Comparables: Sam Gagner
Jonathan Willis
May 20 2010 10:33AM
With all of the young prospects coming up the system and the first overall pick for this summer’s draft in the bag, it’s easy to forget about the importance of the Oilers’ best young forward currently playing with the NHL team, Sam Gagner.
Part of that is the apparent lack of development; Gagner has yet to match his scoring totals as a rookie, although his overall game has improved by leaps and bounds. That lack of offensive development has some wondering what kind of player Gagner projects as. We’ve seen different names tossed out there; Jim Matheson suggest Stephen Weiss while Lowetide generally prefers Vincent Damphousse as a benchmark.
Gagner’s track record is far better than Weiss’s was at the same age, so I suspect that Matheson’s comparable is underselling things, but I think the Damphousse comparison is realistic. Still, I also think we need more players to compare
One of the difficulties of comparing players across different years is that league scoring hasn’t been consistent; in 1981-82 NHL teams scored four goals per game on average, while immediately prior to the lockout they were just a hair over two and a half.
To make comparisons a little fairer, I went back to the 1980 NHL Entry Draft, and looked at all the forwards who a) were 6’1” or shorter b) played in the NHL in their first year of eligibility and c) didn’t have a truckload of penalty minutes. I then took those players and adjusted their offence to reflect an NHL season in which teams averaged three goals per game, and projected that offence over an82-game schedule. I did the same with Gagner, and ended up with a list of five players who had comparable scoring over their first three NHL seasons:
| Player | Year One | Year Two | Year Three |
|---|---|---|---|
| Sam Gagner | 15-40-55 | 18-28-46 | 19-33-52 |
| Jim Fox | 16-22-38 | 24-30-54 | 23-33-56 |
| Bobby Carpenter | 25-27-52 | 25-29-54 | 22-31-53 |
| Vincent Damphousse | 18-21-39 | 11-32-43 | 21-35-56 |
| Pat Falloon | 22-28-50 | 23-23-46 | 20-28-48 |
| Tim Connolly | 16-22-38 | 11-34-45 | 11-40-51 |
Carpenter, Connolly and Fox all have roughly comparable birthdays to Gagner. Damphousse is seven months older over the seasons in question, while Pat Falloon’s birthday makes him almost a full year older than Gagner.
Jim Fox was a tremendously talented player who managed to survive 1980’s hockey despite standing only 5’8” tall. He spent just over eight seasons with the Kings, scoring just under a point per game in each of them, before injury ended his career. He missed the entire 1988-89 season recovering from a knee injury suffered during the stretch drive the year before; he attempted to make a comeback the following year but only played 11 games before being forced to retire.
Bobby Carpenter was the first high school hockey player to jump directly into the NHL. He had a strong start to his career and in his fourth season scored 53 goals and 95 points. Unfortunately, it was a one-off; his scoring returned to its previous levels and more or less stayed there for the majority of his career. After Carpenter turned 30 he became more of a defensive specialist, but in the end he managed to play 1178 games at the NHL level, recording 728 points.
Vincent Damphousse was the sixth overall selection in the 1986 NHL Entry Draft, and after three seasons of decent results in Toronto he broke out in his fourth year, scoring 33 goals and recording 93 points for the Maple Leafs. He would either near or surpass the point-per-game mark every season for the next eight years and ended his NHL career with 1205 points in 1378 games.
Pat Falloon had some up and down seasons after a promising start to his NHL career. He struggled during his fourth season, rebounded the following year with a new team, struggled some more and was dealt again, struggled some more and then got signed by Edmonton; he was okay for the Oilers before getting shipped off to Pittsburgh, and he eventually found himself playing in Switzerland.
Tim Connolly has had an occasionally brilliant career that has been plagued by injuries; when healthy he’s an excellent offensive player and when healthy has been a number one centre since the NHL lockout. Unfortunately, he’s not healthy all that often.
It’s a fairly nice group to be compared to, especially since Falloon’s a full year older and I’m probably stretching to include him in the group anyway. Connolly and Fox both would have had significantly better careers if they were healthy, Damphousse was a very good player for quite a long time, and Bobby Carpenter had a long and relatively successful NHL career.
@OB1 Team Yakopov - F.S.T.N.F
This is outlashing like a child wrote by GSC"So, now we're altering stats in order to make Gagner look comparable to Damphousse...
When is this nonsense going to end?"
