Friday, March 21, 2008


On Consistency

This is Eldrick Woods. You've probably heard of him, he's merely the most dominant professional athlete of my lifetime (and even if you're quite a bit older than me, probably of yours, too).

I realize that baseball analogies are more popular, but I've been thinking about golf a lot lately when it comes to explaining hockey. You play? I still do a bit; when I was younger, in the pre-family days, I played a lot. The summer after I graduated from university, I played 103 18-hole games between the May long weekend and Labour Day.

It's a funny game. You can jump straight from the car to the teebox, mind still racing with work crap, and shoot wonderfully; you can get to the course early, warm up, get comfortable, and stink it up. You can play in a scenario (say with family on vacation) where you honestly do not care how you score, and go low; you can play in a scenario where you desperately want to play well, and be completely hapless. I have been shocked by my own golf score (good and bad) on so many occasions that I was jarred into realizing something one day:
(A) I'm a 9-handicap, and sometimes my score is going to be better than that, and sometimes about the same, and sometimes worse.

(B) If I want to be a 3-handicap, I can practice and play a bunch more, and get some lessons, and maybe it'll happen... but I'm not one today. I can hit all sorts of good shots and make smart decisions indistinguishably from a 3-handicap, but over time, I'll average a few more mistakes and a few less terrific shots each round, because I'm not a 3-handicap.

This is what comes to mind when I'm driving the boy to school in the morning, and I hear the guy doing the morning sports update declare that the Flames put forth "no effort last night" in "another inconsistent performance" losing to whoever. You know, for a tubbo in a windowless room to claim that a pro hockey team didn't try -- that's 18-20 guys who are in the top one-thousandth of hockey players who have ever tried to succeed at it -- is self-evidently shameful, but that's beside today's points. One of which is simply that sometimes people have a bad game despite preparing well and wanting it really, really bad. (God help the poor Olympic athlete who has the wrong bad day.)

The other point is something like this: there are at least a thousand golfers in North America who can hit it as far as Tiger, and make 25-foot putts on occasion. There are also duffers who have played with him in a pro-am and beaten him on a particular hole.

"The difference between Tiger Woods and Other Golfer is their consistency" would be an extremely stupid, nonsensical thing to say, and yet when the contrast is less extreme, people say this all the time. The Leafs' inconsistency has cost them a playoff berth... if the Flames were more consistent they would have clinched the division already... if the Sharks hadn't been so inconsistent at midseason, they'd be ahead of the Red Wings... etc. etc. etc.

I have posted before that whenever a coach or media type says that Local Team X needs "more big _____", that can easily be shortened (and made more accurate!) by deleting the word 'big'. Today's lesson in hockey/sports lingo is that more consistent = better. If the coach says the powerplay needs to be more consistent, then what he means to say (even if it's subconsciously) is that the powerplay needs to be better. If your team needs a more consistent goalie, what they actually need is a better goalie. And if the Flames can only win their 1st round series if they play very consistently, what that actually means is that the can only win if they play very well.

And if I play a round of golf with Tiger Woods, I might tie him on the first hole, but when he beats me by 12 or 15 on the round, it won't be because I lacked consistency.


Oddly, I was defending Huselius on a messagebaord somewhere recently, because there's this public perception out there that Huselius isn't all that good a player because he's "inconsistent". This is a guy that has 4 times scored 20 goals in this league and has the 2nd most points on the team through the last 2 seasons!

The point I tried to make was Huselius may be "streaky", but that doesn't make him "inconsistent". If he was more "consistent", he'd be a $7M, 100 point/year (ie a top tier) player.

You're a 9 handicap? Nice.

I played a lot in high school but never took it too seriously, then couldn't afford it for most of my 20's. I took it back up last year and got a hole in one in July! Broke 100 a few times, this year trying to get down to the low 90's and maybe under.

PS: Good post.

Ironically, about the only criticism of Tiger lately (meaning last couple of years) has been the occasional 'inconsistency' with the Driver. One quibble I've got with Matt's post is about 18-20 guys wanting to win, I'd agree that they'd all PREFER to win, yet some players obviously want to win more than some others. Why there is this difference in 'want' is as inexplicable as it is obvious. Their need is larger for some reason, probably something to do with how they're wired and the sum of experiences/disappointments/victories they've collected along the way.

