Young Men on the Draw
The Edmonton Oilers are a bad team on the draw, and they have been for a few seasons now. Part of the problem has been that they've had bad centers over the last couple of seasons. But that's not to say there aren't reasons for their lack of ability - at least some of the reason, it seems to me, is youth. Faceoffs are, in many ways, just like any other puck battle in the game in that success requires strength, coordination, and a whole lot of practice. But just how much should these young players be expected to improve? And how do guys like Sam Gagner and Andrew Cogliano compare to their peers? After the jump I'll take a look at all of the centers who first entered the league at 20 or younger and who have taken at least 200 even strength faceoffs in each season from 2005-06 to 2007-08.
Unfortunately, there are only twelve players who meet the criteria, so we'll be dealing with a pretty small sample. In fact, because the sample is so small, I thought I'd begin by listing out how each of those twelve players have done in terms of EV FO% since the start of their respective careers:
What becomes clear here is that each player is unique. Most players demonstrate some improvement over the years, but some just flat-out don't (I'm looking at you Evgeni Malkin). Nonetheless, I thought it might be instructive to look at the average improvement by age:
Again, this is a very small sample, so it's hard to come to any really solid conclusions, but it does look like players will improve as they get older. All three of Sam Gagner, Sidney Crosby, and Jordan Staal improved substantially from 18 to 19, and several players made large improvements from 19 to 20 as well. After that, the improvement by age slows down but doesn't stop, which is pretty encouraging for a fan whose team is counting on Sam Gagner to become a reliable two-way player. It's somewhat less encouraging for a guy like Andrew Cogliano who has made decent progress as a faceoff guy this past year, but is starting from so far back that it's hard to imagine him every being passable on the dot.
This should also be instructive if the Oilers decide to play Taylor Hall at center for long stretches this season. It will be very tempting to say that he can't play the position if his faceoff percentage is somewhere between 40% and 45% in his first year, but that's not really the case. In fact, 45% would be the second best showing of any of the 18 or 19 year-old rookies on this list, behind only Jonathan Toews (whose rookie season looks amazing in this context).
I thought it might also be helpful to look at how players improved based on their years of service:
This is a somewhat less encouraging picture for players like Gagner and Cogliano. This is only based on six players, so it will obviously require a larger study to be really helpful, but the same group of players who improved substantially in their first three years taper off in the text two. It doesn't make sense to bet against Sidney Crosby all that often, but I think this preliminary information would help me to take the under if we set the line for his EV FO% at 54.58% (his total from one year ago).
As for the Oilers, let's hope that Cogliano is moved to the wing for good and that Gagner has one more development spike in the circle still to come.
14 comments
|
0 recs |
Do you like this story?
Comments
Interesting stuff Scott.
This little article is one of the reasons you guys are becoming the gold standard in the Oiloblogosphere.
You take the conjecture and subject to the cold hard facts and find out exactly where the bear shits in the buckwheat.
Love it.
You could probably increase the sample of the study by looking at players’ first 3 – 5 years of work (rather than just guys who started as teens). Dustin Boyd played from 20-23 and his F/O ability actually devolved slightly (50.8, 45.5 and 49.7) although that’s just one player.
In addition, Gabe’s recently noted that teams win an average of 52% of O-zone draws, meaning having a high zone-start could technically bump a guys F/O average (and vice versa). Perhaps controlling for zone start would add something to the analysis?
I actually did include guys who started at 20, but Boyd didn’t play enough at 20 to qualify. At any rate, that’s definitely one way to increase the sample, which would be pretty important if we wanted to draw a conclusion with a great deal of confidence.
I must have missed Gabe’s piece on OZ draws. That’s actually the opposite of my expectation since I’d think teams would rather protect poor centermen by having them take more OZ draws, rather than more DZ draws. Controlling for zone would also be pretty difficult, unless Gabe has that data handy for individuals (the NHL doesn’t list it).
by Scott Reynolds on Aug 22, 2010 3:46 PM MDT up reply actions
Due respect, but correcting for ZoneStart would be smoothing out molehills. 52%? Even if sustainable, not that big a deal. Consider:
— Of the 380 forwards who played 41+ GP, in 2009-10, over 90% had an OZS% in the range of 40-60%. Thus there are a lot of cancelling events and the surplus in one end or the other is a relatively small number. e.g. a guy taking 60% of his end-zone draws in the O-zone and winning those at the rate of 52% and winning his D-zone draws at a rate of 48% will wind up with a net percentage of 50.4%. And he’s near the outlying end. For most players the expected difference would be closer to 0.2% or even less.
