Diverging Realities Part Two: The Sharks Fan Base, Jumbo And Patty, And Why The Playoffs Matter (Really! They do!)
Hey, Shark Circle here. Part One of this blog-series focused on what I mean by Diverging Realities, what caused them, and how the phenomenon has divided the hockey world at large. Here is Part Two, which will focus on the effect within the Sharks fan base, in particular the debates surrounding Joe Thornton, Patrick Marleau, and the question of whether players are responsible for their performances in the playoffs. If you haven’t read Part One yet, I recommend you start with it first, otherwise Part Two will be very difficult to understand. Please enjoy!
I spent the last blog detailing how realities have diverged among fans across the NHL, but that doesn’t mean the phenomenon isn’t at its worst, or close to it, in San Jose. Nowhere have I noticed the divide between the people who try to understand as much of reality as they can, and the people who inadvertently create their own through mathematical miscalculations and manipulation**, more than within the Sharks fan base. Specifically, the debate surrounding Joe Thornton and Patrick Marleau’s playoff history. The advanced stats community (aka #fancystats on twitter) has always discounted the entire idea of “performance” and self-determinism in the playoffs, so it’s no wonder that a discussion surrounding great regular-season performancers who decline in the playoffs completely confounds them. And for the first time since these fans have risen to
prominence arrogance within the Sharks community, they are greatly displeased with the way things are going, what with Sharks GM Doug Wilson saying he’s moving towards major decisions they don’t agree with. And as we all know, it’s not until adversity hits that you necessarily see all of people’s true colors. And this group of fans has turned on Doug Wilson very aggressively since he stated his intention to change the direction he takes this franchise in.
**If you doubt any of my claims here about statistics being miscalculated, that blog will provide proof, which is definitely an important thing to have before you believe something. Cough.
Notice I don’t say “some of them” have turned on him, or “some of them” believe something. That’s because typically one or two people set the accepted opinion for this group on any given issue, usually a blogger (a bit like me, only way more enlightened, of course), and from there it just becomes a parade of group-reinforcement. Dissenting opinions aren’t disagreed with or debated, they are corrected, usually with statistical terminology like “statistical noise” or “insufficient sample size.” Those terms especially are mainstays among this community used to essentially “white out” any dissenting arguments or even data (the two go hand-in-hand with this group). Example:
Here’s my data, here’s why I’m right. But what’s your data? …Oh, that data doesn’t count. That’s just statistical variance/noise/a small sample size… Well, I guess that only leaves my data now. Looks like I’m right again!!!
Hard to argue with that, isn’t it? And what’s the latest inarguable group-opinion from these fans? In case you didn’t believe me when I said they’d turned on Doug Wilson, here’s a recent article posted on Fear The Fin titled F*ck the San Jose Sharks. It’s a pretty a funny read; after all, the sarcastic rants of frustrated fans usually do hold some irreverent entertainment value. But the message it means to send is a very clear one of disapproval (says Sherlock). Here are a few quotes from the blog.
1. “Being a Sharks fan blows.”
2. “They fired Drew Remenda.”
3. “the Sharks decided to fire a beloved broadcaster as they attempt to blow up the best team that has ever played hockey in San Jose.”
4. “The man we’ve (almost) universally lauded over the past several years decided the best way forward is without a couple of the team’s best forwards. That sucks.”
5. “… it fucking sucks being a fan … today.”
So being a Sharks fan sucks right now. Message received. And I don’t think anyone would argue with that after what happened with the Sharks recently…
I mean they fired Drew Remenda! Can you believe it! They actually fired Drew Remenda! Just when everything was going right with the Sharks coming off such a memorable and rare feat as blowing a 3-0 lead in a playoff series, the Sharks went and ruined all the good vibes by firing Drew Remenda!
Okay, so that’s not exactly sarcasm, as I really liked Remenda when he wasn’t sugar-coating what was happening with the Sharks for fear of getting fired from his job… damn… but here’s what I’m getting at: the fact that this team collapsed in epic fashion against the Kings so recently is not even listed in the article as one of the main reasons why ‘the Sharks should be f*cked.’ Isn’t that a bit strange? What is mentioned, besides Drew Remenda’s firing, is how much it sucks, and essentially how crazy it is to this group overall, that Doug Wilson would even dare consider trading the aging players who just led the team in completing possibly the worst playoff collapse in NHL history.
And yet the same blog also mentions how these fans have “(almost) universally lauded” Doug Wilson over the past few years. But if he’s so crazy, and such an idiot as many in this group have called him lately, then why did these fans “(almost) universally laud” him until now? Or did he just go from awesome, “clever” GM to totally insane idiot overnight? Because I just checked the #fancystats probability ratios on that, and they’re telling me the likelihood of someone going from an awesome GM to a crazy one overnight is very low. And even when it does happen, it still falls within the expected statistical outliers.
Yeah, so applying the statistical arguments they make to real world situations can be confusing. That’s the problem. None of it makes any sense, does it? What we have is the segment of the fan base that loved Doug Wilson the most just a few months ago now criticizing him more harshly than anyone. Fear The Fin, a blog which has consistently praised Doug Wilson’s drafting record over the years, just used the extension of Roy Sommer as a reason why Wilson isn’t really looking to the future like he claims to be, because Sommer’s history developing talent is “spotty at best”. So the Sharks have been good at drafting under Wilson (and Sommer, who has coached Worchester for Wilson’s entire tenure), but bad at developing those draft picks? How exactly do you make that distinction from the outside? Or without a time machine, for that matter? You can’t. That’s just more dressed-up contradiction, criticizing Wilson in the same areas they used to praise him, but just changing the names around as a way to make sense of their own emotionally-driven, completely hypocritical logic.
But that is what happens when you use a different method to view (or calculate) hockey than everyone else: your conclusions will differ, too, sometimes to the extent where decisions that seem obvious to everyone else, or seem to at least have logical explanations, make absolutely no sense to you. In this case, your team’s GM, Doug Wilson, finally decides to side with what has been consensus around the NHL for years, that Thornton and Marleau are lacking something when it comes to leading their team to a championship, and yet even though many prominent NHL-insiders and even teammates, in addition to Doug Wilson, have reported seeing or hearing about that problem first hand, and yet even though a majority of hockey fans believe they’ve seen it manifest on the ice as well, you’re completely befuddled. How could Wilson do it? It’s so crazy! It’s so far fetched! Where did he even get this idea?
