One of the essential elements of the Glenn Anderson mythology is the belief that he relaxed in the games that didn’t matter and the blowouts. I don’t have a specific quote in mind but I remember hearing during Glennfest discussions about how, when the game was out of hand, guys like Gretzky would be looking to get back at the trough, while Anderson was thousands of miles away, thinking about something completely different. It’s always struck me as being a little fishy but it’s a big part of the Anderson story. I wanted to check it out, so I took a look at one of Anderson’s best years, the 1985-86 season.
I was able to identify 16 Oiler blowouts. My definition of blowout was generous - a four goal win. Here are the games I was able to find:
10/30/85 WIN 3 at EDM 7
11/03/85 TOR 1 at EDM 7
11/08/85 VAN 0 at EDM 13
12/03/85 EDM 8 at LAK 4
2/07/85 MNS 4 at EDM 8
12/20/85 LAK 4 at EDM 9
02/14/86 QUE 2 at EDM 8
02/19/86 TOR 5 at EDM 9
02/26/86 EDM 8 at WIN 2
03/04/86 EDM 6 at VAN 2
03/09/86 EDM 7 at LAK 3
03/14/86 DET 3 at EDM 12
03/18/86 WIN 2 at EDM 6
03/25/86 EDM 7 at DET 2
03/26/86 EDM 8 at PIT 3
04/02/86 VAN 4 at EDM 8
Here are the numbers for the Oilers Big Four forwards for those games (GP-G-A-P).
Gretzky 16 - 14 - 50 - 64
Kurri 16 - 25 - 19 - 44
Anderson 15 - 15 - 15 - 30
Messier 14 - 10 - 15 - 25
In the non blowout games, those same four did the following:
Gretzky - 64 - 38 - 113 - 151
Kurri - 62 - 43 - 44 - 87
Anderson - 67 - 39 - 33 - 72
Messier - 49 - 25 - 34 - 59
For a kind of apples to apples comparison, I put together the points/game numbers for blowouts and non-blowouts and then calculated the percentage increase in points/game rate during blowouts.
Gretzky: 4.0/2.36/69%
Kurri: 2.75/1.40/96%
Anderson: 2.0/1.15/70%
Messier: 1.79/1.2/49%
I wasn’t surprised to see that this doesn’t hold up particularly well. Now, the blowout sample is pretty small but Anderson is in the same ballpark as Gretzky, well ahead of Messier and clearly back of Kurri for blowout games in 1985-86. More seasons would need to be run but, in the 1985-86 season at least, this aspect of the Anderson legend doesn’t seem to hold up.
* * *
On the topic of Glenn Anderson, the thread I posted a few weeks back drew the following late comment:
If Ryan Smyth had the goal scoring skill Anderson had or if Anderson had the heart and sole Ryan Smyth has there wouldnt be a question about who is the better player.
As it is, GA needed motivation and big situations gave that to him. In those situations his drive was equal to RS’s on a constant basis. Consequently he scored big goals, because he had alot more goal scoring skill then Ryan Smyth, and I dont think its even comparable.
Ryan Smyth has little upside in those situations, mainly because he gives you everything he has all the time.
If you dont understand the above, you probably still dont understand why the Oilers didnt pay him the money.
Nothing’s bigger than the playoffs and that’s where Glenn Anderson made his bones, scoring 22 game winning goals. Smyth, to date, has only 4. When you consider the number of games that their teams won though…things narrow considerably. To date, Smyth has skated off the ice for the winning side on just 35 occasions. Anderson’s teams won 151 playoff games. If you calculate the rate of game winning goals/playoff win, you come up with 0.114 for Smyth and .146 for Anderson. Seems like a big difference…until you realize that with just one more game winning goal, Smyth would be at .143. There’s just not much to pick from there, particularly in light of the sample size.
It’s all fine and well that Anderson made the Hall of Fame and I’m not opposed to that - he was an Oiler, I cheer for the Oilers, so the more the merrier. His case is an interesting one thought because, for all of his numbers, it’s really an intangible case. Much as I’d like for the tangible case to be stronger, it seems to be very difficult to make one that doesn’t depend largely on “Count the ringz!!!!111!!” A lot of the other stuff just doesn’t hold up.
