Thrashers 2nd Least Injured Team

James Mirtle has the final end of season man games lost to injury up on his blog. As I mentioned recently in a post the Thrashers were very fortunate during the last two seasons that they were rather healthy. According to Mirtle's ranking the Thrashers were the 2nd most healthy team in the NHL. Were are all those people who wanted to fire the trainer after Kari's groin problems?

Playoff Picks

OK I'll slip on my fortuneteller hat and look into the playoff future.
Note: (winning percentage against other playoffs teams during the regular seasons.)

Eastern First Round

Montreal (.537) v. Boston (.463) Montreal easily dominated the season serious. I think the Habs will make a long run. To me Montreal and the Rangers have the strongest all around clubs in the Eastern Conference once you consider talent and experience combined. Pick: Montreal

Pittsburgh (.478) v. Ottawa (.476) I'm sorely tempted to pick Ottawa. Both clubs had weak winning percentage against other playoff teams. Ottawa has a ton of talent and the Canadian press--which was ready to award them the Cup after a hot October--have ripped them apart with a relish usually saved for the Leafs. If Ottawa wins the first game the young Penguins squad could panic under the weight of expectations. However, it looks like Alfredsson and Fischer will both miss games and that tips the balance back to the Pittsburgh. Pick: Pittsburgh

Washington (.561) v. Philadelphia (.446). The Flyers sort of stumbled into the playoffs and Capitals are red hot. The Flyers must win the 1st game to have any shot in my opinion, the confidence of the Capitals is sky high and they will keep rolling if they can nail down that first playoff win in several years. Pick: Washington

NJ Devils (.500) v. NY Rangers (.543). The Devils have home ice, but the Rangers dominated the season series winning 7 of 8 and have the better record against playoff qualifiers. Jagr sort of mailed it in most of the season but perked up during the last quarter of the season and if he gets going they are a more potent team. Pick: Rangers

Western First Round
Detroit (.694) v. Nashville (.389). Easily the biggest mismatch on paper. If the Wings stay healthy they win the whole thing. Pick: Detroit.

San Jose (.500) v Calgary (.476). There are people who say the Flames are constructed more for the post-season, I agree with that. The Sharks just torched weak teams during the regular season and the joy ride ends in the first round against a tough opponent. Pick: Calgary.

Minnesota (.439) v. Colorado (.427). I don't like either club that much and the Avalanche had a poor record against good teams. On the other hand, Colorado somehow managed to muddle through injuries that would have caused many other clubs to crater. Colorado is experienced--or old--depending on how you look at it. This one is tough for me to call. I'm giving the edge to Colorado for their gutty regular season. Pick: Colorado

Anaheim (.558) v. Dallas (.535). Looks pretty even on based on the regular season but when you add in Selanne, Nidermayer to Anaheim and subtract banged up Zubov it tips strongly toward the Ducks. Pick: Anaheim

Stanley Cup Finals Matchup:
Red Wings v Montreal. Pick Wings.

Separating Playoff Contenders from Pretenders

The NHL playoffs are upon us and it is time for many of us to make our picks to win the Stanley Cup or draft a playoff pool team. The playoff games are much tougher since only the top half of the NHL qualifies and scoring usually drops significantly as these higher quality teams face off against one another.

Who should you favor when making your picks? Most people will take a look at the final regular season numbers. Some will put lean towards teams with a good power play, others favor those with strong goaltending. One thing that I find to be very revealing is to look at games where two playoff teams were matched up against one another. In other words, how did each club fare when facing a quality opponent.

With that question in mind I uploaded the complete regular season results for every singly NHL team and then recalculated some key statistics into playoff and non-playoff categories. One notable exception is that I deleted the extra point for a OT loss or a SO win because those things don't exist in the playoffs. I also didn't count the "goal" credited to the SO winner in the regular season. So what we have is the old school "winning percentage" based on OT Wins and Losses and if the game wasn't settled at the end of OT it counts as a tie. So .500 is the break even point like in the old NHL.

