What I Learned Today: Prediction

Tuesday, 15 April 2014

How Wrong Was I - My Prediction vs the Final NHL Standings

Back in October I predicted the standings for the 2013-2014 NHL season and I was making a larger point that pundits ('puckheads') blather on as the so called experts. These 'puckheads' and I'm on of them to a lesser extent make predictions and rarely ever go back and admit they are wrong. Of course if they get it right they thump there chest (ie Don Cherry) or give some chin swagger (Bob McKenzie.)

I'm going to actually compare my and THN (The Hockey News) result to actual and see how wrong I was:

Eastern Conference:

THN's Prediction:	My Prediction:	Actual Result
1) Pittsburgh Penguins	1) Pittsburgh Penguins	1) Boston Bruins
2) Boston Bruins	2) Boston Bruins	2) Pittsburgh Penguins
3) Ottawa Senators	3) Detroit Red Wings	3) Tampa Bay Lightning
4) Detroit Red Wings	4) Toronto Maples Leafs	4) Montreal Canadians
5) New York Rangers	5) Ottawa Senators	5) New York Rangers
6) Toronto Maple Leafs	6) Montreal Canadians	6) Philadelphia Flyers
7) Montreal Canadians	7) Washington Capitals	7) Columbus Blue Jackets
8) Philadelphia Flyers	8) New York Islanders	8) Detroit Red Wings

9) Washington Capitals	9) Columbus Blue Jackets	9) Washington Capitals
10) New York Islanders	10) New York Rangers	10) New Jersey Devils
11) Columbus Blue Jackets	11) Philadelphia Flyers	11) Ottawa Senators
12) Tampa Bay Lightning	12) Tampa Bay Lightning	12) Toronto Maple Leafs
13) Carolina Hurricanes	13) Carolina Hurricanes	13) Carolina Hurricanes
14) Buffalo Sabres	14) New Jersey Devils	14) NY Islanders
15) New Jersey Devils	15) Florida Panthers	15) Florida Panthers
16) Florida Panthers	16) Buffalo Sabres	16) Buffalo Sabres

From a mathematical perspective:

Odds of predicting the top 8 in precise order is 1 in 518,918,400
Odds of predicting the 8 teams that will make the play-offs regardless of order is 1 in 12,870

Well both THN and I were both completely off in predicting the order of the standings. As for playoff teams I got only 4 of 8 teams right and THN got 6 of 8 right. Both THN and I thought that the Toronto Maple Leafs and Ottawa Senators would make the play-offs. Sadly, this is the worst season in NHL history for Canadian hockey teams.

Western Conference:

THN's Prediction:	My Prediction:	Actual Result
1) Chicago Blackhawks	1) Chicago Blackhawks	1) Anaheim Ducks
2) Los Angeles Kings	2) San Jose Sharks	2) Colorado Avalanche
3) San Jose Sharks	3) Los Angeles Kings	3) St. Louis Blues
4) St. Louis Blues	4) St. Louis Blues	4) San Jose Sharks
5) Minnesota Wild	5) Anaheim Ducks	5) Chicago Blackhawks
6) Vancouver Canucks	6) Vancouver Canucks	6) Los Angeles Kings
7) Dallas Stars	7) Phoenix Coyotes	7) Minnesota Wild
8) Edmonton Oilers	8) Minnesota Wild	8) Dallas Stars

9) Anaheim Ducks	9) Nashville Predators	9) Phoenix Coyotes
10) Nashville Predators	10) Colorado Avalanche	10) Nashville Predators
11) Winnipeg Jets	11) Winnipeg Jets	11) Winnipeg Jets
12) Colorado Avalanche	12) Edmonton Oilers	12) Vancouver Canucks
13) Phoenix Coyotes	13) Dallas Stars	13) Calgary Flames
14) Calgary Flames	14) Calgary Flames	14) Edmonton Oilers

From a mathematical perspective (now you have 14 teams):

Odds of predicting the top 8 in precise order is 1 in 121,080,960
Odds of predicting the 8 teams that will make the play-offs regardless of order is 1 in 3,003

Again, THN and I were way off in predicting the order of the standings. As for playoff teams both THN and I were right for 6 out of 8. Although, the odds are greatly improved with 2 fewer teams in the Western conference.

My Point made Simple:

I was wrong and the odds of me or the puckheads getting it right or close to right is remote. The folly of prediction, but with this I at least went back to see I was accurate. I scanned the local sports pages and I did not see one of these puckheads check their accuracy. I did see a lot of articles with puckheads analyzing the final standings and what went right and wrong for teams. I could sit here and make convenient rationalizations after the fact and talk about player injuries, bad coaching, and players having abnormal seasons. But, it still makes my prediction wrong.

One More Prediction to Go - Stanley Cup Champion:

This my and THN's prediction from back in October. Based on how Boston has played this season, I like my chances. St. Louis Blues are also playing extremely well, but will have to battle out of the West and will be playing Chicago Blackhawks in the first round. That could possbly be one of the best series this play-offs considering the talent of both teams and how they match-up.

Stanley Cup Champion:

My prediction:

Boston Bruins - (Runner-up St. Louis Blues)

THN's prediction:

St. Louis Blues - (Runner-up Pittsburgh)

Comments? Thoughts? Suggestions? Which team do you think is going drink from Lord Stanley's Cup?

