NHL analysis

Well, the last game (Calgary @ San Jose) is just wrapping up, and it looks like a clear victory for SJ (5-1 with 2:31 to play in the third. I’m just going to assume that game finishes as a regular victory for San Jose and move on.

Last night was solid for the predictions. There were 12 games in total and I got 8 right (66.6%). That leaves me at 21-for-33, or 63.6%. I’ve still been very impressed with how high that number is considering the very small sample of data I have to go on. I’m going to track my percentages from day one of the regular season but I’m also going to “throw out” the month of October and use only games from November onward for my “real” total. That’ll give every team about 12-14 games played, under real game circumstances, and should give a solid start for the predictions to work with. Of course as the season progresses I should in theory only get more accurate. In theory.

Busy night last night followed by a quiet Sunday. Here are the games:

Away—Home—Prediction
——————————
CAR–@–WAS—Washington
DET–@–TOR—-Detroit

Well, last night was a tough one, and I only went 2-for-6. Heck, that’s more in line with what I’ve been expecting through the pre-season anyways. That’s 13-for-21 overall (62%). Here’s tonight’s games:

Home—-Away—-Prediction
——————————-
NYR——BOS—–Rangers
OTT——MTL——Senators
TOR——DET——Red Wings
PIT——-BUF——Sabres*
NJ——–NYI——Islanders*
ATL——NAS——Predators
CAL——SJ——–Calgary
ANA—–PHO——Coyotes
CHI——MIN——Blackhawks
LA——-COL——Kings
TB——-DAL——Lightning
EDM—–VAN——Oilers

Busy night tomorrow. The Sabres and Islanders wins are statistically too close to call.

Well, a bit of a weaker night last night, going 3-for-6. That puts me at 11-for-15 overall, or 73%. Considering I was originally planning on waiting at least a month into the regular season to start publishing my predictions, I’m pretty happy with these results. I expected to do alot worse early on.

Here’s tonight’s predictions:

Away — Home — Prediction
——————————-
NYR @ NYI — New York Rangers
PHI @ WAS — Philadelphia
CBJ @ CAR — Carolina*
BUF @ PIT — Buffalo
NAS @ ATL — Nashville*
LA @ SJ — Los Angeles
CAL @ VAN — Calgary

The predictions of Carolina and Nashville are statistically either even or so close they may as well be even. Statistically the biggest winners should be the Rangers, Kings and Flames.

Well, I sure was pleasantly surprised last night: 8 for 9 of my predictions were right!  Hehe, I have to say I really think that was a fluke (there seriously haven’t been enough games for my algorithms to be accurate) but it just goes to show you that stats don’t lie.  I only missed San Jose beating Calgary.  I did catch the “upsets” (based on points in the preseason) of Rangers (6 points) over Boston (8) and Florida (1) over Atlanta (4).  Home teams did almost exactly as expected, as well, winning 5 of 9 games (56%).

So I should probably just stop with the predictions now and go out with an overall 89% accuracy rating, eh?

Naaaaah.

Here are tonight’s predictions:

Away — Home — Prediction
——————————-
Chi    @   Fla  — Chicago
Det @ TB — Detroit
Mtl @ Ott — Montreal
Phi @ NJ — New Jersey
StL @ Min — Minnesota
Edm @ Pho — Edmonton

The two close games should be Montreal-Ottawa and St. Louis-Minnesota. The Montreal-Ottawa game is actually very close, as Montreal has a very poor record on the road (and has had an easy road schedule) while Ottawa has been equally bad at home (with an even easier schedule). Chicago has been just hurtling along in the preseason, as has Edmonton. Phoenix has done alright but has had a very difficult schedule.

Well, although there really haven’t been enough games played to offer useful results, and even though the changing rosters and plethora of minor leaguers in pre-season games make predictions both meaningles and nearly impossible, I’ve decided to post my game predictions for tonight. There are two reasons: first and foremost, to test that I’m ready for “real” hockey, which is only a short week away; and second, just because I feel like it ;). Hopefully this will also give you a taste for what’s to come.

I will point out that the power rankings, which are used to determine these predictions, are still quite variable because of the tiny number of games played (some teams have played only 3 games). There, I’ve covered my butt for when I go 0-for-9 tonight (haha, hopefully not).

Away - Home - Prediction
———————————————–
BOS @ NYR — NYR
MIN @ DET — DET
OTT @ TOR — OTT
ATL @ FLA — FLA
CHI @ STL — CHI
COL @ DAL — COL
SJ @ CAL —- CAL
VAN @ EDM — EDM
PHO @ ANA — ANA

The weakest prediction (statistically) is New York over the Bruins. All other games (statistically) be convincing victories.