The table below lists the results for all 16 playoff teams. What does the data tell us? Don't underestimate Washington. They played extremely well against other playoff clubs--in fact the Caps had a higher winning percentage (.561) against playoff teams than non-playoff teams. There were only four teams that put up a better percentage versus non-playoff teams: Detroit, Anaheim, Washington and the Rangers.

Beware of teams that racked up big regular season point totals against weak teams because there are no basement dwellers in the post-season. Who were the worst offenders? The Sharks absolutely destroyed non-playoff teams to the tune of .705 win % in the regular season, hmmm. The bottom half of that table show you a bunch of teams larded up on weak opposition but struggled against the other playoff clubs. Three of the teams that have home ice advantage have warning flags: Minnesota, San Jose and Pittsburgh. The Penguins have the advantage of facing off against Ottawa which also struggled against quality teams but the Pens might not be a post-season juggernaut. Likewise, Minnesota had a less than sterling record against quality teams but Colorado also struggled greatly versus good teams.

Team	Playoff Team Win %	Non-Playoff Win %	Playoff Team Goal Differential	Non-Playoff Team Goal Differential
DET	.694	.630	+41	+31
WAS	.561	.500	+13	-2
ANA	.558	.528	+4	-1
NYR	.543	.486	+12	+3
MON	.537	.622	+13	+28
DAL	.535	.538	+16	+17
NJD	.500	.583	-11	+16
SJS	.500	.705	-6	+35
PIT	.478	.653	+4	+24
CGY	.476	.549	-9	+11
OTT	.476	.598	-8	+24
BOS	.463	.549	-16	+7
PHI	.446	.639	-11	+29
MIN	.439	.695	-12	+22
COL	.427	.598	-20	+28
NAS	.389	.609	-10	+13

How about those special teams? Which PP and PK units excelled in games versus other playoff teams and which struggled? Again, Washington really stands out here their PP% against playoff teams was an outstanding 24% and their PP unit struggled against the weaker teams at just 16%. Other teams where the PP unit was better against tougher opposition include Pittsburgh, Ottawa and Minnesota. Teams where the PP tended to vanish against better clubs include Boston, Calgary, Dallas, and Philadelphia. Don't count on the a ton of PP goals from the Bruins and the Flyers in the post-season as both clubs saw a huge drop off.

How about penalty killing? In general the gap appears to be smaller than on PP side of things. It would appear that good PK units are good almost all the time. This is one area where the Nashville Predators excel. Most of the stats presented here offer cold comfort to Predators fans but Nashville along with Dallas, Detroit and Minnesota all had the top PK% versus other playoff teams (all in the West I might point out, the 3 top PP units were all in the East).

Team	Playoff PP%	Non-Playoff PP%	Playoff PK%	Non-Playoff PK%
ANA	17.5	18.1	83.9	82.5
BOS	13.5	22.3	77.6	79.6
CGY	14.1	19.4	80.0	83.0
COL	12.7	16.6	79.3	83.4
DAL	15.8	20.6	88.0	82.2
DET	20.1	21.3	86.2	82.3
MIN	20.9	17.2	84.2	86.0
MON	24.1	24.3	80.9	84.1
NAS	14.8	14.8	86.3	84.4
NJD	15.3	16.1	81.3	85.1
NYR	16.3	16.7	83.5	85.9
OTT	19.8	16.9	81.9	80.2
PHI	17.9	27.6	81.7	85.4
PIT	21.3	19.2	78.1	85.2
SJS	18.3	19.3	83.5	88.4
WAS	24.2	15.8	77.3	82.9

Now let's take a look at defensive and offensive efficiency. First we have team save percentage broken down into playoff and non-playoff opposition. Some clubs almost always play their #2 goalie against weaker clubs so we might see some big differences on this one as the #1 netminder will start almost every game in the playoffs. Here are the teams that actually had a better SV% when matched up against other playoff teams: Anaheim, Calgary, Colorado, Dallas, Detroit, and NY Rangers. Teams that saw a sharp decline in SV% against better opposition: Boston, Minnesota, Montreal, NJ Devils, Philadelphia, and San Jose.

