Some projections seem strange at first glance, which forces us to take a deeper look and to challenge our beliefs. For example, the VUKOTA system projects a spike in goal scoring against teams in the Atlantic Division this season, like Pittsburgh, Philadelphia and the New York Rangers. Why?
The Philadelphia Flyers are easy enough to explain. When you lose Martin Biron to the New York Islanders and Antero Niittymaki to Tampa Bay, and replace them with Ray Emery and Brian Boucher, you hardly need advanced statistics to predict a dramatic decline in the crease. As for the Penguins, they lost their two best defensive players, so their hole is on the blue line, not between the pipes.
That leaves us with the New York Rangers, and the real enigma. Henrik Lundqvist was remarkable in 2008-09, earning a fifth-best GVT (goals versus threshold) of 18.8 in 69.2 games. It's easy to react with surprise when you see his VUKOTA projection for 2009-10 is a GVT of "only" 12.2 in 59.3 games. Even if we normalize that to the same number of games, it's still a drop of over 20 percent, which is certainly more than merely a "return to Earth" adjustment. Even though he'll still be ranked as the No. 6 goalie in the league, how can we explain such a big plunge for a 27-year-old perpetual Vezina finalist?
Lundqvist's problem isn't consistency. Some people claim that the shutout is a useless statistic, but I disagree. His shutout total dropped from 10 in 2007-08 to just three last season, but since his goals-against average remained roughly the same, that drop-off is actually a sign of increased consistency. Remember, fewer shutouts with the same GAA also means fewer four or five goals-allowed nights, too. To confirm his great consistency we can look at quality starts, which are awarded only when a goaltender has played well enough for his team to win. Last season, 47 of Lundqvist's 70 starts were of quality, giving him a quality start percentage of 67.1 percent, which was second in the NHL to Boston's Tim Thomas.
Given these amazing numbers, why would the VUKOTA system project such a decline for Henrik Lundqvist? The answer lies in the postseason.
The VUKOTA system uses playoff data as an indicator of future performance. Lundqvist's career playoff save percentage was .907 going into the postseason in 2008-09, which is 10 points lower than in the regular season.
While many will remember some stellar nights against the Washington Capitals in Round 1 of the playoffs (a road shutout in Game 2, a 2-1 win in Game 4 and some insane saves in a 2-1 loss in Game 7), he ended the series with a .908 save percentage, a 3.00 goals-against average, and was pulled twice. Despite his reputation, those numbers are fairly middling and account for about 10 percent of his projection score, with the other half of the projected drop the result of simple regression to the mean.
To see an example of this from the other side, check out Jonas Hiller and the Anaheim Ducks, who are projected to have the second-best goaltending in the league this season. Last season Hiller had a GVT of 14.9. That's good, but not great. Three goalies who had comparable seasons include Biron, Dwayne Roloson and Nikolai Khabibulin, but their average projection next season is only 8.4 GVT, just over half of Hiller's 16.3 GVT projection. Obviously age is a factor since Hiller is 27 and the rest are all over 32, but the Swiss superstar was awesome in last season's playoffs with an amazing NHL-best .943 save percentage in 13 games, greatly boosting his season numbers.
In the end, is playoff performance really enough to project King Henrik's slide out of Vezina contention, or to project Hiller as one of the favorites for the award? We'll all be watching very closely to find out.
A version of this story originally appeared on ESPN Insider .
Robert Vollman is an author of Hockey Prospectus.
You can contact Robert by clicking here or click here to see Robert's other articles.