When in reality, there is logic behind it, even if you just distill it down to the most basic idea that Doug Wilson wants to move on from two players well into their thirties who haven’t been able to get the job done in nearly a decade, one of whom was called “gutless” by a former teammate. And while there are certainly legitimate reasons to question Doug Wlson’s decision-making this offseason, like the recent contract given to Mike Brown, the fact is he’s made signings like that, or like John Scott, his entire tenure with San Jose. It’s just that back then, this group of fans was too busy lauding him to care. So let’s not pretend Mike Brown and John Scott are the reason this group has experienced a sudden change of heart towards Doug Wilson, because that type of signing is nothing new with him (Jody Shelley, anyone?). What’s changed is that after a decade of playoff disappointment, Doug Wilson has finally decided to stop closing his eyes to the problems with his star players just because they’re his star players. He’s finally decided to hold the leaders of the team accountable for their results, like you would do in any other business under comparable circumstances. But these #fancystats Sharks fans can’t even fathom that logic because, in their version of reality, individual playoff performances don’t even count in terms of being self-determined by the players themselves, so there’s nothing for Doug Wilson to even hold the players responsible for.
No wonder this group is so perplexed by anyone critical of Thornton and Marleau. I’d think Thornton and Marleau were perfect too if I could completely discount their playoff achievements and just look at their regular season success.
But in a sport where the ultimate goal is to win a Championship, in the playoffs, discounting playoff achievements would be ridiculous. Unfortunately, that fact hasn’t stopped these #fancstats Sharks fans from doing it. They’ve concocted a fallacy through an amateur misunderstanding of statistics, specifically how sample size and statistical variance work, and their belief in it is the main reason they think the concept of the Sharks moving on from Thornton and Marleau is insane.
But really, it isn’t. This group of fans has just been telling each other for so long that the earth is flat, that now that Doug Wilson comes out and says he agrees with everyone else that it’s round, they think he’s crazy. And this same phenomenon is repeating itself on nearly any subject that comes up within the Sharks fan base. Instead of normal disagreements and debates, you have a bunch of people who think the earth is flat, and won’t listen to anyone who says otherwise, and a bunch of people who think the earth is round, but have given up trying to convince the flat-earthers. Resentment has replaced discussion, and it wouldn’t surprise me if other fan bases were also experiencing this same type of non-debate on various issues that come up about their teams. Just not to this extent. Now, advanced stats pledges would argue over which side are the “flat-earthers,” but widespread fallacies among their community, like the one I just detailed, are proof of which side they represent in that analogy, just like how believing the earth is flat is a fallacy that can be scientifically disproved.
Maybe analogies do more to complicate the point than crystallize it, though. Maybe we’d better served to just put it this way: they are the fallacy side, or false-reality side (in the context of hockey). Why are they the fallacy side? Because they believe in fallacies, one of which I just detailed above. Discounting playoff performance (discounting in this case means to zero) because there are less playoff games than regular season games is a fallacy. If a team plays four-times as many regular season games as playoff games, that’s a four-to-one ratio, not four-to-zero. So right there you have a basic mistake, the type that exemplifies why you need to be qualified if you’re going to publish statistical analysis. They could at least weight the statistics in that way (playoff performance for this hypothetical team gets one-fourth the value of regular season performance, as opposed to zero value, due to the four-to-one ratio in sample size). And even then they wouldn’t be addressing the fact that the playoffs are more important than the regular season, which presents another weighting-dilemma in the opposite direction, one that is impossible to quantify to an exact number but still needs to be accounted for (which they do not do), or that the quality of competition increases in the playoffs, which can affect performance, meaning that not all performance fluctuations from regular season to postseason are “luck”, “regression,” or “chance,” which is how they always characterize it because they don’t have the expertise to know how to account for it or quantify it.
You see, if every player reacted the same to improved competition, then teams could just draft the junior players with the most goals or points, or sign the players from the KHL with the most points, and expect those players to also have the most points in the NHL out of their groups. Teams wouldn’t have to look at anything else. But drafting, and especially signing NHL-aged players from overseas (which is a better example because that eliminates the “development” variable), shows us that some styles adapt to improved competition better than others.
For instance, Nick Ritchie was drafted 10th overall by the Anaheim Ducks, while plenty of other junior players with comparable statistics had to wait much longer to be drafted. Why? Because Nick Ritchie is 6’2″, 230 lbs. He has an NHL body, while many of those other players ranged from 5’6″ to 5’11”. There are always exceptions, but scouts understand that an 80-point junior player who has an elite NHL body is probably going to have an easier time transferring that production to the NHL than an 80-point junior player who is 5’6″. So while the production of two players’ like that might be equal in junior, scouts understand that the way they produce in junior, against a lower level of competition, can effect how well that production carries to the NHL, against a greater level of competition.
Well, when you go from the regular season to the postseason, you are jumping from a lesser level of competition to a greater level of competition. The gap isn’t as big as jumping from junior hockey to the NHL, but it’s the same exact principle that applies. All of this has to be accounted for when analyzing postseason production and success, and deciding how much weight to give playoff analysis. The advanced stats community weighs none of these variables, which has resulted in the creation of, and belief in, a complete fallacy within their community. A complete misinterpretation of reality in hockey. I don’t know how else to say it. And this is just one of their fallacies. Don’t even get me started on “PDO regresses to 1,000,” a “law” they’ve based on a fundamental misunderstanding of what “regress” means
Besides the obvious (that spreading false information and analysis is bad), you might be wondering what other consequences all their fallacies have had on the Sharks fan base, particularly the playoffs fallacy I just detailed. Well, as harsh as the backlash has been against Doug Wilson this offseason among this group of Sharks fans, they’ve reserved their greatest wrath for anyone who questions Thornton or Marleau’s championship mettle, or however you want to put it. After all, there wouldn’t even be a “Doug Wilson backlash” if he hadn’t essentially done just that himself. You see, the advanced stats segment of Sharks fans have been so aggressive for years insulting anyone who dared suggest there might be some merit to the criticisms surrounding Thornton and Marleau’s lack of success in the playoffs with the Sharks, that now that Doug Wilson has seemingly substantiated all those criticisms about the team’s culture (with quotes like “we want players who want to play here, not just live here,” and allusions to the team being more like co-workers than teammates), this segment of the fan base has had to make a choice: either admit they were wrong about Thornton and Marleau, as Doug Wilson essentially has now (to his credit), and that maybe the Jeremy Roenick’s and Mike Milbury’s of the world, who the #fancystats community so detests, were onto something, or simply throw Doug Wilson’s good, lauded name in with all the other crazy idiots like Roenick, Milbury, and most hockey fans with eyes.