Finally, how about offensive efficiency? Which teams finished off their scoring chances when they had the opportunity? I'm using team shooting percentage--which is I'll admit is an imperfect measure but it is all that I have available for the moment. Most teams are going to convert fewer scoring chances when matched up against better teams, a few beat this trend: Anaheim, Dallas, Detroit, Montreal, Pittsburgh and Washington. Teams that suffered a big drop off: Boston, Colorado, Nashville, Ottawa, and Philadelphia.

Team	Playoff SV%	Non-Playoff SV%	Playoff ST%	Non-Playoff ST%
ANA	91.9	91.7	9.0	8.9
BOS	90.9	91.8	8.0	9.6
CGY	90.7	90.1	9.4	10.1
COL	91.2	89.3	7.6	11.2
DAL	90.6	90.3	11.0	10.6
DET	91.7	90.0	9.2	8.7
MIN	90.4	92.6	9.9	9.9
MON	90.5	92.8	11.5	10.1
NAS	90.6	91.0	8.6	10.0
NJD	90.8	92.2	8.2	8.6
NYR	91.5	90.5	7.9	7.9
OTT	89.9	90.4	9.4	11.5
PHI	90.6	92.2	9.6	11.4
PIT	90.4	93.0	11.3	9.8
SJS	89.1	92.2	8.6	9.3
WAS	89.8	90.1	10.7	8.5

Conclusion: Which teams are pretenders hiding behind regular season records puffed up from munching on bottom dwelling clubs. Based on the data above I'd say don't count on Boston, Philadelphia, Minnesota, Ottawa and Colorado and Nashville going deep into the playoffs. But there are some cautionary flags even for favorites like San Jose and Pittsburgh. Which are the playoff contenders? Well Anaheim looks rock solid especially when you consider that Selanne and S. Nidermayer were not even in the lineup for many of these games. Detroit, Dallas, Montreal and the Rangers also appear to be very solid against other quality clubs. But more than anything the data suggests that Washington could be a big time competitor in the post-season despite not qualifying until the last weekend. They have played some great hockey against top teams.

One last question, which non-playoff team might have made things interesting if they had qualified for the post-season? The table below shows the same winning percentage breakdown for the 14 non-playoff clubs. Vancouver, Buffalo and Carolina just missed the post-season so it is not terribly surprising to see them them at the top. Chicago was a surprise to me. The fact they played well against other playoff clubs bodes well for the Blackhawks future, if they can improve versus non-playoff clubs they could take the next step forward. On the other hand, Edmonton and Columbus both did very well against weak clubs and they will need to win more "hard" points against tough teams to climb up the standings next year.

Team	Playoff Win %	Non-Playoff Win %	Playoff Goal Differential	Non-Playoff Goal Differential
VAN	.490	.500	+1	-5
CHI	.489	.474	+5	-2
CAR	.463	.598	-15	+19
BUF	.459	.576	-2	+20
TOR	.439	.455	-23	-5
PHX	.412	.565	-20	+4
FLA	.390	.549	-24	+15
NYI	.389	.464	-12	+22
ATL	.378	.415	-29	-20
STL	.375	.461	-31	+1
EDM	.357	.545	-30	+3
LAK	.343	.435	-44	+7
CJB	.341	.566	-20	0
TBL	.341	.402	-24	-21

Draft Lottery: Three True Outcomes

The NHL Draft Lottery will be held Monday night at 8pm. You can watch it on the NHL Channel if you get it. From a Thrashers perspective there are three outcomes that affect our team.

1. The Thrashers pick #1 because they win the lottery--14% chance.
2. The Thrashers keep their #3 position because either the Lightning or the Kings win or one of the teams way down the list wins--remember that teams that win can only move up 4 spots so if Buffalo wins it does affect the Thrashers one bit--61% chance.
3. The Thrashers pick is falls to #4 because one the three teams just below us in the order (Blues, Islanders or Columbus) wins the pick and leapfrogs over Atlanta--25% chance.

So the largest odds (61%) are that the Thrashers stay at the number three spot with a 25% of falling one slot and even smaller chance of rising two spots (14%).

