DOES YOUR TEAM NEED IMPROVEMENT? FIRE THE COACH,
LEARN HOW TO PLAY DEFENSE,
AND, BY THE WAY, YOU MIGHT HIRE A NEW BACKUP GOALIE

I looked at 20 seasons, choosing the two teams from each season that improved the most in standings points: the median improvement was 24 points. DEFENSE: Thirty of the forty teams improved their GA totals significantly. The median improvement was 18.5 %, or somewhere north of forty-five fewer goals allowed . Improvements in offense were less dramatic at 9.5 %. COACHES: Not surprising that there were so many coaching changes, the 60% (24 of 30) dovetails nicely with the fact that 65% of these teams had declined in the standings during the previous season. Is there a "Bruce Boudreau effect" at work here? ROSTER TURNOVER: I tried to identify a basic 20 man roster and I don't think I'm far off. Obviously there were more roster movements (injury call-ups, rookie tryouts, veterans acquired in trade at the end of the season) but the median is 6 players out of the hypothetical 20, or 30%. Some teams made wholesale changes: in Mario Lemieux's second season, he had 11 new friends on the Penguins roster. ROOKIES: The median is 2 on the 20 man roster. A few like Lidstrom, Malkin (the Pens did not improve by a single point in Crosby's rookie year) were major contributors, but mostly these teams used rookies as supporting cast. GOALIES: The median is 1: there were a few cases where a number one goalie made a major contribution but most of the goalies added were either backups or junior partners in a shared goaltending arrangement. CONCLUSIONS: Anecdotally, many of these teams had made significant changes the year before their big improvement. So if the Chicago Blackhawks add some more defense this year, look out for them. Alternatively, what was Wayne Gretzky Thinking (he was thinking about Kurt Sauer?) by trading away two solid defensemen?

Draft positon and success

Using 200 NHL games as a standard, 90% of defensemen taken among the top ten picks, have achieved that goal, 65% of those taken in picks 11-20, 59% of those taken in picks 21-30. The most successful drafts in terms of d-men were 1979 (13 of 14) and 1998 (10 of 11).

D-MEN TAKE LONGER...BUT NOT MUCH LONGER

The median age at which the 80 defensemen currently in the NHL who were first round picks played at least half of their team's games: 20.5 years. The median for 85 centers picked in the first round: 20 years. Forty-three or 51% of the centers had played at least 41 games by age 19: by contrast,twenty-two or 28% of the defensemen had achieved this by the age of 19.It's a difference but not that significant when you consider that forwards can break in by playing 4th and 3rd line minutes (think Bryan Little, Martin Hanzal) and top flight defensemen rarely do that.
Measuring the quality defensive play is difficult statistically but I looked at offensive contributions and found that of the 38 defensemen who were picked on the first round and who showed offensive promise, the median age at which they first achieved their career average PPG was 22 years.