It’s unfortunate, but it seems these fans have chosen the latter. They wouldn’t consider anything Jeremy Roenick had to say about Patrick Marleau years ago, despite the basic logic involved that maybe he might have seen something from playing with the guy that fans on the outside couldn’t, so why would Doug Wilson’s first-hand knowledge from inside the organization hold any value, either? If you thought they’d finally admit, at the very least, that you can’t know what you can’t know, and maybe give Doug Wilson the benefit of the doubt just because he’d been their leader and General in the defense of Thornton and Marleau for so long, you were wrong. Their emotional investment in Thornton and Marleau, and the fallacies that they themselves concocted around those two players specifically to defend them all these years, trump all. Now Doug Wilson is under the bus with the everyone else who can look past their statistical fallacy into reality. That doesn’t mean there isn’t still room for debate in reality, but you have to get there first, instead of a believing in a statistical fallacy that never gives you the chance to think for yourself or come to any other conclusions in the first place.
But they do so it doesn’t so they can’t, and as a result, Doug Wilson has gone from a favorite among advanced stats Sharks fans, and someone who shared most of their ideals, to an idiot. Just like that. You deserve better, Doug Wilson! At least five-percent! But there’s just no way Doug Wilson has really changed as drastically overnight as this group’s opinion of him has. These people were either wrong about him before, or they are wrong about him now. And both times, they just knew they were right.
Now, I’m no Doug Wilson-cheerleader myself. In fact this group of the Sharks fan base was a lot higher on him just a few months ago than I have been at any time during his tenure as Sharks’ GM. And hey, I’m very encouraging of people changing their minds when presented with good reason. When someone’s understanding of a subject evolves from being wrong to being right based on new evidence, that’s a great and, unfortunately, increasingly rare, thing. But that’s not the same as characterizing the same man in such opposite ways over the course of such a small period of time in such a manic way. When you turn against someone on the basis of statistical fallacies and totally hypocritical, bias logic (i.e. Doug Wilson is good at drafting when you like him, but bad at “developing draft picks” when you don’t), that speaks more to the group itself than to the subject they’re changing their opinion about. You can disagree with Doug Wilson’s statements this summer all you want, but if you were almost universally praising him for the entire decade prior, then there’s no way you should be calling him an idiot, or crazy, based on one summer of statements–not even moves, but statements of intended moves, that you disagree with, especially when your basis for disagreeing with his intentions is a complete fallacy. Not to mention he’s the one who traded for Joe Thornton in the first place. He’s the one who’s kept him and Marleau on the Sharks through his prime years despite heavy criticism. But now he’s an idiot for wanting to trade two 34-year-olds? That doesn’t make sense.
No, if you’re having such strong, stubborn shifts in opinion like that, it speaks to something that’s going on with you, and not, in this case, something going on with Doug Wilson or his sanity. You overlooked his Jody Shelley type signings in the past, but now you focus on them with a microscope. That’s a change in you, not a change in him. My opinion about the reasons for this change what the reasons are for this change, both around the NHL (in Part One) and within the Sharks’ fan base, have already been well-documented, but more than anything, this speaks to a certain segment of the fan base that has been defending Joe Thornton and Patrick Marleau for so long that they’ve developed an extreme bias towards those players, to the point where they care more about being proven right about those two players than they do about the Sharks’ future success independent of Patty and Big Joe, and they will turn on anyone who suggests trading those players using any reason they can.
And that speaks to the issue that arises when passionate sports fans attempt complex statistics. Bias becomes a huge issue. It’s hard to analyze something objectively when you’re not the least bit objective about what you’re analyzing, even having statistics at your disposal can just as often inflame that bias as they can act as an intermediary. And so every nuance within a statistical study just becomes a kind of open space that these bloggers can let their biases seep into without most readers noticing it, often without themselves knowing it either.
But what about when their stats are right? You might ask. Surely they can’t always be miscalculated? Even twelve monkeys can write a Shakespeare twice a day on the big hand, after all, isn’t that so? And that’s a decent enough question to help me segway to the next paragraph. But no one is saying every point someone makes is wrong. I mean even twelve monkeys can write a… I don’t want to type it all again. But I’m just trying to outline the divide that has emerged among NHL fans on the whole, despite the fact that few people are always wrong about every word they say, and to explain the down side of what can happen when these stats are miscalculated or manipulated.
And often the stats themselves don’t even have to be manipulated. All that has to be manipulated is which ones are shown, or, on a subconscious level, which ones your mind is disposed to even look for in the first place when doing a “study.” For instance, even with the most simple analysis, like the Jody Shelley and Mike Brown signings, you might get the studies themselves done right, and you might be able to report, correctly, when each respective signing takes place that they’re all bad moves, but as we’ve just seen, bias can still come into the play when you’re deciding how much weight to give those bad signings. It’s “yeah Doug Wilson acquired Jody Shelley and Colin White, and I don’t like the signings, but he’s still a good GM who makes far more good moves than bad ones,“ verses “Doug Wilson signed Mike Brown and John Scott, and I don’t like the signings, and they prove he’s a bad GM.“
Or take this Fear the Fin blog, for example, titled… well I can’t find it and I don’t remember the title, but hey I’ll link to them later. What I can tell you, though, is that it has become probably the most referenced blog for Marleau-defenders because it contains every pro-Marleau sounding playoff statistic you can think of. And the people who reference it, just as the person who wrote it, that the data proves Marleau doesn’t underachieve in the playoffs. That it’s an objective fact. That the statistics in that blog are irrefutable.
It’s like they’ve never heard the famous line, “Lies. Damned lies. And statistics.” Or they don’t understand that it’s not trying to say that each individual statistic is a lie. It’s referring to selection, presentation, manipulation, and bias.