— Official faceoff percentages also include neutral zone draws which would serve to further mute the effects of ZoneStart in the big picture numbers.
— Those neutral zone starts may be the elephant in the room. I’d be keen to know how many players might have anomalous (high OR low) neutral zone FO%. Haven’t seen the big picture stats to know, but in viewing individual game faceoff data I’ve noticed more than a few where a guy was more effective in either end zone than at centre, or conversely where a guy seemed to be padding his stats by winning more meaningless faceoffs. Could be random distribution or it could be guys playing games with each other or it could be guys not giving a shit at certain points or whatever, there’s lots of possible explanations. But my gut says that that effect will kick the shit out of any ZoneStart effects.
— On the other hand, I think adjustments for special teams duty should absolutely be done. The effect isn’t that much greater – 54% rather than 52 – but you will find way more examples of guys who take a whole lot of faceoffs on one special team and none on the other, so they won’t cancel out near as much.
Writer for The Copper & Blue and primary shareholder of Zorg Industries
"Never be ashamed of who you are" -- Jean-Baptiste Emanuel Zorg
by Bruce McCurdy on Aug 22, 2010 6:34 PM MDT up reply actions
Even if he gets around 50% its should be ok. Furthermore, if he gets to center the 1st powerplay, he might be able to win more draws and bump up his percentage slightly
Sins can be forgiven but conscience is a killer.
These are just EV FO% numbers, so PP and PK don’t enter the equation. His overall FO% should, of course, go up if he spends more time on the PP.
by Scott Reynolds on Aug 22, 2010 3:38 PM MDT up reply actions
Cool, thanks for the info. So last season Gagner actually did about as well as or better than 2/3 of the players you’ve listed (I’m talking ball-park here). If you don’t mind my asking, what’s the rationale for excluding centers that started when they were older than 20? Do players that debut at a later age generally start with better face-off percentages? Or is there typically an improvement period for all new NHL centers regardless of age?
I don’t know if I like mustached Gagner, but it definitely makes him look a lot less “elvish.”
by Stephen's Beaven on Aug 22, 2010 11:23 AM MDT reply actions
I chose guys who started between 18-20 because I wanted to compare them with Gagner and Cogliano and look forward a couple of years beyond where those two are at now. I actually expected there to be more than just ten other guys. I may expand on the study by either including players who came into the league later or using for seasons of data. The problem with those seasons is that they’ll run into the lockout so everyone will have a blank year.
by Scott Reynolds on Aug 22, 2010 3:41 PM MDT up reply actions
Thanks for clarifying. It would be cool to expand the study, but what you’ve done already is very informative.
by Stephen's Beaven on Aug 22, 2010 6:39 PM MDT up reply actions
More excellent work, Scott, you’re really on a roll!
I’ve long been convinced that faceoff percentages improve as a function of age, and this seems to confirm that. Would be interesting to see more data on a larger group of players down the line, but this does establish a baseline.
Good idea to to focus on EV FO% only, since special teams duties will corrupt expectations/outcomes somewhat depending on what ST role(s) a given player might have.
Writer for The Copper & Blue and primary shareholder of Zorg Industries
"Never be ashamed of who you are" -- Jean-Baptiste Emanuel Zorg
I’ve long been convinced that faceoff percentages improve as a function of age, and this seems to confirm that.
Except Cogliano and Malkin
Editor of The Copper & Blue, and leader of The Cult Of Hartikainen.
Hey now, Cogs was quite a bit better last year than in his rookie season. It’s just that his starting point was so bad that quite a bit better is still really bad.
by Scott Reynolds on Aug 22, 2010 7:45 PM MDT up reply actions

by 


