I think anyone who's ever played any kind of competitive hockey game has noticed that, that all the little things can wrong no matter how well you ate that day or how much time you spent mentally preparing yourself in the locker room. Passes all miss, your timing is off, the one time your mark gets away from you comes back to bite you.

I haven't noticed any rhyme or reason to why it happens some games, and the complete opposite happens other games.

My only quibble with this post is that I think Roger Federer has been more dominant in his sport than Woods has been in his. I like the rest of it.

Nice post. But I have two quibbles:

1) I think I'd go with Edwin Moses as the most dominant in your lifetime -- perhaps Federer.

2) I'm not quite willing to throw out the concept of consistency just yet, or to equate consistency with better.

While the phrase is certainly overused in the sports world, I still think it is probably the case that coaches are trying to articulate how someone or something needs to get better.

To use a hockey example: there are lots of ways a powerplay can stink. Two obvious ones are that it can stink pretty much all of the time or it can look dynamite sometimes and awful at other times. In the first case, "inconsistency" isn't the major problem. In the second, it is. To just say "our powerplay needs to get better" to describe both contexts seems a little spare to me.

Consistency is only one element of an athlete's overall "talent". Shot mechanics, speed, power, vision, etc. I don't think there is anything wrong with attempting to isolate consistency as in need of improvement, anymore than there is with telling someone they need to work on their shot. Even if consistency is something somewhat more difficult to isolate.

Tremendous post. Point well made.

Hey MF

This is a good topic. Some further thoughts:

One of the strange things about "consistency" is that it seems to come about through two (at least) different routes.

In one sense, consistency is simply a cumulative property of getting better at some particular thing. Shooting freethrows, for example.

For freethrows, consistency in result is the only thing that matters. As such, it is pretty hard to separate "better" from "consistent" and, consequently, to separate consistent from the physical skills of throwing the ball through the rim.

In this case, getting better = becoming more consistent. It is relatively easy (Shaq being the exception to prove the rule) to improve someone's consistency in freethrow shooting by working on their mechanics and making them shoot 1500 shots a day. You and I both did it in 5 days with kids in summer camps.

I may be wrong, but I read your golf example the same way -- consistency is a cumulative property tied to overall improvement in the underlying skillset.

I think it is a very defensible position to extend this to other types of skills or sporting phenomena. But, I'm not sure if it can be extended to all sporting phenomena or to all instances of inconsistency. I think there are other avenues to "consistency" that are not as directly tied to the underlying skillset -- at least not a skillset as conventionally defined.

We can all think players from whom you pretty much know what you are going to get when they step on the field of play. We can also think of other players who you have no clue what they might bring to the table -- sometimes they are world beaters sometimes they are bums (Kovalev?, Vince Carter?).

In the later case, the lack of consistency isn't a product of their lack of skills (at least not physical skills), but rather something a bit more difficult to pin down: confidence, mental skills, concentration, effort, . . . ?

It is those types of players that drive coaches, teammates, fans crazy. How can they be so good one day and so bad the next? And part of that craziness is a product of not being able to pin down exactly what needs fixing. It isn't like the freethrows where you simply need to practice more. It is something else.

I don't really know what it is, but there is some aspect of "consistency" that isn't simply an cumulative property of getting better at something.

Forgive me for the shallow analysis, but there's been a dominant figure in men's or women's (or both) tennis for most of my life, the name simply changes. I'll take anyone's word for it that Federer is the most dominant of the dominators, but then again, there's no particular tournament or course-type that has proven impossible for Tiger to conquer. Perhaps if they played the British Open on clay. As for Edwin Moses, Sergei Bubka, et al... they weren't pros. (Off on a technicality!)

I don't want to throw out the word 'consistency' either, but it's appropriate and sensible as a descriptor for some things, and for others, not.

Sensible: principles; effort; sauce.
Not: performance

It makes perfect sense for a *coach* to talk a lot about consistency, because his team/players can't control their skill level, and they can't control the vagaries of good day/bad day per above, but they CAN control their effort and preparation. And things that his team can control on a day-to-day basis is exactly what a coach should be concerned about.

That said, that's not an excuse for the media or anyone else to play along, especially when they apply different standards to the home team as they do to others (ever read an Edmonton media describe a team below the Oil in the standings as being better than the Oil, but not as consistent? In 30 years?).