Because even if every statistic* and statistical calculation** in that blog were true, it was still meaningless and wrought with bias. I don’t even have to read it again because I knew it the moment I read it the first time. How can that be if the stats cited were all true (IF they were? Because the author’s bias was obvious. The blog was written by a huge Patrick Marleau fan, who was a bit emotionally invested in Marleau to start with simply from rooting him on for years but then, it seems, became extremely self-invested in him once he made it a major goal to prove his own reality of Patrick Marleau to all the idiots who saw it differently, and as a result this fan who had developed a clear bias towards his subject sat down at his computer and, before studying or investigating anything at all in any honest or objective attempt to the find the truth, typed out the title to the blog he was going to write no matter what the stats said:
Now would be such a great time to remember what that damn blog was called.
Anyway, he typed out the title to his blog, Why Patrick Marleau is so great and the statistics show he’s actually super great in the playoffs even though it always looks like he sucks and the playoffs aren’t self-determined anyway except for when I can make them sound favorable or whatever it was (I’m bad with titles), before he even looked up the statistics. You could tell just from reading the blog.
Now, I guess this one’s just my opinion, especially since I can’t even find the blog for you to decide for yourself, but I don’t think he even cared what the stats were going to say. He seemed to have his mind made up before he even did his research for the blog. Instead of going from data to conclusion, and letting the data dictate the conclusion, he went into the blog with the conclusion already decided in his mind, and the only data he was interested in finding, or posting in his blog, were the stats that made Patrick Marleau sound good. Patrick Marleau has scored the most playoff goals in the NHL since (whenever). Nevermind that he’s also played more playoff games than almost anyone else since (whenever), too.
• I’m sure they were
** I’d be much less sure if there were any complex ones in that one, and I don’t remember
It would be just as easy for someone who wanted to paint a negative picture of Marleau’s playoff history to rank his points-per-game average in the playoffs since the start of his career against everyone else in that time period, and then you’d have a very negative statistical one-liner to throw around about Marleau.
In reality, the stat that combines these two measures would actually be the most accurate, where you combine career points-per-game (or goals-per-game) averages in the playoffs with a minimum-games cutoff somewhere between ten and fifteen games. And yet not once in years of debate surrounding Marleau’s playoff history have I ever seen anyone look that stat up. The most important and accurate stat relevant to that discussion, and not one person has looked at it. That’s because no one actually cares about the truth, these people just care about backing up their pre-conceived agendas. You might counter by saying, “well you just said the anti-Marleau people haven’t cared to look up that stat, either,” but the difference is, those people aren’t writing statistical studies. They’re just reporting what they see, and being called idiots for it by the side that does go to great lengths to find stats they can use to, well, call people idiots with. They are the advanced stats community, after all. Shouldn’t they, of all people, be interested in the most relevant stat on the subject?
Bias like that is rampant among what are supposed to be objective advanced stats studies. In case you wanted further proof that bias is also influencing the advanced stats community’s stance on Thornton and Marleau in particular, here it is. Let’s take a look at an actual advanced stats blog written two years ago about Thornton and Marleau’s future.
First, let’s establish what their “stance” is. I’ll try to characterize it as best I can. They have many stances on many subjects, obviously, but when it comes to Joe Thornton and Patrick Marleau, the short of it is that they claim to believe the two are still very productive players who are worth the contracts they are being paid, and that the Sharks were one win away from eliminating the 2014 Stanley Cup Champions in the playoffs with them on the roster, and that as a result Doug Wilson should not trade them this offseason because the Sharks could still win the Cup with them, and the Sharks would be worse without them in the immediate future.
So where’s the bias in this seemingly unobjectionable line of thinking? Well, besides the statistical fallacies that are informing these opinions at their core, that is? Here’s the simplest, most obvious one: Joe Thornton and Patrick Marleau are both 34-years-old, so to be 35, and as it turns out, the actual stats in #fancystats don’t have very good things to say about 34-year-old players. Or 34-year-old golden retrievers. Miss you, Bugsy… But hey let’s take a look at some quotes from Fear The Fin’s own blog, written in 2012 by The Neutral, titled How many years of productivity do Thornton and Marleau have left?:
On Thornton: “This group as a whole experienced a fairly steep decline in both their 33 and 34-aged seasons, going from averaging 77 points a year from 30-32 to 67 at 33 and just 59 at 34.” (Emphasis mine).
So that analysis is very clear. The Neutral articulates the information well: his own calculations two years ago told him that Joe Thornton was likely already in his declining-years at age 32, and that it would likely just get worse at age 34 and on. Now it’s two years later, Joe Thornton is nearly 35, and yet The Neutral is among those criticizing Doug Wilson in absolute disbelief that Wilson could even suggest the notion of trading Thornton. What more obvious an example of bias could there be?
The Neutral’s own calculations, his own stats, which are precisely what the #fancystats community claims to value above all else for their analysis, tell him that if Thornton is not already on the decline at age 34, the odds are extremely high that he will be in sharp decline over the course of the next three years of his contract. A contract which has a no-move clause in it. So if you think it’s hard to trade him now, just wait until his game goes the way of Dany Heatley–a player with a similar build and physical skill set, in two or three years, and see how many teams want to trade worthwhile assets, or anything at all, for a player like that making almost $7M a year.
That’s the type of situation The Neutral’s stats warn us about. And yet these days he’s arguing the opposite of what his own study told him two years ago. Essentially his brain knows that at age 34, Thornton is likely to get worse very soon, but because he is invested in this player, and because he’s spent years defending this player and he wants badly for this player to lead the Sharks to a Cup and prove postseason-performance-dissenters like himself right, he’s letting himself forget those stats that clearly show this is likely the exact time to trade Joe Thornton, while his value is still relatively high, because a drop-off in production is all but likely coming very soon and not later.
If this were a player on any other team, The Neutral would simply cite the stats that say keeping a forward at top dollar from age 34 to 37, expecting steady or improved performance, is folly. And remember, the Sharks likely need improved performance, not just steady. Thornton will likely have to outplay Kopitar, Toews, Getzlaf, and so on if the Sharks are to win the Cup, possibly all three in one postseason. But since it’s his player, it seems The Neutral has subconsciously chosen to ignore his own calculations on the matter and, in fact, argue against the very implications of his own stats. That is the definition of bias and of selectively enforcing the stats only when the conclusions they draw suit you.
