL (a whole lot more if he gets to play with Sedins)

Left wingers

D.Sedin 20 20

Higgins 17 12

Prust 9 4 (Mtl)

Baertschi 14 9 (7 Ut)

Kenins 7 4

Center

H.Sedin 18 18

Sutter 26 21 (Pitt)

Horvat 20 13

Vey 10 10

Defence

Edler 12 8

Tanev 5 2

Hamuis 8 1

Weber 12 11

Sbisa 8 3

Bartkowski 3 0

Corrado 3 8 (7 Ut)

282 237

Tell me what you think. If you disagree, please tell me why; but most importantly post your predictions.

If you agree with this it means total of 282 goals or 3.44 goals per game.

Lets put this in perspective now: If my prediction is 36 goals less that would mean 246 goals = 3 goals per game. If you disagree with me please tell me by how much and if you can get that below 246 goals or less then 3 goals per game. If we can manage anywhere near 3 goals a game, we should not only manage to get into the playoffs but I would argue place any where from 2-6 in our conference.

If this is unreasonable please explain why.

For you grammar police the title should be Canucks.

How I get to my predictions: Average career goals by season after factoring out injury seasons ,worst and best season, and factor in last season totals at .666 if player wasn't injured or last season gets factored in at .500 if player was injured. So D.Sedins goal total is 20 last season at .666 and his career season total at .333 minus injury season 09-10 and minus his lowest and highest goal totals. The .666 counters for decline in his play and .333 a counts for career #s his range or margin of error is plus or minus 3.

If you want to know where the .666 and the .333 comes from take advance probability courses in math. Correct 19 out of 20 times with a margin of error of + -3.

Well, a couple of things since the Canucks best goals per game over the last 10 years was 3.27 in 2009-10 and the Canucks aren't that team this year.

1. Trends - players age and develop, perhaps weighting more recent years would yield better results. It's a team sport and line mates make a difference. There have been lots of changes.
2. Reasonableness Test - I think that you'll get less variance, error and a better forecast of future performance if you try to work from the top down by using team totals and work towards individual performances rather than from the bottom up and try to build a total from individual performances.
3. Goal differential is probably a better measure of success and overall team play anyways

GF GPG GA Diff Pts

05-06 252 3.07 251 +1 92

06-07 217 2.65 197 +20 105

07-08 207 2.52 206 +1 88

08-09 243 2.96 213 +40 100

09-10 268 3.27 218 +50 103

10-11 258 3.15 180 +78 117 Cup run, Presidents Trophy

11-12 241 2.94 191 +50 111 Presidents Trophy

12-13 122 2.54 115 +7 59 Lockout year

13-14 191 2.33 216 -25 83 Torts year

14-15 236 2.88 220 +16 101

The Torts year can be thrown out all together.

Since there have been substantial changes to the team since the Presidents Trophy years, I would suggest that only 12-13 and 14-15 could be used to reasonably predict next years goal totals. The average of those 2 years is 2.71 ppg +/- 5% would give us a range of 2.57-2.85 gpg or 211 to 234. So the reasonableness test starts here.

Let's go from there and try to build some individual stats.

Lets just say everybody scores the same as they did last year whether they were in junior or not, that's 237 total and close enough to a reasonable total.