It's a variation on George Carlin's bit, the one another commenter brought up in a different context (possibly at Lowetide): ever notice how everyone else's stuff is shit, but *your* shit is stuff?

I too enjoyed the post, and the replies.

Another thing I'd add: it's not just the consistency (in sacamano's two senses) of individuals at play here, and I think that's where the golf analogy--where there's no real team play--starts to break down for hockey. Team play makes the set of variables that much more complex. When the then "Mighty Ducks of Anaheim" defeated the much mightier (in the regular season) Detroit Red Wings in 4 games in the first round of the playoffs, it wasn't because the Red Wings were being inconsistent. In fact, if I remember correctly, the Wings dominated most of the games. Yes, Giguere was amazing, but the Ducks' team play, their trapping, clearing rebounds, etc. was the reason for their success. (Good goaltending is partially a matter of coordination b/w goalies and the defensive play).

So, if you accept that, the issue of consistency becomes more complex, b/c it's not just a matter of the players as individuals, but how the team is constructed as a collective. After all, isn't that why we spend so much time obsessing about linemates, etc.

Which brings me to a question I keep wondering about: why do teams in the NW keep going through long losing steaks, often after long winning steaks, and thus stay bunched together? Minnie, which is for now leading the NW, just went through a 5-game losing streak, and you can find this for almost any of the teams. Is it inconsistent play?

Parity, if that's what we want to call it, it seems to me isn't just about the fact that no team can now keep all their best players b/c of the salary cap. It also seems to be about the fact that all teams use some kind of trap part of the time, and therefore have similar systems to break the trap. Part of that seems to be the "dump and chase" fore-check. Is what seems like inconsistent play in the NW an effect of the mental and physical toll playing those team systems, the trap and the fore-check?

Antro, re: the first half of your comment, I think that makes the golf analogy even more illustrative. If a golfer can be prepared and focused and still turn in a rather lousy performance -- with a lot fewer variables, opportunities for unlucky bounces, and "things that they can't control" -- then the performance of a hockey team, well, shit.

As for the streaks, it's pure chance. IIRC, a look at binomial distribution says that a .500 team has nearly a 50% chance of having a 7-game losing streak at some point during an 82-game season. A .600 team is more likely than not to have a 5-game losing streak. Etcetera.

I'm still not convinced that consistency should only be used to describe effort, prep, etc. and can't be used to describe performance.

As you point out, inconsistent performances happen regardless of preparation, effort, etc. Consistency is something -- but I'm not sure what.

Consistency is something -- but I'm not sure what.

Regarding performance/results, I don't think it is, at least in any meaningful or useful sense.

Imagine a team with a 20-10 record in mid-December, that has achieved it by winning 2 and losing 1 ten times in a row. Would the columns in the local media be praising their consistency? Not a fucking chance -- they'd be furious/befuddled over why the team can't be consistent for more than two straight games.

There's never been a 40-goal scorer in the history of the NHL who scored once every other game, and if there was, pundits would be down on his ability to impact any one given game.

Consistency is a meaningless concept when it comes to results (per those two examples), and it is very nearly meaningless when it comes to effort & prep, because it is so hard to identify inconsistency there without using results as the context.

I usually agree with Mr. Fenwick on 90% of what he says, but on this point I can't disagree more. There is certainly a difference and a playing "better" and playing more "consistent".

To use his golf analogy, if a player with a 9 handicap shoots 100 rounds a season, and not a single one is outside of the range of +8 to +10, he does not need to improve the "consistency" of his game, he simply does not have, or has not yet developed, the skills to shoot better. If a 9 handicap shoots 100 rounds, with half his rounds below or at par, and half of them at +18 or so, he obviously has the skills to be a scratch golfer, he simply needs to execute those skills all the time, or more consistently. He doesn't necessarily need to get "better" he just needs to play at or near his ability more often.

In the hockey sense, it comes down to those individual simple moments that players just need to do all the time. Some defencemen are not gifted at sliding along the blue line to get the shot through to the net, others are. Where a defenceman has that skill, but only half the time does it successfully he doesn't need to get "better" in the sense of developing the skill, he just needs to get more consistent in its use.

