Sharks-Avs, through five
by Hawerchuk on Apr 23, 2010 1:26 AM EDT
Email
PrintIs it a sign that you don't live in a real hockey town when you go to a bar on a Thursday night that has hockey paraphernalia on the walls and they don't have the game on, and nobody seems to care? Undoubtedly, yes.
At any rate, here are the Sharks-Avs shooting stats for the Sharks perspective through five games:
| Down 1 | Tied | Up 1 | SF/-1 | SF/0 | SF/1 | SA/-1 | SA/0 | SA/1 | |
| Game 5 | 0 | 1705 | 116 | 0 | 35 | 2 | 0 | 20 | 2 |
| Game 4 | 0 | 2889 | 1335 | 0 | 57 | 24 | 0 | 43 | 16 |
| Game 3 | 0 | 3651 | 0 | 0 | 93 | 0 | 0 | 33 | 0 |
| Game 2 | 2438 | 1484 | 0 | 75 | 14 | 0 | 24 | 13 | 0 |
| Game 1 | 971 | 2629 | 0 | 12 | 45 | 0 | 13 | 32 | 0 |
| Totals | 3409 | 12358 | 1451 | 87 | 244 | 26 | 37 | 141 | 18 |
| Corsi | 70.2 | 63.4 | 59.1 |
Games 1, 3 and 4 were examples of what happens when the Avs get outplayed but things generally break their way. Game 5 was an example of what happens when they get outplayed and things go wrong for them.
Here are the scoring totals by San Jose's lead:
| F/-1 | F/0 | F/1 | A/-1 | A/0 | A/1 | |
| Goals | 6 | 4 | 1 | 0 | 8 | 1 |
| Shots | 87 | 244 | 26 | 37 | 141 | 18 |
| S% | 6.9 | 1.6 | 3.8 | 0.0 | 5.7 | 5.6 |
San Jose still has a ridiculously low shooting percentage in tie games, but they're outshooting Colorado 71-41 per 60 minutes, which lets them overcome their bad shooting luck.
32 comments
|
0 recs |
Do you like this story?
Comments
re: A Quick Programming Note
I may disagree with you on some points, but I for one am grateful for the time and effort you put into this. And grateful to Tom and Vic and the other guys who also work on this stuff.
As far as the Sharks, Finally, eh?
by DoctorMyBrainHurts on Apr 23, 2010 8:41 AM EDT reply actions
I feel slightly naughty commenting on a post that had comments disabled.
This behaviour isn’t peculiar to sports fans. If all it took to discredit woo-woo was to provide good sound evidence, chiropractors and Republicans and Don Cherry would all be out of jobs.
We’re pattern-seeking animals. When something happens that goes against rational expectations we construct a narrative to explain it after the fact. The alternative – to accept that shit happens, that dumb luck exists – is a difficult thing to come to terms with. It means that sometimes the better guy loses, that some results are not based on entirely merit, that you don’t always get what you deserve. Even within the bounds of a contest with relatively simple rules, the outcome can be unjust. That makes us uncomfortable so we rationalize the outcome post hoc.
Now, to the chemistry/clutch-under-pressure crowd: why are you wasting your time arguing on message boards? Since you have special powers that give you greater insight into the game than the Vegas oddsmakers who rely on mere evidence to calculate probabilities, you must have won a lot of money this week. Congrats! Get out there and spend it!
To copy and paste a comment I made on C&B:
I’m not sure what’s controversial about the notion that players come up against a hot goalie, start fine-tuning their shots, then fine-tune them too much and start shooting wide/into the chest a little more. Seems like human nature to me. In a small sample, I bet that happens all the time. It’s also eminently correctable, as last night’s results show pretty effectively.
Randomness happens a lot in a game as inherently chaotic as hockey, but to deny human nature as an element of a game played by humans is to put on a big-ass set of blinders when doing your analysis, if you ask me.