Personally I'd be thrilled to see Atlanta win, but mostly I hope not to slid. The top three players look to be very promising with star potential.

End of the Season: What Went Wrong the Thrashers?

Tonight marked the 8th Fan Appreciation Night I've attended as at Philips Arena. For me it was by far the most depressing one of all. In previous years I could console myself with the fact that the franchise was still young and the future was bright. All that suffering during the previous 82 games would lead to a high draft pick and a superstar to build around. Here we are 9 years later and the future is all we have to talk about. Honestly it just makes me sick.

When you consider that more than half (8 of 15) teams in each conference make the playoffs each year--the bar is not that high. Unlike say baseball where only 8 teams out of 32 get into the post-season, in the NHL 16 out of 30 do. You only need to be AVERAGE to make the playoffs and yet despite the level playing field provided by the salary cap we've been above average just once.

I think the sorry state of affairs can be summed up pretty succinctly. Out of roughly 500 players in the NHL Ilya Kovalchuk was second only to Ovechkin in putting the puck in the net. Kovalchuk is an elite offensive player. Despite having that singular talent on their roster the Atlanta Thrashers will finish the season ranked 27th out of 30 teams in the NHL. You don't have to be a genius to figure out that this franchise has fundamentally failed to find complimentary talent to fill out the rest of the roster. It doesn't matter whether you look at free agency, drafting or trades there are errors and mistakes on all fronts.

God's honest truth is that the roster wasn't good enough. You can make all sorts of excuses such as the distraction of the Hossa contract or the "surprising" down seasons from Kozlov, Zhitnik, Exelby, and McCarthy. But at the end of the day EVERY team has pleasant surprises and unpleasant surprises. No team has everything go right for them. Good teams evaluate and weight risk, bad teams are frequently "surprised" when things go badly.

The truth of the matter is that the Atlanta Thrashers were unusually healthy. You will not hear GM/Coach Don Waddell tout this fact but go compare the number of games played by the Thrashers stars to some other teams around the NHL. Notice that other teams lost key heart and soul superstars like Lidstrom, Sakic, Crosby or Brind'amour for extended periods of time. Sure Kovalchuk played hurt this season--I can find beat up guys in every single NHL locker room who dress anyway. The Thrashers lost very few man games to injury this year--that is not why they finished among the worst clubs in the league.

What killed this club more than anything is that their older players started playing like--well, older players. The Thrashers organization appears to have put together a roster in which they assumed (or hoped) that every player on the roster would play at or near their peak. That's like opening up a car wash and assuming it will always be sunny and dusty. Any good business plan starts with a realistic model of what to expect. If you assume that conditions will always be favorable you're going to go out of business rather quickly.

You can't assume that everything will work out in your favor. You can't assume that Slava Kozlov, who had a career high 80 points at age 35 will defy Father Time forever and sign him for his age 36, 37 and 38 seasons. You can't assume that Bobby Holik is going to be the same shut-down center at age 37 that he was back when he was 30. You can't assume that notorious slow starter Alexei Zhitnik is going to find the Fountain of Youth at age 35. You can't assume that a 2nd line center like Todd White will remain effective until he turns 36. The Thrashers bet against the house repeatedly heading in the 2007-08 season and house cleaned them out. The Thrashers either need to place smarter bets or they need to find a luckier GM.

Are there old players who remain effective past age 35? Yes, but take a good look at the list of names, most of them are Hall of Fame level talents (like Mark Recchi or Nick Lidstrom or Chris Chelios) those type of players can be expected to remain effective as they get near 40 or pass it. On the other hand, if your name is unlikely to ever appear on a Hall of Fame ballot, the odds are very good that if you're going to experience a performance crash between 35 and 40. You shouldn't be surprised when high-risk investments crash and fail to provide value.

The Thrashers 2007-08 roster betrays an inability to accurately forecast player risk factors in my opinion. A NHL team is a multi-million dollar business and yet NHL teams are run in a seat-of-the-pants fashion that would be unacceptable in many other business fields. Any successful business must make accurate judgments about the risks and rewards associated with their business and take the route most likely to result in a positive outcome.