What has been infuriating watching the Flames this year is that for spans of time they seem to be able to do everything right - and then they can't seem to make even the simplest plays. In particular they seem particularly unable to make simple clearing plays out of the defensive zone on a regular basis. This isn't a question of being able to make these plays, but to simply do it,for lack of a better term, consistently.

As to MG's analysis, I think consistently inconsistent is a perfect way to describe Huselius over his career or a season, but there is no doubt that game to game Huselius does not execute at the same skill level all the time. This isn't a lack of physical gifts, but a matter of mental focus or willingness.

anoncanuck, I appreciate you stating the other side so clearly, but I think it's wrong.

The fact that I occasionally hit a ball 315 yards down the middle, or birdie back-to-back holes, does not make me an inconsistent version of Tiger Woods. The repeatability of the skill is so utterly integral to the concept of the skill itself, it's a joke to characterize it any other way. Every centreman in the NHL wins at least one-third of their faceoffs... to say anything that implies that the 40% guys are equally skilled, but simply less consistent, than the 60% guys is absurd on its face.

The other problem is that the two extreme examples of the 9 handicap that you use are, uh, extreme... that doesn't happen. And complaining about the inconsistency of the 2nd guy is, in reality, a complaint that he should be a lower handicap than he is, not a complaint about the variability of his scores per se. Which is my point.

As for the streaks, it's pure chance. IIRC, a look at binomial distribution says that a .500 team has nearly a 50% chance of having a 7-game losing streak at some point during an 82-game season. A .600 team is more likely than not to have a 5-game losing streak. Etcetera.

I guess where I agree with sacamano is the assumption that a .500 hockey team *is* such a hockey team right from the beginning of the season. As opposed to *becoming* a .500 hockey team over some interval (say the 20 game segments that you often post). I don't dispute your observation, and in fact I like it, and will adopt it. But it only captures part of the picture, let's call it the "static" aspect--yes, given what players *are* (measured w/ all the fancy statistics of timeonice and behindthenet, etc), if they were more "consistent", they would be "better." Here, like you point out, the two words mean practically the same thing. Iginla is a better player than Tucker b/c he is consistently able to execute at a higher level, and we see that season over season, game after game.

But there is a "dynamic" aspect too, and I think what sacamano is pointing out is there are real changes over the course of the season, especially for something as complex as a hockey team. Some of these are clearly individual: there are injuries, there are young players that actually improve, etc. And then there are changes that affect team play: a line starts to "click," the power play improves and therefore becomes more consistent, the D start clearing rebounds better due to a shift made by the coaching staff. And the individual changes interact with the team changes: a certain star gets injured, say on a fore-check, and the team weakens.

So that's why I still wonder if the hot-and-cold play of the NW teams might have to do with the team systems used to either enforce or break the trap, which involve so much constant scrumming and hitting along the boards. There's only so much your players can do before their level of intensity starts to lag, and the small and large injuries start to pile up.

I don't have time to check, but there have been unbelievable amounts of injuries to the Wild, Canucks, Avs and Oilers this year, and you gotta wonder if it has to do with the prevailing style of play.

If there's one thing a semester of sport psychology has taught me, it's that much-derided concepts like "clutch" and "consistency" are perfectly legitimate measures of an athlete or a team of athletes -- just not the way sportscasters and the athletes themselves use the term. That line from The Princess Bride springs to mind: "You keep using that word. I do not think it means what you think it means."

I think you can accurately characterize a player like Alexei Kovalev as inconsistent on a career-scale basis, because some years, he looks like a top-flight player (this year, for example), but others, he looks like he'd rather be at his dacha with a couple of honeys (last year, for example). I think your argument about consistency being a load of bunk is predicated on the MSM definition and (ab)use of it, and not what I would consider the actual definition of it. Specifically, I think talking about "consistency" with regards to a discrete skill, like face-offs or free-throws, and "consistency" with regards to an overall performance (e.g. "Dustin Penner has been inconsistent this year.") are two entirely different things. The former is blatantly obvious horseshit, because as you note, repeatability is the skill, but I don't think a larger-scale application of the concept is inherently flawed. I mean, Penner has been inconsistent this year. For the middle 40ish games of the season, he (largely) looked like a legitimate top-six forward, even if he clearly wasn't driving the results. For the 15 before and 20 since, he's largely been invisible. There's obvious reasons for it, but that doesn't change the fact that, by definition, it is inconsistency. (I also object to the characterization of streaks and streakiness as "pure chance." Players don't flip coins to decide whether they're gonna play like shit or not. I'm know that's not what you meant, but that's what the statement implies, when the truth is, there's tons of perfectly obvious, real reasons for a given streak, one way or the other.)