Don’t believe me that the #fancystats community is letting their hearts influence their heads? Check out this comment from a most recent Fear The Fin article that is the only comment (at the time of writing this) “recommended” by three or more posters in the entire comment section other than one joke comment about puffing a cigar, and one meme.
“I love my team Cup or not, I would rather have Awkward, Teeth, the Joes, and Pickles than a lineup of douchebags that may have a better chance at a cup.”
Yes, this poster and, apparently, many others in the community who agree with him, is so biased towards players like Thornton and Marleau that he would rather the Sharks not win the Cup than win the Cup. As long as those players get to stay on the team.
What??? I just… What??????? He literally said he’d rather have a lineup with less chance of winning the Cup than a lineup with more chance of winning a Cup. That is the definition of illogical. But I guess because the team with less chance of winning the Cup is filled with affectionate nicknames, while the team with more chance of winning the Cup would obviously be “douchebags,” now it makes sense. Now I agree with it.
I wonder what would happen if you went on Fear The Fin and wrote the same comment, just with those affectionate nicknames and the word “douchebag” switched around a bit. “I would rather have a lineup of douchebags than have a better chance at a Cup with Awkward, Teeth, the Joes, and Pickles in the lineup.” The substance and logic of the comment would still be the same, but I doubt it would get such a positive response.
If this doesn’t show you what a big part bias, fandom, and emotional-manipulation, essentially, play in presenting stats, then you’re not paying attention. That comment may not cite any stats, but it motivated multiple other readers to agree with a fundamentally illogical concept simply through the use of emotional manipulation, and logic is what’s behind any (real) statistical argument. By contrasting the nicknames of beloved Sharks players with a hypothetical team of “douchebags,” this comment got Sharks fans who read it to completely ignore the real statement of the comment. All they latched onto was, well I’d rather have my favorite players than douchebags, too. Who could disagree with that? The comment represented, and played off of, a bias so strong, that everyone involved, from the author of the comment to the people who hit the “rec” button on it, didn’t even realize that their thoughts were falling into the completely absurd. In a league where the ultimate goal is to win a championship, they all just signed off on a comment that stated they’d rather have a worse chance of winning a championship than a better chance. Try quantifying that in an excel document and your computer will probably explode.
Yeah, I guess I don’t really know how computers work, do I? Still, you see my point, yes? And it drives at something deeper than just normal sports fan bias. Normal sports fan bias usually comes in support of a team. This is different. They said they’d rather lose with certain players than see the team win without those players. That is supporting individual players not because they are on your team, and not even independent of whether they are on your team (i.e. supporting them on another team), but supporting them instead of your team. You would rather your team’s odds of winning suffer than see your team without those players on it. That is very unusual, and points to what I spoke of earlier. Many in this advanced stats community have become so invested in Thornton and Marleau through this statistical fallacy of theirs that has convinced them, basically, that Thornton and Marleau are being unfairly targeted, that they now care more about seeing Thornton and Marleau prove this false foundation of theirs right, essentially, than they care about the Sharks as a franchise. These fans have begun to root for themselves, essentially, rather than their hometown hockey team. They’ve begun to root for the advanced stats almost as if it were a team, and they see Thornton and Marleau as the greatest representation of that on the ice because they don’t know that their defense of those players is actually based on fallacies, not real, correctly calculated advanced stats.
I’ll use a hypothetical example, just so I can give the advanced stats crowd a break for a minute. It’s like when you’re watching junior hockey, and you peg some prospect no one has ever heard of as the next great thing based on something you saw. You stake your reputation on this player’s potential. Everyone says you’re crazy, that they’ve never heard of this guy, but you stand firm to your position, that he’s going to be great. Well now you’re going to start rooting for the player, right? But why? Does he play for your team? No. Is he from your hometown? No. It’s because you identified him. You’re essentially rooting for yourself, your prediction, for the rewarding feeling that comes with being proven right when everyone else said you were wrong.
Now take that desire, and multiply it by a thousand. These fans have invested their emotions in defending these players for years, insulting anyone who disagreed and putting them down while pumping themselves up ever higher, all on the basis of these fallacies that they believe are true. They don’t know any better.
Their own advanced stats study even shows that Thornton is the one player whose production drops so much in the playoffs that he even shows up outside of the study’s falsely contrived expected chance variance (or whatever they called it this time), which was designed to greatly limit the number of players that actually showed up as having increased or decreased production in the playoffs. It was only Joe Thornton at one end of the spectrum, and Danny Briere at the other. Joe Thornton is literally the playoff’s anti-Briere. It’s ridiculous. And they know that. Their own stats show them that. And yet still, you’re an idiot if you question Thornton’s postseason play. Even though it’s right there in their own study.
Because for some reason, even though using the postseason’s anti-Briere as one of your representatives would seem to be setting yourself up for failure, the situation with Marleau and Thornton has come to represent the whole advanced stats community, at least within the Sharks fan base. So there is so much emotion pent-up in this one situation. So much has gone into this. They truly believe Thornton and Marleau are being wrongly persecuted, basically, by ignorant idiots. They see themselves as the great defenders of not only these players, but logic and reason. It has gone to such an extreme that you now see comments all over the place like the one I posted earlier. I’d rather see Thornton and Marleau get a few more chances to prove me right with the Sharks than see the Sharks win a Cup without those two players, where I no longer have any chance of being proven right by them.
That’s basically what it is. Strange things can happen, psychologically, when you invest yourself in sports that you have no control over, over a long period of time, and then you combine that emotional, often extremely frustrating and disappointing experience, with statistical fallacies that fit your narrative, and statistical studies you can manipulate. It comes to a point where you’re not just manipulating or misleading other people, or being mislead or manipulated yourself by other people, but you’re actually manipulating and misleading yourself. And not just logically, but emotionally as well. And I make no moral judgement about it. I’m just pointing it out. There are people within the Sharks fan base right now who walk, talk, and act like Sharks fans, and who think of themselves as Sharks fans, who don’t actually care about what’s best for the Sharks going forward. Strange as it sounds, it’s the truth.
Now one last look to The Neutral’s blog about Thornton and Marleau. We already went over his comments on Thornton. Here’s what he said about Marleau.
“Again, the group as a whole suffered a pretty sharp decline going from averaging 37 goals a season from 30-32 to 31 at age 34 and just 26 at 35.