A short discussion here: the reason the statistics offer so much confidence right now is that I’ve differentiated between home and away records. The prediction method is simple: I compare the home team’s home power ranking to the away team’s away power ranking. How big a difference there is decides if the confidence is high. Tonight, the only moderately close power rankings that are faced off are Boston and New York. The problem of course is that even teams that have played 5 or 6 games have only played 2 or 3 games each of at home and on the road. So as far as the stats are concerned, these teams have played literally only a handful of games.

Anyways, I was looking at these numbers anyways, just to mostly test out that everything is working as expected, so I pass them on to you as well.

Bill Guerin, what can I say? Four goals last night, one assist. I wondered why Dallas would buy out his last year. Did they not realize that they are paying him $2.25 million to score goals for someone else?

Players in situations like Guerin are such wonderul short term additions. He has nowhere to go from up and has oh so much to prove to himself, his team and the fans. No question he had a dreadful season last year, but I would say there’s also no question he has at least a few more seasons left in his tank.

For any who don’t know, buyouts work like this: the team pays two thirds of the players salary, all in the one season (it would be over X number of seasons if the player had X years left on his contract), which counts against the cap.  If only one year remains, they can also choose to split the buyout over 2 seasons, which is why Guerin’s buyout is $2.25 for this season and next season.  To put this into perspective of the Dallas team this year, Dallas currently has about $1.9 million in space under the cap after Guerin’s buyout. They are paying Jere Lehtinen $2.77 million. They pay Philippe Boucher (their number 2 blueliner) $2.5 million. Yeah, the buyout is that big.

To give the flip side of the story, I’ll give Dallas GM Doug Armstrong the benefit of the doubt here. He must have tried shopping around Guerin and couldn’t have had much luck. He obviously would have been happy even with a late round draft pick, just so he wouldn’t have to pay that buyout.

Buyouts are an interesting new feature in the CBA and in the salary cap era NHL. It’s turning out they are one clause, at least, that really favours the players. Even if Guerin had chosen to retire, he’d still be making more than $2 million this year. And why? Because he stunk so badly last year that Dallas prefers to pay him not to play.

You know, come to think of it, I could easily have a worse NHL season than Guerin did last year. Maybe I can get Dallas to pay ME $2 million to not play for them!

Well, every team has played at least one game so far (some as many as four). The Edmonon Oilers, Chicago Blackhawks, New York Rangers, Minnesota Wild and Colorado Avalanche all are undefeated. Home teams have racked up 23 regulation wins compared to 12 on the road, but road teams have done well in OT, winning 4 or six games that went into overtime, as well as the only shootout. Home teams have outscored road teams 132-106. Phoenix has had the most difficult schedule so far, facing Dallas (2-2) twice and Edmonton (3-0-1) twice. Note that schedule difficulty is very innacurate, since there have been so few games played so far.

Leafs’ fans everywhere are paying more attention to the pre-season victory over Ottawa than the 7 losses last year. I (a Leafs’ fan myself) am looking forward to October 4th to see what the boys can do in a real game situation against their Ontario rivals.

Injuries are piling up but luckily noting really serious yet - Malkin dislocated his shoulder and is questionable for the opener but should be back soon; Jagr is also questionable for the opener but likewise is not expected to miss much time. Fedorov of course is out for a month or more. And plenty of very minor injuries are adding up - Chara in Boston, Kaberle and Kubina in Toronto, Alfredsson in Ottawa, the list goes on. These injuries are usually not even injuries - the teams simply don’t want their players risking further injury by playing on a sore ankle or knee during pre-season play.

Team stats are ready to go and working nicely through the pre-season testing; can’t wait to try them out for real in another few weeks. Individual player stats are likewise progressing nicely and hopefully I’ll have more than expected by the time the season gets underway.

So the Columbus Blue Jackets, plagued by injuries for the last few seasons, got more bad news today: Sergei Fedorov has injured his shoulder and is expected to be out for 4-6 weeks recovering.

This is bad news for a team that lacks scoring depth up front. Fedorov had an off year last year, managing “only” 44 points in 67 games, but he was still the 4th highest scorer on the team (and Nikolai Zherdev, the man above him, appears likely to be a hold out this year). More, much more, was expected of Fedorov when he was acquired last season. Injuries and a slow start combined to keep him from having the impact his $6 million salary suggest he ought to have.

From a salary cap perspective, the only good news is that the Blue Jackets are off the hook for paying Fedorov while he remains on the injured reserve list. If he misses 6 weeks he would be out until about early November, missing about 8-10 games. 8 games would free up about $580,000 of actual cap space. Not a whole lot, especially considering that the Blue Jackets are already well below this year’s $44 million cap. So the good news isn’t really all that good after all.

Looks like the fans in Columbus may have another year of disapointments ahead of them. The other additions, notably Fredrik Modin, should help to offset Fedorov’s loss, but the team was really hoping to field the full roster they’ve put together this year. They will not be able to replace Federov, even when Fedorov isn’t fully on his game.