Great job trying to get browny points, do we get browny points now ?
For the most part I'm on your side with this argument, JW, yet this statement is a bit loaded.
While raw statistics are impartial, the use of stats aren't. It's the analysis of stats that make them relevant and frame their use, as any good politician could tell you (if there's such a thing).
For the record, I think that your manipulation of points to reflect different eras in hockey is entirely appropriate and one of several good ways to compare players.
Yes it is, however if that was the first time it would probably be ignored. However it isn't.
Hey JW. I think the article was a great read and an interesting comparison of how Gagner looks compared to players with similar stats. I am no math whiz and calculus is something best left in my distant past, but I feel that given what you had to work with, you did a good job here.
Personally, I would like to see comparisons of other young Oilers using the same method. Maybe Cogliano, Gilbert etc...
@OB1 Team Yakopov - F.S.T.N.F
Oh shizz really, my bad. ~Hopes for browny points from the Ogden family~
@ Andrew W:
That's a fair point to make; my argument lost something because it was so concise.
A better statement would be that raw statistics are impartial, and that any analysis of those statistics can be checked with logical arguments or with data for how accurate it is.
For instance, my argument could be checked in several ways:
1. If there was evidence that individual scoring didn't fluctuate with league scoring, my argument would be invalid. As this assumes that players who became skaters in the mid-1980's are just more talented, as a group, than today's skaters, and that the reverse is true for goaltenders, I think we can safely discard it.
2. If I excluded players from consideration. All anyone would have to do is go back over the last 20 years of entry drafts, using my criteria, and double-check. The data's publicly available and since I've already established method it would presumably take less time than my original article did.
3. If my math were wrong on the players included. As I've included my process, this is remarkably easy to check - though I note none of my critics have bothered.
What I'm getting at is that people who say stats can be manipulated to say whatever the writer wants don't understand statistics or are simply too lazy to check the arguments presented.
Hey man, essentially calling someone a liar/fraud when you have nothing to back it up is pretty serious in my books... kind of like something a child would do.
Obviously not...but comparing a player like Gagner to a player like Damphousse? Really?
Damphousse - who is arguably 2 inches taller and anywhere from even in weight to 10 lbs heavier than Gagner, by the way - managed to continue a high level of production throughout his career, as I mentioned previously. He didn't just put up big numbers in the 80's, he did it throughout the 90's and into the new millennium, as I mentioned previously.
In fact, here's Damphousse's production through the 80's (4 seasons): 315 GP, 256 PTS, 0.81 PPG, rough average of 66 PTS per season
And the 90's (10 seasons): 690 GP, 673 PTS, 0.97 PPG, rough average of 80 PTS per season
And the 2000's (5 seasons): 373 GP, 276 PTS, 0.73 PPG, rough average of 60 PTS per season
Career: 1378 GP, 1205 PTS, 0.87 PPG, rough average of 71 PTS per season.
So, again, what purpose does placing Gagner in the 80's era serve? It doesn't account for Damphousse's best years in the mid 90's when the average goals per game in the NHL dropped nearly a full point from the 80's. It also doesn't account for the fact that Damphousse scored 70 PTS in 82 GP (.85 PPG) in the 99-00 season and 46 PTS in 45 GP (1.02 PPG) in the 00-01 season, where the average goals per game were down to around 5.5 and he was over 30 years of age. Pretty impressive for a player in the twilight of his career and facing a more defensive-minded league, yes?
What I'm saying is that Damphousse was a proven, high-level scorer throughout his career, and to compare Gagner to such a player because the stats may be changed to reflect the higher levels of offence in the 80's doesn't at all tell the whole story. It also doesn't account for the fact that Damphousse showed consistency and productivity throughout his career. Hell, he still averaged 60 PTS per game when the goals per game average were down in the 5 goal range. I don't see a Vincent Damphousse in Sam Gagner, and that's why.
Yeah name calling etc isn't required
Again, my argument centers on Damphousse because you stated very clearly that you believed him to be a comparable to Gagner, which I don't think is anywhere close to being the case. That's where I disagree and I showed why in the comment above.
The inclusion of Falloon is irrelevant when you said it yourself that Gagner is comparable to Damphousse before you even jumped into showing your results. That's what I focused on. Notice I didn't focus on anyone else on that list for a reason: because I disagree with your assessment of Damphousse and Gagner.