No, that’s not it. What makes Anderson great is that he scored key goals in key situations during Stanley Cup runs that ended with Stanley Cups. It’s a shame that Smyth didn’t have those opportunities as an Oiler.
Could have done and should have done just aren’t the same as done. The difference here is probably as simple as having teamates like Gretzky and Messier, but in the end if the count of cups are an important measure, then Anderson beats Smyth by a landslide.
Now if someone suggested Smyth wouldn’t be a hall of fame candidate today if you’d sent him back in time and subsituted him for Anderson on those 80’s teams, I might then call foul.
I think this is a little like the Shawn Horcoff debate, whether or not you get it is a tell about you as opposed to the player in question.
Years ago Vic compared his reg season scoring rates and his post season abilities and came up with a completely rational result.
Glenn Anderson’s ability to score big goals in big games has more to do with his being IN big games and performing at predictable levels.
Some of the greatest pitchers in history don’t have have a Sandy Koufax reputation because they didn’t get there. Just ask Juan Marichal.
I think you missed the point they were making. They said after a game was well in hand, Anderson would stop trying to pad stats. Your analysis is basically irrelevant to answering that question.
The relevant statistics to compile would be statistics based on the goal differential in the game. This of course is probably hard to do.
But if one is looking at blowouts, then one should be looking at the difference between Andersons statistics each period in those games.
As well, one should be looking at Andersons statistics in third periods of games decided by one goal vs. his statistics at other times.
Shorter #3: In order to disabuse me of this myth (which there is no good or even interesting reason for me to believe), you’ll have to do a lot better than that.
Nice one Tyler. Clearly some of your fellow fans don’t appreciate it, but this is conceptually valuable stuff, I think (same with the Fuhr stuff).
Break “Why did the Oiler win 4 Cups in 5 years?” into a pie chart, and most of the pie is the piece One Epic and Many Other Very Good Hockey Players. Everything else — coaching, clutchness, making the big save, walking past the Islander dressing room, etc. etc. — is fighting over the tiny remaining slice.
Godot -
If it was true, does it not strike you as bizarre that Andy’s points spiked by the same percentage as Gretzky’s in those blowouts? It seems to me that, if the myth is true about Andy in 1985-86, Gretzky must have been doing little when the game was close.
I can and will check that out; I’m going to do a couple of other seasons for which the data is available too. It seems to me though, that a far more likely explanation for Gretzky seeming to pile up the points in the blowouts is the same reason he seemed to get more points than Andy in the close games: he was a much better hockey player.
11/08/85 VAN 0 at EDM 13
Good lord.
The relevant statistics to compile would be statistics based on the goal differential in the game. This of course is probably hard to do… As well, one should be looking at Andersons statistics in third periods of games decided by one goal vs. his statistics at other times.
I think that’s the obvious rebuttal and it’s a fair point. I’m not sure there’s going to be a sufficient enough sample size, even over the course of his career, to reach any definite conclusions on the issue, except for one that can reasonably be drawn: if there is a “meaningless game equals Anderson’s head in the clouds effect,” a big if at this point, it is almost certainly overstated by the media and the desire for a convenient narrative. I know Gretzky talked about this supposed Anderson trait in his autobio, but I don’t think he’s untouched by the narrative.
The times I watched Anderson play, especially before he got old, I thought he was a damn good player. I thought he was the only forward not named Gretzky or Lemieux that had any business being on the ice against the Soviets in the 87 Canada Cup, though that admittedly might speak to style as well as effectiveness.