Tuesday, 7 January 2014

Oracle of North York almost predicts Canada's Men Olympic Hockey Team

Earlier this year I wrote an article predicting the outcomes of the NHL season and discussed the folly of prediction. My point was that making predictions is pretty useless, but we all enjoy making predictions. It makes for good TV (or radio.) However, key to all predictions is going back and comparing your result to actual. Were you right? Were you even close? Are you better than the average. Well I bothered to compare my prediction and the prediction of professional hockey analysts to the actual result.

Picture from Hockey Canada Website

Back on Thursday, December 12, I wrote a blog with my prediction for the Men's 2014 Olympic Hockey Roster. Highlighted in yellow are the players I wrongly predicted.

My Prediction:

Forwards:	Left Winger	Center	Right Winger

Line 1:	Chris Kunitz	Sydney Crosby	Steve Stamkos

Line 2:	Patrick Sharp	Jonathan Toews	Logan Couture

Line 3:	Rick Nash	Ryan Getzlaf	Corey Perry

Line 4:	Martin St. Louis	John Tavares	Patrice Bergeron

Defencemen:	Left Side	Right Side

Pairing 1:	Duncan Keith	Brent Seabrook

Pairing 2:	Shea Weber	Drew Doughty

Pairing 3:	Jay Bouwmeester	Alex Pietranagelo

Goaltenders:	Roberto Luongo	Spare:	Jamie Benn (F)
	Carey Price	Matt Duchene	PK Subban (D)
	Mike Smith		Kris Letang (D)

I picked a 23 man roster. At the time that was accurate, but they subsequently announced that the rosters will be 25 players. So, I was one defencemen and one forward short. Looking at my notes (chicken scratch on a now crumpled ruled paper) I predicted Kris Letang would be the 8th defencemen and Matt Duchene as next after Jamie Benn for forwards.

Actual Result:

Forwards:	Left Winger	Center	Right Winger

Line 1:	Chris Kunitz	Sydney Crosby	Steve Stamkos

Line 2:	Patrick Sharp	Jonathan Toews	Patrick Marleau

Line 3:	Rick Nash	Ryan Getzlaf	Corey Perry

Line 4:	Jeff Carter	John Tavares	Patrice Bergeron

Defencemen:	Left Side	Right Side

Pairing 1:	Duncan Keith	PK Subban

Pairing 2:	Shea Weber	Drew Doughty

Pairing 3:	Jay Bouwmeester	Alex Pietranagelo

Goaltenders:	Roberto Luongo	Spare:	Jamie Benn (F)
	Carey Price	Matt Duchene (F)	Dan Hamhuis (D)
	Mike Smith		M.E. Vlasic (D)

My accuracy as compared to the Professionals:

Kieran - Forwards (12/14), Defencemen (6/8) and Goalies (3/3) - Overall (21/25 or 84% accuracy)

The Globe and Mail Writers:

Eric Duhatshek - Forwards (10/14), Defencemen (6/8) and Goalies (3/3) - Overall (19/25 or 76% accuracy)

James Mirtle - Forwards (10/14), Defencemen( 6/8) and Goalies (2/3) - Overall (19/25 or 76% accuracy)

Sean Gordon - Forwards (10/14), Defencemen (6/8) and Goalies (2/3) - Overall (19/25 or 76% accuracy)

The Hockey News:

Rory Boylen (THN Editor) - Forwards (9/14), Defencemen (5/8) and Goalies (2/3) - Overall (16/25 or 64% accuracy)

The Toronto Star:

Damian Cox - Forwards (10/14), Defencemen (7/8) and Goalies (3/3) - Overall (20/25 or 64% accuracy)

Sportsnet:

Nick Kypreos - Forwards (11/14), Defencemen (6/8) and Goalies (3/3) - Overall (20/25 or 80% accuracy)

Doug Maclean - Forwards (12/14), Defencemen (6/8) and Goalies (3/3 Overall or 84% accuracy)

The Sports News (TSN):

Darren Dreger - Forwards (11/14), Defencemen (7/8) and Goalies (3/3 or 80% accuracy)

In Review:

Well I beat the pundits on average by two correct picks. It seemes like the majority of pundits made the same mistakes. On forwards the majority of pundits predicted Logan Couture (including me) who did not make the team. He is also going for hand surgery tomorrow (Jan 8, 2014) and it is unknown when he'll return per the Globe and Mail. Steve Yzerman and company likely knew this and had him out of the the running. Few people predicted Jeff Carter to picked as well as to be a forward. On defence everyone predicted that Brent Seabrook would be selected because of his strong play this season and him being Duncan Keith's defensive partner. Dan Hamhuis was not on anyone's final roster from the lists I looked at as well. Otherwise it was small group of players on the bubble. What I learned really is that everyone pretty much predicted 75% of the team and it was the same players. What is more telling is how many of these pundits will publicly compare how accurate they actually were. Will they own up to being wrong? Furthermore, almost all of them hedged there bets in their original article. Giving them some wiggle room.

Closing Comments:

In conclusion... To quote Ralph Wiggum: "I beat the smart kids, I beat the smart kids!" Well me and Doug Maclean... I gave myself a pat on the back. Now lets see if Steve Yzerman and his team put together the right guys to get the job done in Sochi. Also props to the Globe and Mail for having the only article published clearly showing their prediction as compared to actual.