I've been perfectly clear regarding my stance on that comparison since my first comment to this article. I didn't focus on Falloon, or anyone else for that matter, because I disagree on your belief that Gagner is comparable to Damphousse.
Did I call Willis is a liar? Show me where, please.
I said that he altered the stats and it shows in making Damphousse somehow look comparable to Gagner. And he did alter the actual stats, there's no getting around it.
@GSC
I don't think Gagner is a 70 point player also, but its good to get a article to see where others think he is heading.
I thought the calculus comment was kind of clever...
Canada is getting killed by Russia
@ GSC:
Here's what I wrote in the original article:
Why did I include that? Because we do need a range of options, like I presented.
The article doesn't say he'll be Damphousse. It suggests a Damphousse-like career is a possibility. It also suggests a Connolly-like, Fox-like, Carpenter-like and Falloon-like career are possibilities.
And at this stage of the game, I don't think a Damphousse-like career can be ruled out. If Gagner doesn't take a jump next season, Damphousse is no longer a comparable, but until that happens, he is.
By the way, how would you alter the stats to reflect the difference in scoring by era, since it's obviously unfair to compare them straight up? I'd be open to looking at a better method, if you have one. If you don't, I don't see what you're complaining about.
Only way Gagner get 90 point in a season is if the NHL goes to 144 games. Thats going by his current stats and not projecting that he will get almost double stats
Jonathon, don't bother. These guys are still trying to sell 1500sqft houses for $500,000 - because that's what they were going for 5 years ago. 1500sqft is 1500sqft - the conditions of the market don't matter.
GSC wrote:
So, you acknowledge that Gagner's numbers would be better, because "Well, duh... stats were inflated back then," but later on you argue that adjusting the stats is an attempt to paint a flattering portrait of Gagner.
So, are we supposed to leave the stats the same, or adjust them for era? Because your argument strikes me as hypocritical.
A 'yes' or 'no' would have been fine.
If you agree that comparing stats straight-across in different average-scoring eras is unreasonable, it logically follows that stats need to be normalized prior to objective comparison - doesn't it?
Do you think a 100 point player today = a 100 point player in the early/mid 80's?
I don't understand why you throw out Weiss as a comparable when he's far and away a better match to Gagner in virtually every way -- size, style of play and the era he plays in -- than the players you used. In those terms, Damphousse is the biggest reach and Carpenter is a close second.
You say it's because Gagner's track record over the first three years is far better, but that's a case of getting fixated on the numbers instead of really looking at and knowing the players involved. Weiss played just seven games as an 18-year-old while Gagner played 79. In Weiss' first three seasons, he played 134 games, while Gagner played more than the equivalent of a full season more, 223 games.
Weiss is five-foot 11 and 190 pounds. Gagner is five-foot-11 and 191 pounds. If you reject Weiss, then why stretch the size parameters so far, to six-foot-one or less and why go back so many years? Is is because there aren't enough players who are five-foot-11 or six-feet from the last 7-10 years who played as 18-year-olds?
Wouldn't that make for a better match in stature and eliminate much of the adjusting of stats to reflect the difference in scoring from different eras? Damphousse, for example, was every bit of six-foot-one and was very lanky. Sam is squat with a very different body type and reach.
In terms of style of play (and mental make-up), Weiss is also a very good match. In 2009-10, Weiss took 180 shots and Gagner 170. Weiss had 40 PIM, Gagner 33 PIM. In 2008-09, Weiss took 154 shots and had 22 PIM. Gagner took 156 shots and had 51 PIM.
I fully expect Gagner to turn out very much like Weiss, who had 60 points this season and 61 points last season. I also expect the Oilers to be pleased when that happens. Matheson had it right.
It was pretty strongly implied.
I noticed that I did that too...my mistake, attribute it to trying to make a timely reply to the other comments.
I would say that adjusting/altering the stats for era in order to make Gagner look comparable to Damphousse is flawed. It may work in other comparables, like you mentioned (and by the way, I have no argument with the other comparables, which is why I didn't include them in my comments). With Vinny, however, I just don't see the correlation despite your adjustment.
So yes, I think adjusting the stats to make a comparison of Gagner with Damphousse is painting a flattering portrait. I hope I've clarified my argument.
Next poll question.
How many points will Gagner get in his best year ?