He scored big goals, but he also scored beautiful goals. Gretzky talked about this in his autobio too (it’s funny what you remember reading at 12), how Anderson scored a silly pretty goal on an until-then unbeatable Billy Smith in their first Cup win, with Gretzky claiming it was a back-breaking goal for the Islanders and opened the flood gates for the rest of the series, etc. The stat crew is going to HATE this, and I wouldn’t want to overemphasize the point, but I think there is a psychologically difference in effect when someone dances through an entire team and roofs it versus a some-weird-shit-happened-in-front-and-whadya-know-the-puck’s-in-the-net goal. A recent example of this would be the Backstrom-Semin passing drill goal against the Flyers in the playoffs last year. Any goal at a crucial time in the game changes momentum, but I think that ridiculous display of skill did lift Washington’s tail and hang Philly’s heads more than your average goal. And rightfully so, that was just an ungodly display of skill. If you’re Joffrey Lupul or Scottie Upshall, how are you not at least subconsciously thinking, “I can’t play with these guys.” Of course Washington lost the series, so it’s not hard to see how slight the effect is (though they won that particular game).
It’s more than the media who will tell you Anderson often went for a skate late in blowouts. I attended many of said blowouts, and it was pretty obvious a lot of nights. Andy would put it in cruise control, esp. on the defensive side of the puck. Some of those nights it was impossible not to score anyway. Messier was the same way, although occasionally he would use garbage time to, uh, “take out” the garbage.
For all your good work here Tyler, Godot points out the flaw in the study: it doesn’t account for game situation. Also, it by definition omits games like this one.
But as for running up the points in blowouts, it’s simple arithmetic:
16 blowouts: 131 goals, 8.19 (!!) GPG
64 others: 295 goals, 4.61 GPG
… meaning the team scored 77% more in those games. So Anderson and even Gretzky scored a slightly smaller share of the points in those games.
Years ago Vic compared his reg season scoring rates and his post season abilities and came up with a completely rational result.
I’d be interested to see that. I know I did my own analysis on those scoring rates specific to the Big Five’s careers in Edmonton. As an Oiler Anderson’s points-per-game soared from 1.07 in the regular season to 1.22 in the playoffs, despite the fact the Oilers as a team saw their scoring rates decline by about 0.3 G/G during his years on the club. Put another way, Andy was involved in about 23% of Oiler goals in the regular season and 28% in the playoffs. He really did rise to the occasion.
//If it was true, does it not strike you as bizarre that Andy’s points spiked by the same percentage as Gretzky’s in those blowouts? It seems to me that, if the myth is true about Andy in 1985-86, Gretzky must have been doing little when the game was close.//
Maybe because Anderson got his points early in those blowout games, whereas Gretzky’s were evenly distributed.
Gretzky’s point was that Anderson scored when it mattered.
1) In the playoffs (easy to check again and prove, see oildroppings)
2) His stats should be better when the game is close.
3) His stats should be better in the first period of blowouts than in the third period of blowouts.
The analysis MS gives is interesting, but irrelevant to proving or disproving Gretzky’s assertion.
You know, it gets really, really tiresome to try to prove again and again that there’s no such thing as the big goal scorer. Kinda like trying to have a discussion about dark energy and starting with proving the world is a sphere all the time.
Let me try an Austrian, as opposed to econometric, approach. What would we have to believe in order for the “big goal scorer” to exist?
1. A skillful player, who has employed his skills his entire life, deliberately underperforms when faced with little competition. And in fact underperforms greater than the decrease in the level of his competition.
2. A skillful player overperforms his level of competition when faced with a “mentally tough” situation.
3. The source of this overperformance is a positive force field that emanates from his head and scares his fellow professionals.
4. The source of underperformance is a French-style insouciance and Zen-like inability to be motivated by extrinsic rewards (such as contract bonuses).
Seems like a stretch.
If Gretzky said it, it must be True.
There are professional gamblers. The argument for a big goal scorer is the samw argument for professional gamblers.
Professional gamblers are successful because they know when to bet big and know when to bet small and know when not to bet.
A big goal scorer knows when to exert extra effort, and when to exert less effort, and when to exert no effort.
There’s a religion in this thread, swear to God. Glenn Anderson’s scoring rates in the postseason are well within the range of random and in fact this is basically true in all sports. Bill James fell in love with “Game winning RBI” and “Victory important rbi” and Eddie Murray was a GOD by this metric right before it turned around on him and he wandered back to the median.