As we see in the chart of Thornton’s comparables, the numbers seem to balance out after that but that could be as much about survivorship bias as anything else–the only players who even make it to 36 in the NHL are ones who are still productive.”
More of the same. And in The Neutral’s conclusion to the article?
“So what does this all tell us? Well, nothing too unexpected–even players who have historically been as productive in their early thirties as Thornton and Marleau tend to tail off pretty substantially in their 33- and 34-year-old seasons.” (Emphasis mine).
End of story. When Neutral was able to look at the stats as hypotheticals back then, knowing there was no immediate danger of actually losing either of his beloved Sharks mainstays, he had no problem being objective, for the most part (I’ll get to that in a minute). But now that it’s two years later, and those 34-year-old seasons (35, actually) are a reality that Doug Wilson has to deal with, The Neutral’s fan-instincts have kicked in. He doesn’t want to lose Thornton or Marleau, period, stats be damned. And for a #fancystats guru, that is the ultimate hypocrisy, and that is bias.
In fact, even the study itself showed bias because it was very selective in the criteria it used for Thornton and Marleau. Somehow this particular study of Joe Thornton suggested declined-production wasn’t likely until well after age 30. I distinctly remember that when the same author did the same type of analysis discussing possible free agent or trade targets for the Sharks in the past, such as Ilya Kovalchuk and Rick Nash, he found that the age forwards started to decline at was 26 or 27 by quoting this study done by Hawerchuk of Arctic Ice Hockey.
Here’s the graph taken from Hawerchuk’s study on the subject, which clearly paints an even dimmer picture, and, oh yeah, a completely different one by about six years, than the study The Neutral did on Joe Thornton on the same question of “peak age.”
But, well, Thornton already showed he was productive up to age 32! That changes the scenario! You might say.
And Kovalchuk showed he was productive up to age 27. Not to mention Thornton’s point production went from 114 in 2007, to 96 in 2008, to 86 in 2009, to 89 in 2010, to 70 in 2011, to 77 in 2012, to a 68-point pace in 2013, to 76 in 2014. 110+, 90s, 80s, 80s, 70s, 70s, 60s, 70s. You could argue Thornton’s been in slow decline since he was around 27, too. But The Neutral didn’t.
Well scoring has gone down across the NHL in those years.
I don’t remember that absolving Ilya Kovalchuk or Rick Nash in the analysis done on them. That would be just another example of giving favorable considerations to our own players in instances we didn’t care to for neutral players. That’s bias, even if the considerations you care to give for your own player actually result in the more accurate study, because it’s still a double standard.
Well Thornton is big and strong, and has never relied heavily on speed for his success, so those kinds of players can keep producing longer! Yeah!
Rick Nash is big and strong, too, and he started out faster than Thornton, too. So are Dany Heatley and Ryan Clowe, who were never speed-demons either. Players who never relied on pure speed probably never had top speed to start, which means once they start to slow down, they can get really slow and actually decline much faster and even fall out of the league almost overnight. See Jonathan Cheechoo, or Scott Gomez, or Mike Richards who has quickly fallen to the 4th line in LA.
But… well Marleau is super fast and skilled! Those types of players can keep producing longer than the ones who are just big and strong but never relied on speed! Yeah!
So was Ilya Kovalchuk.
But… Well, maybe it’s, Hawerchuk’s study is for when the average forward declines, right? But Joe Thornton and Patrick Marleau are better than the average forward, so that’s why The Neutral had to do a new study for them! Yeah, that’s it!
The Neutral thought Hawerchuk’s study was accurate enough, and applicable enough to all NHL-forwards, to use it as the basis for his analysis of Rick Nash and Ilya Kovalchuk. And they weren’t even 30 yet. Were they not better than average? Were they not star players? Did they not warrant their own studies on the basis of talent or performance just as much as Thornton and Marleau did? Yes, they did.
Want to know something else that’s concerning? Eric T. of SB Nation did an article last March on the same subject as Hawerchuk where he mentions that “Hawerchuk’s study of points per game is probably the most-cited analysis of hockey aging. Its simple elegance provides compelling evidence that points per game peaks at around age 25.” (Emphasis mine). 25. Not 30, not 32, certainly not 34 or 35. 25. Here are a few more quotes that raised my eyebrows…
“His method … was prone to problems as players started dropping out of hockey … (and) couldn’t reliably tell us how steep the drop-off would be in a player’s 30’s. (Emphasis, parenthesis, the word “and” and two ellipses I think I actually did right, mine. Fucking blogging).
So we have what is “probably the most-cited analysis” of hockey aging, in case you thought I’d just found some small-time blogger’s work to poke holes in or that the miscalculations I referred to earlier were the exception, and yet Eric T. admits this commonly cited and spread analysis is “prone to problems” and cannot “reliably tell us” certain pertinent information. The defense of this would be that the study wasn’t meant to tell us the drop-off past 30, and yet The Neutral and many others used it as their basis for analyzing signings that went well past the age of 30, like Ilya Kovalchuk. He also mentions how Hawerchuk’s study didn’t even make use of the “best practices developed in baseball analysis.” Because why would you want to use the best tools available for the most-cited study on the subject, and thus the one most spread throughout the hockey world.
Eric T. then goes on to mention how the method he recommends for studying “peak age” isn’t even the method Hawerchuk used, and that the conclusion of Hawerchuk’s study was a year off for forwards, that his study showed the peak age was actually 24, not 25, which doesn’t even have to do with the drop-off after 30. They couldn’t even agree on the early twenties. And so we’re now going on three different versions of the same, or very similar, “statistical analysis,” with three completely different conclusions, and tons of contradiction and inexactness in what should be an exact science. Not to mention both Hawerchuk and Eric T’s studies sound way, way off because good NHL players who start to decline at 25 are the exception, not the rule. Their studies must be including players like Jed Ortmeyer, which would skew the statistics.
And this isn’t to critique Fear The Fin exclusively, or Hawerchuk, who does the hockey world a great service providing all the corsi-based stats at his blog behindthenet.ca. It’s just that, as prominent advanced stats blogs, they have examples in print that can be referenced more easily than, say, old tweets of fans on twitter that would be difficult to find, plus it has to be said that Fear The Fin, and The Neutral in particular, has taken a bit of a leadership role for this segment of Sharks fans. It’s hard to ignore the irony that Thornton and Marleau, and the group of fans that most supports them, are perhaps both being led in the exact opposite of ways given that the criticism of Thornton and Marleau has always been that they’re too laid back and unassuming.
