In analyzing game-by-game performance, don't forget the role of luck
As basketball fans, we tend to over-analyze our team's performance from game to game, particularly on the defensive end of the court. But much of the variation we see in the outcome of games can be attributed to simple luck. Over the course of the season, luck evens out and the teams with the best offenses and defenses establish themselves by their record of performance. But in any given 3 or 4 game sample, luck can easily be a big enough factor to obscure much of what we think we're seeing.
Take for example the last two Blazer games. In Houston, the team gave up 111 points to the Yao-less Rockets and were lambasted for their defensive performance. In Oklahoma City last night, the Blazers held the Thunder to just 74 points and received praise for their defensive effort. While I think it's probably true that the Blazers played marginally better defense last night, luck certainly amplified the difference.
The primary (but not exclusive) way that luck manifests itself in basketball is in an opponent's outside shooting percentage, which is best represented in the traditional box score by the 3 point percentage. While it's certainly true that good defensive play (such as getting hands in shooter's faces, not leaving good shooters open, etc.) can reduce an opponent's outside shooting percentage on average over the course of a season, for individual games, much of the variation in shooting percentage is due to luck. Sometimes shooters are hot and other times they aren't. Over the course of a season, 3-point shooting percentages across the league fall within a relatively narrow band. Last year, the team with the best percentage (Boston) made 39.7% of its threes. The worst team (PHI) made 31.8%. (Portland was 4th at 39.3%). Defensively, the range is similar. The best team (Cleveland) allowed opponents to shoot 33.3%, while the worst team (SAC) allowed opponents to shoot 40.6%. (Portland was 19th at 37.4%).
So even the best defensive team allowed opponents to make a third of their threes, and the best offensive team made only 40% (just 7 percentage points higher) . What that means is that, in an individual game, when a team makes 50% of its threes (as Houston did the other night) or 20% of its threes (as OKC did last night), much of that variation is attributable to luck, not the defense. Had either Houston or OKC had an average night shooting from downtown, the outcomes of both of those games might well have been different.
When a team is hot from outside, it has a huge effect within the game. Defenders are forced to play farther out which opens up space and allows for easier dribble penetration and interior scoring. When a team is cold from outside, the opposite happens. Defenders cheat back and clog up the middle. They intercept more passes and are better able disrupt penetration and post scoring. Good defenses can suddenly look terrible when the other team is hitting threes and mediocre defenses can suddenly look brilliant when the other team's shooters go cold.
That's why you want to be really careful when assessing a team's performance on a game-by-game basis. Your judgment, given the small sample size, can easily be skewed by the role of luck. The reality is that Portland wasn't as bad on defense against Houston as the score and stats suggest and Portland wasn't as good defensively against the Thunder as the score and stats suggest. This is why you play a whole season.
The season is still very young and we don't have a whole lot of data to work with yet. We should get a better sense of where the Blazers are both offensively and defensively after another week or so.
8 recs |
45 comments
| Add comment
Comments
Indeed
We probably win the Denver game without the Oden FTs if one of the following happens:
1. Any of our ridiculous number of layup misses roll in.
2. Melo’s shots rim out and not in.
"Good evening Blazer fans, wherever you may be!"-Bill Schonely
by skywaker9 on Nov 2, 2009 9:06 AM PST reply actions 0 recs
Are you making the argument that we win the game if our shots go in and the other team's dont?
Blazer Fan
by leeroyjenkins on Nov 2, 2009 9:22 AM PST up reply actions 0 recs
In terms of getting the rolls
Yes, which is luck.
"Good evening Blazer fans, wherever you may be!"-Bill Schonely
by skywaker9 on Nov 2, 2009 9:39 AM PST up reply actions 0 recs
Nice post
Good offensive play and/or poor defense can lead to more open looks from distance, but often it does just come down to whether shooters are in rhythm or not.
I can tell you how to make an Excel spreadsheet that proves Portland wins 62 games this year.
by jscot on Nov 2, 2009 9:17 AM PST reply actions 0 recs
Some teams get close-in baskets and that makes their 'rythm' a lot more consistent
Blazer Fan
by leeroyjenkins on Nov 2, 2009 9:23 AM PST up reply actions 0 recs
But that isn't the point
The OP is arguing (correctly) that a big part of the story is outside shots going in or not. That was the difference in Houston, they were really hot from outside. The difference in OKC is they weren’t. That’s not close-in baskets that made that difference.
I can tell you how to make an Excel spreadsheet that proves Portland wins 62 games this year.
by jscot on Nov 2, 2009 9:28 AM PST up reply actions 0 recs
Preseason games are games
I know that because the team’s performance was bad during the preseason we want to pretend those games don’t exist, but they anytime your team is on the floor vs. another team the games count and they should be included in your sample.
10 or 12 games is 1/8 of an NBA season.