1) 90+ 2) 70-80 3) 60-70 4) 50-60
I'm going with option #3
I'm not suggesting a better method. I'm suggesting that the Damphousse example doesn't add up, in my opinion.
I know that you made other comparisons, and again I don't disagree with those. What I disagree with is Damphousse, which I've tried to make clear from the get-go. We're just going to have to agree to disagree on that one.
Honestly, I hope you're right and that I'm wrong. I'd love to have a Vinny D in the Oiler lineup.
Also, Damphousse hit 90 points four times. But those 90-point seasons were put up back when the NHL was a much higher scoring league. Using this year's numbers, here's how those seasons equate; I'll even walk you through the math:
Damphousse scored 90+ points in the following seasons:
1989-90: 94 points 1992-93: 97 points 1993-94: 91 points 1995-96: 94 points
Here's how many goals per game per team were scored in those years:
1989-90: 3.69 1992-93: 3.63 1993-94: 3.24 1995-96: 3.15
Last season, NHL teams scored 2.84 goals per game per team. We can adjust for league scoring by taking the number for each year, dividing it by 2.84, and then multiplying that ratio by Damphousse's offensive production in those years, and then adjusting for games played, like so:
(2.84/3.69)*94*(82/80) = 74 points.
Here are those 90+ point seasons, adjusted for era and games played:
1989-90: 74 points 1992-93: 74 points 1993-94: 78 points 1995-96: 87 points
So, using the Damphousse projection and this methodology, I'm predicting Sam Gagner could be an 80-point player in today's NHL. I don't think that's a prediction that falls out of line with reality.
@Robin Brownlee
I couldn't agree with you more. 65 points would be a good season for Sam, now and in the future. Finally someone that is in reality
Possibly...but I think the Damphousse-Gagner example shows how this method isn't perfect.
@ Robin Brownlee:
I think Weiss is a strong stylistic comparison, for all the reasons you suggest. The reason I toss him out is because Gagner was so much further along at a much younger age.
Weiss didn't crack the 40-point mark until he turned 23; Gagner did it right after his draft year.
At the age of 21, Weiss put up 54 points over a full AHL season.
So while stylistically Weiss is a great comparable, I just look at their accomplishments relative to age and I can't justify that kind of ceiling for Gagner, because when we compare them at the same age Gagner comes out ahead offensively, each and every year.
@ GSC:
Are you suggesting that Damphousse's numbers were abnormally low in some way compared to the way he picked up his game later in his career?
I know each player follows a different development path, but what leads you to think that the first 3 years of their respective careers is nothing more than a coincidence? Is it their skill set?
@GSC
"I would say that adjusting/altering the stats for era in order to make Gagner look comparable to Damphousse is flawed."
Your missing the whole point, it was adjusted to give a baseline.
@OB1 Team Yakopov - F.S.T.N.F
*You're*
If I failed calculus, perhaps your failed English?
*You're*
If I failed calculus, perhaps your failed English?
@ GSC:
Correct me if I'm wrong, but here's how I understand your argument:
1. You think it's fine to adjust statistics for era.
2. You don't have a superior method of doing so then the one I just used.
3. But, for your own unspecified reasons, you think Damphousse's offence is simply too high a marker to aim for, even though you realize those 90-point seasons aren't 90-point seasons in the modern game, and thus you reject the method out of hand.
Is that an accurate representation of your point of view?
My weakest subject, skimmed by with a C though :)
@ Ogden Kings 'n' Pretty Things
A lot of lower-tier stars (i.e. good offensive players below the superstar level) don't break out until their third or fourth NHL season.
But I actually approached the argument from the reverse - I asked what kind of career we could expect from a very young player with decent offence but little upward curve over his first three seasons in the league.
I don't neccessarily think Damphousse is the best fit here; I think he represents the ceiling of possibilities, as Falloon represents the floor. It's a realistic possiblity, and not one I'd rule out, but if I were betting money I'd put it on a Connolly/Fox-esque career.
@Robin Brownlee
I've always liked Weiss. However Weiss is putting up those #'s playing first line with better teammates. I think Gags will out produce him if given the same opportunities next year.
I realize that you can probably find data that suggests that Zack Stortini projects to a 70-point player based on how some 70-point player played before he broke out. I think that might be what GSC is getting at, though. Once you've found a comparable in Damphousse based on past data (ie first three seasons), is there anything else that leads us to believe that Gagner might break out in the same fashion?^I didn't like the way I said this.