LT: Well within the range of random? I beg to differ. Come playoff time scoring outputs generally drop off, yet during his time with the Oilers Anderson’s PPG soared by 0.15 PPG, or over 13%. That’s a pretty extreme outlier.
I’ll bet you can’t find five guys in league history who have, say, 50 playoff games who have seen a similar bump in scoring. Off the top of my head, I can’t come up with even one.
Do you really think Anderson’s annual post season hot streaks were just random? He dialled up the intensity and the focus much more than most and the production shows that. It was obvious at the time and it is still obvious in the statistical record.
And I’m not a religious man.
Come playoff time scoring outputs generally drop off, yet during his time with the Oilers Anderson’s PPG soared by 0.15 PPG, or over 13%. That’s a pretty extreme outlier.
I’ll bet you can’t find five guys in league history who have, say, 50 playoff games who have seen a similar bump in scoring. Off the top of my head, I can’t come up with even one.
Can I cut out their bad years so as to skew their numbers like you did? Anderson for his career was 0.97 PPG in the regular season and 0.95 PPG in the playoffs. That’s not that impressive - here are the alltime PPG leaders in the playoffs, you can click through lots of them and see that Anderson’s feat was accomplished by many.
Also, according to H-ref, Andy put up 1.12 PPG in the playoffs as an Oiler and 1.07 PPG in the regular season. Again, while that’s notable, it’s not that special.
//Also, according to H-ref, Andy put up 1.12 PPG in the playoffs as an Oiler and 1.07 PPG in the regular season. Again, while that’s notable, it’s not that special.//
The vast majority of players PPG goes down in the playoffs.
The average goals per game goes down in the playoffs.
Does anyone know of a website which has the yearly NHL scoring race from the present to the seventies, including at least the top 150 players?
Ditto for the playoffs?
Most sites like hockeydb.com only have the top 20 or so.
I’m reasonably certain that Anderson’s playoff performance as an Oiler is a black swan event (or at least highly anomalous), and can be demonstrated to be so by statistical analysis fairly conclusively.
I may make it my mini-project of the spring.
It’s possible that some players could be better playoff scorers than others without having to resort to clutchiness (clutchness?) as the explanation.
Scoring drops in the playoffs. It’s a well known fact. Some of that can be explained by:
- a drop in the proportion of special teams play
- goons being healthy scratches
- the involvement of only the best 16 teams *
If these don’t completely explain away the scoring drop in the postseason, we’re left with a change in the way the game is being played. Most would probably agree that they see a difference between playoff and regular season play in the NHL - call it ‘tighter checking,’ ‘more intense,’ whatever you like. If the change in style of play is real, isn’t it within the realm of possibility that some players cope with it better than others?
This is all minuscule stuff, impossible to nail down with any certainty. That’s why I just assume that good Irish-Canadian boys from Kingston are playoff money.
* - Taking the best 16 out of 21 isn’t quite the same as taking the best 16 out of 30. If this is a real effect, the Islanders/Oilers players in the 80s might see a smaller drop in playoff ppg than players today.
I have done some work on the 1980s Oilers and playing to the score effects, so I happen to have third period numbers for both Gretzky and Anderson for all playoff games between 1981 and 1987 broken down by the game score after the first 2 periods.
In third periods that started with the Oilers leading by 3 goals or more (19 periods):
Wayne Gretzky: 4 goals, 14 assists, 18 points
Glenn Anderson: 3 goals, 2 assists, 5 points
In third periods that started tied or with a one goal differential (49 periods):
Wayne Gretzky: 10 goals, 20 assists, 30 points
Glenn Anderson: 8 goals, 15 assists, 23 points
I think that is compelling evidence that Anderson did score a lot more in close games than in blowouts, while Gretzky scored a lot regardless of game situation.
The reasons probably have very little at all to do with clutch play or effort (”lack of effort” strikes me as the catch-all justification for everything that goes wrong in hockey). It likely has a lot to do with playing to the score. The Oilers probably shortened their bench and took a lot more offensive risks when the game was close, and tried to shut it down when they were well ahead.