I agree that too much credence is given to intangibles by sportscasters and fans alike (Bob McKenzie recently mentioned something he was told by a GM, that fans and media talk far more about “momentum” than players and coaches do), and frankly most of the arguments in favour of psychological factors are kind of terrible, because they try to explain everything and deny luck altogether, which is foolhardy: shit happens, and some nights the bounces don’t go your way. But here’s another fact: NHL teams also look at chemistry and psychology, too. Most NHL teams employ sports psychologists. Most NHL teams account for guys who’ve played together before, or guys with complementary play styles, in building their teams. That doesn’t seem like a foolish notion to me: if two guys have played together, then they already understand how each other play the game, and know where to expect a guy to be on a given play, making execution easier.
Hell, my exercise physiology prof once told us about a study where the placebo effect actually resulted in increased performance in elite-level athletes — they thought taking asthma medication would result in improved race times, because it opened the lungs. When the numbers came out of the wash, the inhaler group’s times actually weren’t significantly different from the placebo group, though both groups did improve their times. When you’re in the best shape of your life, and the difference between success and failure is measured in milliseconds or millimetres, depending on your sport, what’s going on upstairs can be absolutely crucial, because at the end of the day, what you do is determined by what’s happening in your brain. The technical term for not getting maximal performance out of a muscle in an action despite the muscle being capable of more is “central fatigue”: the muscle can do it, but it’s not being activated or coordinated properly by the brain or spinal cord, for whatever reason. It’s not voodoo: it’s good old-fashioned science.
Humans aren’t automatons, something that people who lean entirely on luck to the exclusion of other mitigating factors seem to forget. Humans also aren’t so fragile that the slightest change in circumstances creates a tailspin, as others have argued here, either. The truth, as with most things, is a little of A, a little of B. I just wish some people would get out of their entrenched positions and actually consider what others are saying in good faith, instead of deciding their arguments are lazy cop-outs based on pseudoscience, which both the stat-heads and the…well, Avs fans, have done here.
SNN Sports - A theoretical Oilers blog (i.e. theoretically, I write stuff there). Link now 100% less broken.
I don’t deny the role of human factors in the game. We can quantify 50% of winning, but the other half is a mystery.
There are several things that bug me about the psychological explanations:
1) They presume knowledge of the player that we simply don’t have. Does Joe Thornton under-perform against adversity? Maybe. But how the hell do any of us know?
2) They assume that Doug Wilson and Todd McLellan are complete failures as motivators and psychological analysts.
3) Even though they’re made up after the fact to match the scoreboard, they don’t explain why the Sharks shooting percentage is normal when they’re down by 1, but terrible when the game is tied.
Agreed fully on the first two points. We can only work with what we can observe from the outside, and I’m sure there’s a lot of projection going on. For #3, if I wanted to try an explanation, maybe the Sharks aren’t trying so hard to find the perfect shot when they’re down, they’re just trying to get the damned thing on net like they’re supposed to; simplifying their game, essentially. I admit it’s sketchy, but it makes a certain amount of sense, if you accept the premise that the Sharks are trying to pick their corners too much against Anderson in the first place.
I’m trying more to address the larger fact that people like R O and Sisu will shit all over anyone who even tries to make a psychological argument (even agreeing that most of the arguments presented make me facepalm in a serious way, and suggest that these people have never actually taken a sports psychology course), and I think that that’s just as invalid as trying to deny that luck plays a major role in the game, especially over the small samples. Just because it’s not something you can assign a number to and run a correlation on, doesn’t mean it’s not a relevant factor, especially over the small samples where so much fluctuation occurs.
SNN Sports - A theoretical Oilers blog (i.e. theoretically, I write stuff there). Link now 100% less broken.
I guess the problem I run into here is that you came up with an explanation to match the facts after the game was played and after I posted the numbers.
I don’t think anybody who watched the game with an eye for psychology came up with that particular theory on the fly. In fact, I don’t think anybody else noticed! All I heard was that, in general, the Sharks couldn’t score because they were frustrated by Anderson and second-guessing themselves.
The other issue I have is that I don’t think most of the people who offer up these explanations actually believe them. MGL had a nice post on his blog about how you get different answers of the odds of something happening if you just ask people or if you ask them to put money on it (they’re much better at handicapping when there’s money involved.) The Avs were huge dogs last night (+220), so the people who were convinced the Sharks lacked mental toughness should have put money on the game like Jeff suggests.
