Tuesday, June 22, 2010

More on Timmins' slugging percentage at draft table

Robert Lefebvre at Habs Eyes on the Prize has linked to a thorough NHL-wide draft analysis done by Daoust of Leafs blog Pension Plan Puppets

The impressive study has a section which examines the years 2001-2007, which precedes Timmins' arrival in 2003 but is a pretty accurate portrait of his time here. The work shows that while Montreal is among the league's best at finding players who play in the NHL, finding impact players has not been quite as successful. In those years Montreal has drafted one player that has averaged 40+ points per season: Tomas Plekanec, who has averaged 50.8 points per season over his five full years in the NHL. So the Canadiens have been very good with quantity, not so hot on quality.

Having said that, it should be noted that 11 teams did not draft a single player that has averaged 40 points per year, more than a third of the league. Also, Montreal is one of eight teams to have found only one such player in that time frame. So that's 19 out of 30 teams which drafted one or zero such players, and that only makes the Canadiens a non-elite team when it comes to finding upper-echelon talent.

I wanted to get an idea of how the quality played out on a per game basis of drafted players who made the NHL. Daoust's chart includes each team's total number of games played and total number of points for their drafted players, so using the two numbers I figured out the points per game per team. 

What I found is that Montreal's points per game is 20th in the league over the period of 2001-2007, even though the Canadiens had the sixth-best average pick over that time. Of course, Montreal only had one top-five pick in that period and used it on a goalie, and the next highest selection at #7 was used on Mike Komisarek in 2001, prior to Timmins' arrival. Montreal's points per game of .380 is below the league average of .411. Here's the full list, and please let me know if my math or methodology is wrong (I don't think you'll be shocked to see the top two).

  1. Washington - .567 points per game
  2. Pittsburgh - .545
  3. Philadelphia - .489
  4. Buffalo - .467
  5. St Louis - .462
  6. Atlanta - .460
  7. Colorado - .458
  8. New Jersey - .456
  9. Anaheim - .440
  10. Carolina - .430
  11. Vancouver - .426
  12. Los Angeles - .425
  13. Chicago - .424
  14. San Jose - .421
  15. Ottawa - .408
  16. Toronto - .3944
  17. Calgary - .3942
  18. Florida - .393
  19. Detroit - .3804
  20. MONTREAL - .3802
  21. Edmonton - .379
  22. Minnesota - .372
  23. New York Islanders - .3501
  24. New York Rangers - .3496
  25. Dallas - .3487
  26. Nashville - .3482
  27. Columbus - .347
  28. Boston - .342
  29. Phoenix - .309
  30. Tampa Bay - .198
There are some teams I'm frankly shocked to see so low. A club like Atlanta, for instance, appears to be masking a real drafting deficiency with the prolific scoring of a no-brainer pick like Ilya Kovalchuk, who has over half of the Thrashers total points from drafted players in the study (642 out 1,111). And what to say about Columbus, who has Rick Nash accounting for about 32 per cent of his team's total (422 of 1,321).

A lot of factors go into a list like this, especially first round draft position and how many defencemen and goalies were selected by each team. But I think it gives a pretty good picture of the Canadiens so-called draft success of late.

If finding players who simply play in the NHL is what you're after, the Canadiens have been among the best. But if finding players who make a difference in winning you games is your ultimate goal, then there's some improvement needed.

In other words, it's time for Timmins and the Habs to start hitting some long balls instead of line-drive singles. 


beaker said...

You have already stated that the habs draft more NHL quality players than the average team.

Thus they have more NHL Games Played than the average team.

Thus the Points per game is not an effective analysis.

A guy like Greg Stewart or Matt D'agostini, or Ryan O'Byrne, or Jaroslav Halak, actually hurts the habs PPG average... whereas a team that drafts a guy who never plays a single game in the NHL is doing better because their first rounder is given more weight.

I can't accept an analysis where drafting a guy in a late round who makes the league as a low scoring 4th liner or 3rd pairing defenceman actually makes your ranking worse than selecting a bust in the same late round.

Arpon Basu said...

I think you'll see throughout my piece that I admit it is flawed on many levels, including the level you point out. But other than Plekanec, Halak, Price, Streit and maybe Komisarek (plus several players not included in this study, led by Subban) the Habs have not really drafted an impact player of any kind over this period. Is that a good track record? I'm not sure.

MathMan said...

Our expectations about drafting are unrealistic because we're always going on about the guy that got away... and never stop to realize that over 8 years the "guy that got away" went to 8 different teams.

Holding up Timmins to the combined draft success of 8 teams (and teams that have had some success) hardly seems fair.

HabsWatch said...

Ranking teams in draft performance by points per game is a poor way to judge how teams have actually drafted. Teams who horde top 5 and top 10 picks had better produce more impact players than teams who constantly pick mid or late round because they've remained competitive year in and year out.

You have to consider how well teams have drafted given the their opportunities and where they usually pick in the draft. Also, going to 2007 is pointless because so many teams have prospects who will become solid NHL players who are still in College, overseas or perhaps just played their first AHL season.

TK said...

Interesting analysis, although as other posters have noted this method does have its limitations. Not the least of which being a team may actually bring down their overall PPG by drafting solid defensive d-men on a consistent basis.

I am not saying Montreal fits that category, but it would be interesting if you break down the GP category by position. Such a number would probably be a better indication of how well certain teams are in drafting for the D-position.

On another note, this reminds me of that book "Freakonomics".

Sliver24 said...

It seems to me that the fact that the Habs have more man-games by draft picks is actually a bad thing to a certain extent.

To me that says the Habs are putting players on the NHL roster that have no business being there at that point in their careers. Instead of vetting a player in the AHL they're doing so in the NHL, leading to a simultaneous increase in games played and decrease in PPG. Examples include D'Agostini, Pacioretty, Latendresse.

I guess that's part of being a perennial bubble team with a fan base that's desperate for something to cheer for.

MathMan said...

@Silver24: Either that or they're digging up more NHL-caliber players than anyone else.

I remember reading that there are more players in the NHL that were drafted by the Habs than by any other single team. I don't know if this is still true but Montreal would still be up there as team-of-draft for an unusually large proportion of NHLers... regardless of what team they play for. That means a lot of teams employ Habs draftees as part of their NHL rosters.

beaker said...

D'agostini was not rushed. He had 2.5 AHL seasons on him before he was called up. He was called up in December 2008. In September 2009 he no longer would have been waiver eligible.

You can make the case for Lats and Pacioretty, but not Dagostini.