But I don’t have anything against any of these people. I don’t even know them at all outside of this fan-divide. It’s just that I don’t think what’s been happening with this group of Sharks fans is a good thing. When you combine incorrect information or assumptions, statistical fallacies, and bias, with a high level of boldness and a messiah-complex type of mind-set relating to hockey analysis, it’s not a recipe that benefits anyone.
And you may be reading this wondering what makes me any better? For instance who am I to question the qualifications of someone else? The difference is that when people read what I write, no matter how firmly I believe it or how sure I am of myself, they understand it’s still an opinion. I don’t claim to be the messiah for hockey analysis. I don’t claim my opinions are statistics, or facts, when they’re really just opinions. That’s a big difference.
The approach of the #fancystats community runs in stark contrast. It’s only natural that many Sharks fans have a bias towards Marleau and Thornton, the two biggest stars in franchise history, but rooting for them because they’re your team’s biggest stars is a lot different from rooting for yourself through them because you have an imagined persecution-complex on their behalf based off of statistical fallacies and years of staking your personal reputation on their success when it never would have had anything to do with you, personally, if you could have just come to terms with their shortcomings. But the problem is that the latter group always starts out as the former. Even many fans who just like Marleau and Thornton on a surface level as stars of the Sharks can’t help but like hearing that the criticisms befalling their beloved players aren’t really true. And this narrative isn’t just being sold to them as a dissenting opinion, but rather they’re being told, “These are the one true facts. You have the truth in your hands now. No one else. Everyone else just has opinions. You are the one with the facts.” When in reality, they’re not getting closer to the truth combining their bias with incorrect information, they’re getting further from the truth.
So what you ultimately get is a group of fans who are straying further and further away from reality and into the absurd, having their natural biases encouraged by fallacies masquerading as facts, while at the same time believing more and more that they are the only ones who know the true reality of the situation.
And when it’s all said and done, you have a group of people who are simultaneously the most arrogant, and the most sure of themselves and their knowledge, that any group of fans could ever be, while at the same time being more biased, and more full of incorrect information and just plain wrong, than any group of fans could ever be.
It is literally the worst combination of attributes any sports fan could encompass not counting Dodgers fans (is it too late to play to the Bay Area homers?). And some of these fans get so lost in their own world where they are always right that they don’t even bother with reality anymore because they literally have their own one in their heads.
If you can’t believe it, just look at what’s happened with the Sharks recently. This group of Sharks fans has already forgotten about the collapse. Why? Because the collapse never even happened! Not in their reality. What collapse? These Sharks fans are already talking about winning the Cup next year. We can’t trade Thornton or Marleau because that will prevent us from winning the Cup next year. We were right there with the Kings. If we hadn’t lost Vlasic, we would have beat them for sure. Nevermind that Vlasic has come out himself and said they still would have lost had he been healthy. Of course, he’s supposed to say that. But that is how this group of Sharks fans thinks in their own world.
It’s like the last decade of playoff failure never happened. It’s like the collapse against the Kings never happened. In their minds, we took the Kings to seven games (I AGREE WITH THAT PART! I want that on the record), and would have beaten them without the Vlasic injury, and since the Kings won the Cup, that means if we just keep the team together for next year, we’ll have a great chance at winning the Cup.
Nevermind that Doughty was banged up for the Kings to start the series. Nevermind that Pearson, Carter, and Toffoli did not play together until game three of the series. Nevermind that Anaheim and Chicago also each took LA seven games, and they had to play against a healthy Doughty, and the Pearson, Carter, and Toffoli line, for all seven games. Nevermind that there are only sixteen teams in the playoffs each year, and that if you lose in the playoffs every year, the team that beats you starts out, at minimum, in the last eight, and one of those teams is probably bound to go on to win the Cup at least once during that ten-year span. Did the team that eliminated us last year (the Kings) go on to win the Cup that year? No. What about the year before (the Blues)? No. What about the year before (The Canucks)? No.
That is reality. It’s been nearly a decade. The Kings core is young, with Kopitar, Doughty, and Carter at the peak of their games. The Blackhawks core is young, too. Meanwhile our core players of this last era, Thornton, Marleau, and Dan Boyle, are 34, 34, and no longer a Shark. But these Sharks fans expect Thornton and Marleau to improve at age 34 (or 35), and accomplish in their declining years what they were unable to accomplish during their primes? Thornton couldn’t outplay Anze Kopitar this year, but these Sharks fans expect him to be able to do it next season, when Thornton is a year further past is prime? The reality is that you need to win four playoff rounds in a row to win the Cup, and we’ve only won one playoff round in the last four years. Although one in four has a nice symmetry to the required four-in-one, that’s just not going to get it done. And even if we’d gotten out of the first round this year, which we didn’t, Anaheim and Chicago would have been just as difficult to beat as the Kings. Harder, in my opinion. Those teams couldn’t be exploited with speed quite as easily as Robyn Regehr, Matt Greene, and even Willie Mitchell at times could. Well, except for Bryan Allen.
It just doesn’t make logical sense. This is a group of fans dedicated to stats, a group that validates all their opinions through stats and all that is quantifiable, and yet when it comes to these two players at age 34, and what this team should do with them, all that logic they’re supposedly dedicated to just goes flying out the door. And where their misconception of how statistical analysis works really does a number on them, and encourages them to ignore reality, is in regards to the collapse. As I alluded to earlier, they deny the collapse even happened at all. Those games could have gone in any order. In truth the Sharks didn’t collapse or really “fail” at all, since only one team out of 30 wins the Cup every year. When you look at the full picture, the only indisputable, statistical fact is that the Sharks took the eventual-Stanley Cup Champions to seven games. Collapse? Ha! That’s an ACCOMPLISHMENT! And if Vlasic hadn’t gotten hurt, we would have beaten the Kings, who won the Cup, which means we would have been Cup favorites! How can Doug Wilson possibly dismantle the Cup favorite team from this season!