The data you've provided shows us that it's not unheard-of for a player to put up Gagner-like numbers and go on to do great things, like Damphousse.
We all agree that Gagner has some upside, but what specifically tells us that the correlation might be a little bit more than correlation?
Same with me and calc!
Cheers.
Unspecified reasons? Look at the damned comment I made above (#54). I made my reasons clear.
I said that in this instance, that being the Damphousse-Gagner comparison, the adjustment of statistics produced a flawed result.
As for #1, yes it can be a useful method (albeit not the be-all, end-all). As for #2, I don't use the same mathematical methods as you so no, I don't have a "superior" method.
It all makes sense now, you reflect your dislike for calculus onto those that use math to disect hockey.
;0
So you are saying it's impossible for Gagner to make the same type of leap in production that Damphousse made?
@ Ogden Kings 'n' Pretty Things:
Nothing.
I'll tell you this, though: I did look at better than 30 years of players, and using the limits I set (explained at the outset) this is what there is.
And I wouldn't bet on Gagner becoming Damphousse either. That's one possibility, but not to my mind the most likely (just the most optimistic that still falls within the range of realistic possiblity, IMO).
What I'm getting at, I suppose, is that it's wrong to completely rule a Damphousse-like offensive career out, as GSC has done. That's not to say that it will happen, but it could.
Like I said, I think Connolly or Fox is a more realistic comparison, and those are fairly nice players to be compared to.
If I'm not mistaken, the point Willis has made with this article is exactly what he just stated a few posts up:
Here is his projected lower bound, here is his projected upper bound.
A reasonable analysis.
Why everyone is harping on Damphousse is unclear. Sure, he was a taller, lankier guy but he never struck me as a physical, speed demon superstar. He was methodical, intelligent and played creatively in the right situation. I don't think that's too far a stretch from the type of player that the Oilers thought they were getting when they drafted Gagner.
I do have to ask the same question raised by Brownlee though: were there not enough sub-6' players to use for the sample?
I hated calculus, but enjoyed statistics and algebra.
Impossible? No. Improbable? Yes.
Again, I hope that I'm wrong. Humble Pie can be quite tasty, and was also a pretty good rock band.
GSC wrote:
Unspecified reasons? Look at the damned comment I made above (#54). I made my reasons clear.
You pointed to successful seasons Damphousse had at a later age than Gagner, and his consistency, and said you didn't see Gagner as Damphousse as a result.
We have no point of reference for Gagner over those years; we don't know what he'll do. So, yes, it would be wrong to say "Sam Gagner's career is pretty much exactly like Vinny Damphousse's, adjusting for era" but it's not wrong to say "Sam Gagner's first three seasons are pretty comparable to Vincent Damphousse, and it's possible he follows a similar career path."
You're ruling out future events based on things we don't know, and I don't think that's a solid reason.
As for the difference in height and body type, that's a realistic point, but while it's indicative I don't regard it as definitive.
Ball Buster wrote:
Sadly not. I wish I had a list of 20 players to compare to, but the criteria I have are rather restrictive.
Put another way, there aren't that many players who go top-10 in the NHL entry draft, make the NHL immediately, score like Gagner, don't rack up the PIM's, and do it all while standing under 6' tall. It's just not a big list.
We could eliminate Damphousse and Carpenter (both 6' or taller) and we'd be left with three comparables, but I made the judgement call not to do that and to keep the list at five. I also made the call not to jump Falloon's numbers up a year, which would have eliminated him from the list as well.
Those are actually the reasons I view Fox or Connolly as better comparables.
So he alterd stats to make an improbably, but not impossible outcome look possible.
Here is a novel idea.
Why compare players statistically from different eras at all?
There is no point really and I never really get people saying Seguin is comparable to Yzerman, or Hall is comparable to Messier.
Why not compare players of the same era if you need to do it at all?
In most of the advanced stat/number fudging I have seen, conformation bias is in full effect. This does more harm to the "field" than good.
For instance Horcoff was proclaimed as an outscorer and corsi warrior two years ago.
Fast forward to this season and you hear crickets about those stats and Horcoff. Now all of a sudden it is on ice sv% and relative corsi that is used to prop up the player.
Well were was that two years ago? Nowhere thats where. Stats get made up/pulled out of the closet when deemed necessary to bolster a preconceived idea and not to investigate who may or may not be effective on the ice.