Well, obviously, you can only evaluate the results when you actually have results to evaluate. I predicted the Sharks would win in 4 going in because I didn’t think that this would happen, and came up with the explanation as a way to try to figure out why that didn’t happen. Maybe I’m wrong. Maybe I’m overplaying Devil’s Advocate. I dunno. Just trying to suggest alternative explanations, because “it’s all luck” doesn’t sit well with most people.
I also don’t think it’s an absurd lack of mental toughness or self-doubt that would lead to changing up your shots. I really think it’s just a matter of, “Well, that didn’t work, let’s try this. Well, shit, that didn’t work, either, let’s try this,” until something does work. I mean, it’s not unreasonable that if a goalie’s got a hot glove that you start shooting blocker side instead, right?
SNN Sports - A theoretical Oilers blog (i.e. theoretically, I write stuff there). Link now 100% less broken.
I love this Marleau quote:
“The puck just isn’t going into the net,” Marleau said. “It’s just one of those things that comes and goes. But it’s not going to stay this way…”
I don’t think hockey players spend as much time doubting themselves as we think they do. One of the NHL execs I know put it to me this way: “We beat a team 5-0. Everybody thinks we played great. But we got outshot early, played like shit and got a couple of lucky goals. And the guys never want to hear it.”
Like I said, I don’t think it’s self-doubt. I think it’s just natural reaction and adjustment to circumstances in the game.
SNN Sports - A theoretical Oilers blog (i.e. theoretically, I write stuff there). Link now 100% less broken.
Dig up some old Steve Yzerman quotes
He has talked about the randomness of the game many times and even said that the Olympic Tournament was basically a “crap-shoot.”
The Sharks got all the way down to -130 when they were down 2 games to 1, so I went after it. That was just much too pretty to pass on, considering the domination in the series to that point.
I do put my money where my mouth is. Regularly.
Editor of The Copper & Blue, and leader of The Cult Of Hartikainen.
I’m not sure what’s controversial about the notion that players come up against a hot goalie, start fine-tuning their shots, then fine-tune them too much and start shooting wide/into the chest a little more. Seems like human nature to me. In a small sample, I bet that happens all the time
You’re making a testable prediction here. In fact, you appear to be willing to bet on it.
Define “hot goalie.” What if a goalie posts a sv% one SD better than his career average in G1 of a series. Is that a reasonable definition?
You claim the opposing team will often continue to shoot below their natural skill level in G2 and/or will miss the net more often than usual. By how much? Say, half the teams will shoot half a SD worse than usual? Perhaps a quarter SD? And the same for missed shots?
Now, do you really think this effect will be statistically significant?
I’m going to predict your response. “Just because an effect does not appear to be statistically significant does not mean the effect isn’t real.”
That is true. Unfortunately it does mean that your theory, like any other random woo, serves no practical purpose.
Is there such a thing as “appearing” to be statistically significant? I’ve always understood it as p < 0.0x (usually 0.05), and that’s the end of the discussion. Hell, are within-game samples even big enough to get statistical significance?
I’m also not even talking about between games. I’m talking about making adjustments within a game because nothing’s going in. You’re implying broader strokes from my statement than it contained. All I’m saying is, if a goalie has a 20-save period (or something like that), shooters might try aiming in different places because where they were aiming before obviously wasn’t working. They’re still outchancing, outshooting, and outCorsiing, but until something goes in, by luck or by skill, they’re going to try different things periodically, because whatever they were doing before wasn’t going in.
SNN Sports - A theoretical Oilers blog (i.e. theoretically, I write stuff there). Link now 100% less broken.
This sounds reasonable. It would be an interesting thing to check. Don’t some of the NBC games have a shot location thing?
by Scott Reynolds on Apr 23, 2010 3:21 PM EDT up reply actions
Of course there is. Anything at or around 0.05 is what is often referred to as marginally significant. 0.05 is an arbitrary number something sitting just on either side is marginal. Additionally there is the issue of multiple comparisons. Add in those occasions were we are looking to see if a pattern we think we saw is reflected in the numbers or just straight up numerical fishing trips and a p-value can appear significant when it is not. Also most of what we do in sports are not experiments but observational and therefore violate most of the hypothesis assumptions there are all kind of reasons to appear significant when it is not significant in a meaningful way.
by Mogen_david on Apr 23, 2010 11:35 PM EDT up reply actions
These are only a factor if one of the teams doesn’t do it at all, or does it very poorly – just like managing injured players, cap space or travel times. If you don’t do them to an average level they will hurt you badly over the course of a season or more, but I doubt the effect will be measurable on a game by game basis. You have to maintain your athletes at their peak as much as possible throughout the season, everyone has to do this, but it is likely over the course of a season that there will be more variation within a team than between teams, unless a team completely fails at it.