What kind of opposition did Messier’s line face early in the game? Is there reason to believe that would change late in the game? Gretzky was presumably double-shifted in important situations, did he play more with Anderson late in close games? What kind of ice time were the Oilers’ third and fourth lines getting in blowouts compared to in close games or OT?
I’m as skeptical as anyone else about clutch play, but I think there are lots of reasons why players could score at a higher rate in the playoffs or in other so-called “clutch” situations.
Also, according to H-ref, Andy put up 1.12 PPG in the playoffs as an Oiler and 1.07 PPG in the regular season. Again, while that’s notable, it’s not that special.
. So what can I say but Mea culpa (Latin, meaning “I fucked up.”)
Let’s start there. You are right. My figure of 1.22 was in error. A stupid formatting error in Excel which I didn’t check thoroughly enough. Computers don’t make mistakes, right? Maybe not, but humans sure do, including this one (on rare occasions
So certainly that weakens my claim about Anderson’s outperformance in the playoffs, but it still doesn’t negate it.
Can I cut out their bad years so as to skew their numbers like you did?
One man’s cherry picking is another man’s research. Especially when one is considering “bad years” which are typically 70+ GP in the regular season but can range from 0 to 20+ in the postseason. I chose to focus on the Big Five’s Edmonton years not to skew their numbers but to to remove the skew of regular: playoff GP ratios which happened in their subsequent career (more on this below). I figured a decade in one place, playing 150+ playoff games apiece was more than an adequate sample size. Besides, as an Oilers fan I am far most interested in their performance as Oilers, where each of them started their careers and compiled the huge majority of their Hall of Fame credentials.
Anderson for his career was 0.97 PPG in the regular season and 0.95 PPG in the playoffs.
Right you are. I cited those exact same numbers in my post. However, I then explained how they were skewed by GP effects. Let me try again …
That’s not that impressive - here are the all-time PPG leaders in the playoffs, you can click through lots of them and see that Anderson’s feat was accomplished by many.
Well I clicked through the top dozen and found that on average their scoring rates dropped by 6.2%, which coincidentally is exactly the drop-off in scoring from regular to postseason on the Oilers during Anderson’s time there; don’t know what the league-wide average is in that respect but a few percent less scoring in the playoffs is certainly consistent with what I have observed and tracked over the years. No doubt somebody has exact stats somewhere, but if it falls much beyond the range of 5-8% lower in the playoffs I’ll be surprised.
I found exactly two guys in that top 12 whose career PPG rates were as much as +0.05 better in the playoffs, as Anderson’s was with the Oilers. Here you go:
Mark Messier 1.07 –> 1.25
Jari Kurri 1.12 –> 1.17
Kurri is a good case in point as to GP distribution effects.
Edmonton reg.season: 754 GP, 1.38 PPG
Edmonton postseason: 146 GP, 1.38 PPG
Elsewhere reg.season: 497 GP, 0.71 PPG
Elsewhere postseason: 54 GP, 0.57 PPG
Career reg.season: 1251 GP, 1.12 PPG
Career postseason: 200 GP, 1.17 PPG
So if you accept the career PPG rates at face value, it appears that Kurri performed at a higher level in the playoffs, whereas in reality in the two distinct segments of his career he was at and below par respectively. What has skewed his career PPG rates is the distribution of GP: Kurri played 40% of his regular season games after he left Edmonton, but just 27% of his playoff games. Similar skews exist for Anderson and especially Messier, but in opposite directions. Let’s break down the careers for all three superstars into three parts (Edmonton, second team, playing out the string) as career rates really plummet in that last part for all three as their skills eroded and they entered the Dead Puck Era:
Jari Kurri
Edmonton reg.season (1980-90): 754 GP (60%), 1.38 PPG
Edmonton postseason (1980-90): 146 GP (73%), 1.38 PPG
Los Angeles reg.season (1991-95): 331 GP (26%), 0.89 PPG
Los Angeles postseason (1991-95): 28 GP (14%), 0.71 PPG
NYR-ANA-COL reg.season (1995-98): 166 GP (13%), 0.37 PPG