These are the arguments they make. And when you ignore reality, when you put no stock in what actually happened, then of course you’re going to disagree with the people who do. Doug Wilson actually watched the series against LA. Seven games. He actually analyzed that series. What happened in each game. The order of the games. How the coaching staff even warned the players against complacency after game three, but the players, led by captain Joe Thornton, didn’t listen, and didn’t respond. How the players in the locker room literally told Doug Wilson that they felt more like co-workers than a team, and that maybe that explains why the team didn’t fight towards the common goal of the Cup together the same way that the Kings did, or that the San Antonio Spurs did all season long in the NBA.
Doug Wilson actually analyzes all these things. That happened. Things that happened. In the playoffs. Which also happened. But when you subscribe to a complete misinterpretation of statistical analysis that says everything that happens over a “small” sample size (i.e. the playoffs, if you’re a Thornton-Marleau fan) is just “noise” and “chance” and “variance,” guess what the result is? That’s not statistical analysis, that is statistical ignorisis. (That’s a word, right?) As in you ignore everything that happened in the playoffs, ignore everything the players experienced personally with each other as teammates, ignore every decision and adjustment the coaching staff made during the series, and instead just say, “Well we were rated x in fen-close over the regular season, which gave us x-percent chance at winning the Cup, and even though it didn’t happen this season, that’s just how it works with chance and bounces, so all we need to do is repeat the same process next year, and as long as we’re still one of the better fen-close teams we’ll have a chance, and eventually the game of weighted-dice that is the Stanley Cup Playoffs will roll in our favor, maybe.”
That is literally the implication of their argument, that the playoffs are a game of weighted dice, and you can’t judge anything that happens during a series because none of the players are self-determining anything, they’re simply each playing to the mean of their overall corsi-ability and chance is determining the rest over a small sample.
Again, such a false concept that completely takes away all agency from the players themselves, and completely discounts the decisions they make in the playoffs, the level of effort they put in, the individual plays they make, the shots they block, the pucks they get out of the zone or fail to get out, the penalties they take or don’t take… you know, the entire game of hockey that gets played hundreds of times every postseason It’s like how a stats blog (don’t remember which) used corsi as a team stat to judge the New York Rangers. Corsi, a stat that doesn’t differentiate between shots that are blocked, and shots that make it to the net. The New York Rangers, the top shot-blocking team in the NHL that year. They literally said, Look at the Rangers’ corsi! Can you believe how much they have to rely on Lundqvist? No, there’s a stat called save percentage that’s for that. Not corsi. The whole reason the Rangers blocked more shots than any team in the NHL that year was so the pucks didn’t get to Lundqvist. Only the shots that make it to the net count. That’s why most people use fenwick to analyze teams, not corsi.
But those are the kind of mistakes they make, over and over. And as for all the defenses they’ve cooked up for Thornton and Marleau around sample size and variance, that is what happens when you give people incorrect information and tell them it’s irrefutable fact, which is essentially gives them license to use it as a tool kit to support and dress up all their biased opinions as “facts” as well.
We’re going to try to weigh Thornton’s overall value as a player, regular season (where he’s always done well) and playoffs (where he hasn’t always done well) statistics included. Except without the playoff statistics because those don’t count due to insufficient sample size.
Yes, it is pretty hard to lose a debate when only the arguments you make count. That’s like having a wine contest between a white wine and a red wine where the wine judges are not allowed to drink or vote for the white wine. I wonder which wine will win then? It’s just beyond absurd. I heard Deadspin wrote a great article about how the advanced stats community is now using “luck” to explain away anything that doesn’t fit with their narrative or that they don’t know how to quantify, and that’s a dead-accurate
spin observation. Doesn’t it just work out so conveniently for the Sharks fans who want to defend Joe Thornton’s playoff performances that they can now suggest the fact is they shouldn’t even have to because there isn’t even such a thing? Of course, that leaves room for only one answer. For instance, if regular season performance represents the number 10, and playoff performance represents the number 4, well you’re trying to find the average of 10 and 4, but then not counting the 4. Well then you’re not averaging or answering or calculating anything. All you have left is the 10. It’s absurd. That is not how statistical analysis is done.
But that’s what they believe. And all current postseason performers fall under the expected parameters for statistical variance, except for, as I mentioned, Danny Briere (on the good end), and Joe Thornton (the bad end), and even those, while both outside the expected chance variance zone, were still “well within the expected number outside of the expected chance variance zone,” or whatever.
So you’re investigating and trying to statistically quantify the historically rare playoff-underachievement of the Sharks in the Thornton and Marleau era, but then even if you decide not to just ignore all the stats because: “sample size, and even if you find you can quantify that Marleau and Thornton really are a rare example of playoff-underachievement, well then it still doesn’t count because hey, it’s rare, and a small number of rare “variant examples” are expected in any analysis, so they can be explained away by statistical variance. “Is Joe Thornton’s drop in postseason production one in a thousand? Yes but you’re going to get one of those out of every thousand just by chance in any statistical model anyway, so it doesn’t mean anything.“ So even when it is, it isn’t. Also known as a self-determining equation that always comes out to “no.” It’s beyond belief that the people doing these studies don’t understand that they’re committing such basic logical errors, but they don’t. If only someone had told Jonathan Toews about all this because I don’t think he got the memo that you don’t even need to try in the playoffs anymore because all your performance hinges on is statistical variance.
And that’s what the advanced stats community uses to explain away any data that doesn’t support their argument. And in this case where it’s a group of fans in question who are looking to defend Joe Thornton and Patrick Marleau, it would hardly surprise anyone that they would try to play up the significance of their regular season success, and marginalize the significance of their postseason failures.
And that would be entirely forgivable, normal, expected “fan” behavior if they weren’t pretending to be objective statisticians.
So, does this mean I would trade Thornton and Marleau this offseason? What direction would I take the team in if I were in charge? Check back for Part Three in the Diverging Realities blog series soon. Thanks for reading!
Written by Shark Circle
PS: I don’t care if this changes the course of hockey analysis for the better forever, and brings fans across the NHL together, this blog was so not worth it. After I publish Part Three, I swear, the hockey world can stay as dumbed down as it wants to just so long as I stay smart enough to never attempt writing about complex subjects like this again.
Subscribe to Shark Circle by entering your email and clicking the SUBSCRIBE! button in the top right corner! Your readership is appreciated!