I follow a couple of individual sports to a fairly close degree, and these issues do crop up in that one or two athletes will find an ‘edge’ (training method/tactical approach/whatever) or a country will develop a training system or broad base of recruitment in a marginal sport that ends up just flat out outperforming other countries/athletes for a season, or even half a decade. But then someone will hire a coach out of the system, a retired athlete out of the system or heck marry into the system (happened partially in biathlon), and the results will even out over time.
In conclusion, in the incestuous world of hockey (coaches/managers/scientists in the field are a small community at the highest levels, and they all talk to some extent) everyone will be grasping for ‘an edge’ in managing the players. And to quote the Incredibles: “When everyone is super, no one is.” In the big picture all of these things wash out unless you fail at it (see discussion of Edmonton’s injury management in the wake of Souray’s comments), or the ice is tilted (travel effects).
What a lovely country you have
When I was in your area last week I couldn’t believe how hard it was to find a hockey game on TV. I’d go to sports bars with 5 big screens and no hockey, and get strange looks when I asked for one of them to be switched.
Come on – California has 3 hockey teams, 2 of which are in the play-offs. Quebec, the birthplace of the sport, has only 1 team. About a fifth of the world’s professional players are from Ontario, and even they have fewer teams than California.
Time for some league contraction.
California has the world’s 5th largest economy by itself, and has more people in it than Canada (36M to 33M). Canada has 6 teams, Cali has three.
The New Improved Avalanche. Now with Real Coaches!
Jibblescribbits: C'mon over and waste some time
by Jibblescribbits on Apr 23, 2010 10:44 AM EDT up reply actions
It’s actually the 8th largest economy, but who’s counting?
http://en.wikipedia.org/wiki/Economy_of_California
Also, population means squat if it’s not interested in the product. I’m sure per capita Canada can support easily 3x the teams of California.
Not that I think contraction is relevant; all three teams in California are sufficiently thriving that it would be silly to suggest it.
Sorry. At one point i had heard 6th.. my mistake
The New Improved Avalanche. Now with Real Coaches!
Jibblescribbits: C'mon over and waste some time
by Jibblescribbits on Apr 23, 2010 1:01 PM EDT up reply actions
And they have 5 MLB teams, 4 NFL teams and four 4 NBA teams Totaling 16 Pro teams
Canada 1 MLB, 1 NBA team and 8 CFL teams Totaling 16 Pro teams if you include the CFL.
by Mogen_david on Apr 23, 2010 12:35 PM EDT up reply actions
If you’re in the San Francisco area, you’ll have no trouble finding hockey games on at bars, at least from my experience.
..:Fear The Fin:..
Perceived ref bias - penalties vs. zone time
I think an interesting statistical perspective on the perceived ref bias in the SJ-COL series thus far would be to compare the ratio of time spent by each team in their respective zones compared to the ratio of penalties by team, all by game, and in the series as a whole.
..:Fear The Fin:..
I don’t remember, I think Vic showed a convincing correlation between the two.
I should clarify, what I want to do is try and relate power-play minutes to zone time (as proxied by EV shots). I mean off the top I suspect there’s correlation, after all the game on the ice dictates that the defined infractions are taken more often when players don’t have the puck.
But, I don’t know if that’s actually the case, I mean officiating is just off the charts sometimes. Still, there’s likely a significant luck component in “bad” officiating, and it’s impossible not to be biased against refs so it’s not like I’m an impartial observer.
Which leaves the remaining undebatable hockey truth of how penalties are (more often) taken. Should be testable, but I’m lazy and need to update my code to parse the gamesheets.
Question
on the shot table, what do the followings mean?
F/-1 F/0 F/1 A/-1 A/0 A/1
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