Again, the Thrashers failure this season is not a failure of any one person or the result of a devastating injury to a key player. The Thrashers failure this season is because the team simply did not properly evaluate the risks associated with their players. Before the season I ran a crude forecasting model of each NHL team and my simple model had the team finishing 11th but within shouting distance of the 8th spot. Basically my model said the Thrashers would need to be lucky to make the playoffs. My model incorporated age related declines for older players and increases for younger guys. It turns out my model was too optimistic since the Thrashers will finish 14th (although my model had Hossa playing a full season). My question is: do the Thrashers even have a model? And if so, are they working on improving it--because it sure blew up this season.

I know that in my day job I'm evaluated every single year. Those evaluations have caused me to change what I do at work, and I think they have made me better at my profession. If I don't perform in a satisfactory manner I don't advance. What about the Thrashers? What is their evaluation of this season? If they just chalk this season up to "bad luck" then they haven't learned anything at all. If they decide to turn over a new leaf and try to be more accurate in their evaluations of both the upside and downside of players, than I would have more hope for the future. Only time will tell.

Lehtonen's Rebound Control

Over on the Smirkin Chicken the poster Clambake pointed out that Kari Lehtonen faces more shots per minute of ice time so far this season than either Pavelec or Hedberg. Clambake concludes: To me, that suggests that Lehtonen is giving up more rebounds. Had he faced shots at Hedberg's clip, he would have faced 100 fewer shots so far (2.13/game), which works out to 8 fewer goals for the season.

I found this pretty interesting. Now it is possible it is just a fluky random thing and Lehtonen has simply been unlucky this year in terms of starting on nights when the team has played especially poorly. If it is just random bad luck then it is unlikely to be present in other seasons, so let's take a look.

Shots Faced Per Minute Played
2005-06
.518 Lehtonen
.485 Other Thrasher goalies

2006-07
.527 Lehtonen
.473 Other Thrasher goalies

2007-08
.578 Lehtonen
.534 Other Thrasher goalies

OK, the data indicates that this a "real trend" since it is consistent over three season. I must admit I'm a bit surprised to see this much of a difference. Now I will say that it is possible that Kari gets the starts against all the tougher teams as the team's #1 but I seem to recall Moose facing his share of tough opponents. The "strength of opponent" hypothesis by using an Expected Shots per Opponent framework to see if Lehtonen happens to face teams that shoot more but at the moment I'm too busy for that, but maybe this summer I can work it up. My expectation is that this pattern can't all be explained by the opposition.

So for the moment let's accept the argument that Kari's rebound control is below average. If that is true how much of a difference is it making? If we calculate the "extra" shots Kari has faced and assume a league average 10% shooting percentage, the numbers suggest that the higher shot differential has led to 7 extra goals against in 2005-06, 21 extra goals in 2006-07 and 12 extra goals against in 2007-08.

How much of a difference would that have made in the standings? Using the Pythagorean ratio which predicts non-shootout winning percentage very accurately, I estimate that it cost the Thrashers 2 standing points in 2005-06, 8 standings points in 2006-07 and 3 standings points in 2007-2008. The first two years are rather significant differences. If the team have finished two points higher they might have made the playoffs in 05-06 and the extra 8 points in 2006-07 might have meant that they don't need to make all those trades at the deadline. This season the team is so bad that the extra three points simply don't matter.

Conclusion: Until I plug in the strength of schedule numbers I'm still not completely convinced this differential is all rebound control. On the other hand, something is clearly going on. I will point out that a lower number of shots could also be produced if the defense has less confidence in the netminder and goes crazy trying to block everything. If you have a lot of confidence in your goalie you get out of the way and let him stop it--he's wearing more padding.

One last thought--it Lehtonen is more prone to rebounds than other goalies, this is something the Thrashers ought to keep in mind when going after free agents. Get out the game tapes and look at which UFA defensemen are great at marking their opponents near the net so opposing forwards don't have a lot of time to put home the loose pucks. Those sort of defensemen could be especially valuable.