NYR-ANA-COL postseason (1995-98): 26 GP (13%), 0.42 PPG
***
Mark Messier
Edmonton reg.season (1979-91): 851 GP (48%), 1.22 PPG
Edmonton postseason (1979-91): 215 GP (73%), 1.30 PPG
New York* reg.season (1991-97): 421 GP (24%), 1.23 PPG
New York* postseason (1991-97): 80 GP (27%), 1.14 PPG
(* first go-round)
VAN-NYR** reg.season (1997-2004): 484 GP (28%), 0.69 PPG
VAN-NYR** post-season (1997-2004): 0 GP (0%), — ppg
(** second go-round)
***
Glenn Anderson
Edmonton reg.season (1980-91, ‘97): 845 GP (75%), 1.07 PPG
Edmonton postseason (1980-91): 164 GP (73%), 1.12 PPG
Toronto reg.season (1991-94): 221 GP (20%), 0.71 PPG
Toronto postseason (1991-93): 21 GP (9%), 0.86 PPG
NYR-STL reg.season (1994-1996): 63 GP (6%), 0.57 PPG
NYR-STL post-season (1994-1996): 40 GP (18%), 0.51 PPG
… showing that Anderson was a playoff outscorer for most of his career but his rates were dragged down at the very end when he played a disproportionately high number of playoff games. Whereas Kurri and especially Messier benefit from the opposite effect.
Which is a very long-winded way of making my main point, which is that is dangerous to compare career regular season to playoff rates. It’s great as a first test, but one needs to go deeper.
Goddamit, I’ve done it again! Besides the usual misplacement of italics that happens all too often in the no-preview no-edit realm, I made a small statistical mistake in the above. The top 12 playoff scorers have an average reduced production of .062 PPG in the playoffs, not 6.2%. Expressed in percentage terms, the typical drop-off of is a shade under 5%. Doesn’t change the conclusion that scoring drops a little in the postseason, but I am a slave to accurate numbers first, even at the cost of some small embarrassment of admitting another screw-up.
Sigh. As if I needed more proof of how skills decline with age.
1. A skillful player, who has employed his skills his entire life, deliberately underperforms when faced with little competition. And in fact underperforms greater than the decrease in the level of his competition.
2. A skillful player overperforms his level of competition when faced with a “mentally tough” situation.
3. The source of this overperformance is a positive force field that emanates from his head and scares his fellow professionals.
4. The source of underperformance is a French-style insouciance and Zen-like inability to be motivated by extrinsic rewards (such as contract bonuses).
Seems like a stretch.
Then you and I are clearly going to disagree on this, because I can perfectly believe that (a) a motivated player will outplay an unmotivated player every damned time, and (b) the same guy can be more motivated in a close game and less motivated in a blowout. Neither of these strike me individually as alien concepts, and it furthermore seems reasonable to me that ’80s Gretzky was the sort to never stop going while ’80s Anderson stopped giving a shit when the GD hit 4, though that could be the mythology talking.
I guess it’s that Goddamned sports psychology that can’t be quantified rearing its ugly head again.
Also, this seems like the sort of thing that, while it will show up in the stats if you approach them the right way, the simplest thing to do is simply to ask someone who was there. I’m not inclined to believe that Bruce is making shit up when he says that Andy tended to treat the third period of a 4+-goal game like a free public skate.
I was there, too, and I remember Anderson taking a lot of stupid, gawdawful penalties in close games. I don’t want to go so far as to say he’s a Rob Brown with better luck, but the team carried him more than he carried the team.
We are going to disagree, because motivation shows up in numbers. All numbers. The very definition of motivation is someone who performs when he doesn’t have to. It’s a habit, not a state of being. Think about the myth of the big goal scorer applied to any other profession, such as medicine. “I left that sponge in there because it was an easy operation.” That’s what we call a terrible surgeon.
I’m very, very good at my job. Being good at my job requires years of training (mostly in math), years of practice, good work habits, attention to detail, and persistence. Similar habits that are required in any professional endevour. It beggers reasoning that somebody can consistently turn these habits on and off.