As it pertains to the Flames specifically, my prof, based on reading the quotes they give, wonders if it's not a problem of poor mental preparation, and not knowing how to address it. I mean, it's not like they suddenly lose their physical skills from one night to the next. Not having your head in the game, on the other hand, is an excellent way to come out flat and not play to your potential. And Mike Keenan, despite his mellowing with age, doesn't strike me as the sort who'd hire a sport psychologist to teach his boys about visualization and positive self-talk and stuff like that.

As for the streaks, it's pure chance. IIRC, a look at binomial distribution says that a .500 team has nearly a 50% chance of having a 7-game losing streak at some point during an 82-game season. A .600 team is more likely than not to have a 5-game losing streak. Etcetera.

Yeah, this is an argument that is often repeated, and is technically incorrect. It is common for people to take the results of statistical tests as a demonstration that a given result was "due to chance".

In fact, no test can tell you that a result is due to anything. What they can tell you is that a given result is not distinguishable from chance.

These are two different animals. There are all kinds of reasons a coin can turn up heads (chance, it is weighted on one side, etc.). The identical results cannot be used to interpret identical causes.

There are myriad reasons why teams go on 7 game losing streaks, and it seems a bit unimaginative to chalk it all up to "pure chance". Having said that, perhaps it doesn't matter since the point of these analyses isn't really to find out what causes the losing streaks, only to demonstrate that most teams have them.

But perhaps antro is onto something -- why isn't there more analysis of the factors that contribute to losing streaks (injuries, scheduling, sex the night before games, etc.).

Last one before watching the Oil whip the Avs.

Lots of people run stats on events to see if the results are "statistically significant", which generally means, "distinguishable from chance".

But, much of the time, "chance" is simply unrealistic as a causal factor.

If I punch 5 Flames fans directly on the nose, and it turns out that 5 noses break and 5 don't, the tests will tell me that potential for one of my punches to break a nose is indistinguishable from chance.

But it would be pretty damn premature, even nonsensical, to say that it was "chance" that decided, which noses broke and which didn't. How about we first look at how hard each punch was thrown, how big the nose was, bone density, the angle of impact, etc. If we can control all of those variables, then maybe we can start talking about "chance".

I'm not convinced anything whatsoever has been controlled in the "pure chance causes 7-game losing streaks" argument.

Why can't Oiler fans punch consistently? Tune in tomorrow.

"Imagine a team with a 20-10 record in mid-December, that has achieved it by winning 2 and losing 1 ten times in a row. Would the columns in the local media be praising their consistency? Not a fucking chance -- they'd be furious/befuddled over why the team can't be consistent for more than two straight games."

This paragraph brings it home for me more than the original post did. No one can play their best ALL the time. As much as I personally dislike the idea, December was not the "real" Flames any more than November was.

Incidentally, why do people isolate the NW for hot and cold streaks, or inconsistency? Which division (or even team) has been consistent, or avoided hot and cold streaks this year?

Why can't Oiler fans punch consistently? Tune in tomorrow.

Believe me -- I'd make sure to find the sweet spot on your nose.

Jeezuz I wish Roloson was more consistent this year!

And Stoll. And Reasoner. And Penner.

Well written, Matt.

At a player level, I do think that young players do tend to appear more inconsistent though, just because they tend to take of lot of ill-advised chances.

So if a talented rookie forward keeps trying to beat guys at the blue lines, some nights it will go wrong a bunch, and his team will pay the price. Other nights he will get away with it, and show up on highlight reels and in the three star selections.

As he gets a bit more experienced, and hopefully learns to respect his teammates and coaches, he'll play the odds a little better, and help his team more, and probably be a bit less spectacular to watch, even though he's improved his skills.

Of course any fair person arguing that a young player "just needs to improve his consistency" should accept that the decrease in crap nights goes hand in hand with a decrease in wonderful nights.

Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?