Blazer Fan
by leeroyjenkins on Nov 2, 2009 9:22 AM PST reply actions 0 recs
p.s. what's funny is that if the team went undefeated in the preseason and through 4 reg season games people would be going nuts
about how amazing the team was. You can’t have it both ways.
Blazer Fan
by leeroyjenkins on Nov 2, 2009 9:24 AM PST up reply actions 0 recs
Um, no, preseason games don't count
Never have, never will.
Don’t count in the standings, don’t count in the stats, which is why the coaches do silly things with their lineups, rather than actually making it a focus to win the game.
I can tell you how to make an Excel spreadsheet that proves Portland wins 62 games this year.
by jscot on Nov 2, 2009 9:30 AM PST up reply actions 0 recs
+92
"And in the end
The love you take
Is equal to the love you make." -The Beatles
by 92wastheyear on Nov 2, 2009 10:46 AM PST up reply actions 0 recs
Maybe so, but he's right about people going nuts in they were undefeated.
by Stryder9 on Nov 2, 2009 1:56 PM PST up reply actions 0 recs
People went nuts when we lost in preseason, too
If the definition of “counts” was “people go nuts over it”, we’d be in big trouble. Especially if it was “people go nuts on the Internet over it”.
I’ve seen people go nuts over Skittles. I’ve seen someone go nuts because he thought someone looked in his direction.
I can tell you how to make an Excel spreadsheet that proves Portland wins 62 games this year.
by jscot on Nov 2, 2009 10:52 PM PST up reply actions 0 recs
Very good point!
I am actually not all that concerned with our defense. What does concern me is that on offense we remain a perimeter oriented team and even more so last night with LMA out for much of the game. The only inside game is on dribble penetration from Roy and occasionally Miller. Thus, our poor shooting made the game much closer than it should have been. We need to develop Greg Oden into more of a consistent low post threat on offense! Last night he took only 5 shots, and 2 or 3 of them were on offensive rebound putback dunks. That is crazy!! He should be getting around 10 shots a game from the low post. We need to go through the inevitable growing pains now and it will pay big dividends down the road later in the season and post season. Otherwise we are destined to the same fate as last year.
by socalblazer on Nov 2, 2009 9:22 AM PST reply actions 0 recs
I agree about Oden.
the consistent inability for more than one or two perimeter guys on this team to give Oden a good entry pass is disconcerting. It was the same last year. Now, when Oden does get those passes, he hasn’t looked great with the ball, I’ll admit. But how is he gonna get better if they don’t consistently give him a look. Seemed there were a few times in the OKC where they were fronting Oden and our passers just gave up on getting him the ball even though there was no one playing between Oden and the basket. Is that really too hard a pass to make?
by DrivetheLane on Nov 2, 2009 9:33 AM PST up reply actions 0 recs
Regarding fronting Oden
I saw the fronts and though and easy was to pass Oden the ball would be to just bounce the pass off the backboard. That would be a simple pass over the defender that leaves Oden room to score. It may be ever easier than an alley-oop.
Anyone have any ideas why we never see such a pass when a defender fronts in the deep post?
by NWfan on Nov 2, 2009 9:40 AM PST up reply actions 0 recs
Interesting
I’ve had the same though before. But I’ve literally never seen anyone do that. It must be harder to execute than it seems conceptually.
www.blazerguy.com
by Blazer Guy on Nov 2, 2009 9:45 AM PST up reply actions 0 recs
Angle of Incidence = Angle of Refraction
If you’re off to the side and hit the back board, the ball will bounce off the backboard towards the other side of the court. If you’re head-on with the backboard, I think a pass might not be a bad idea, but I bet the timing needs to be perfect for an allyoop.
by Anim8rguy on Nov 2, 2009 10:27 AM PST up reply actions 0 recs
I've seen it twice
But both times (one was McGrady in the allstar game back in his Orlando days) the player bounced it off the backborad to himself. McGrady’s attempt was succesful the other one (i don’t recall who attempted it) was not.
by NWfan on Nov 3, 2009 9:25 AM PST up reply actions 0 recs
Scorers would rate it a missed shot on the passer and no assist
Regardless of the degree of difficulty it is a double negative stat play.
"Either way we have two phenomenal units. I'm excited to play with either one." - Martell Webster
by lee3022 on Nov 3, 2009 12:51 AM PST up reply actions 0 recs
Luck also comes into play on loose balls.
Especially against Houston the other night, I saw the Blazers work hard and get deflections on defense, only to see the loose balls end up in the hands of the Rockets for easy buckets.
by Benjamanic on Nov 2, 2009 9:46 AM PST reply actions 0 recs
Rebounds as well
Good rebounders will always out-rebound bad rebounders if you make the sample large enough, but within a quarter or in a series of key possessions, the ball can take unluckly trajectories off the rim and/or backboard and end up back in the hands of the other team. That happened late in the Denver game. Denver got quite a few lucky rebound in th 4th quarter.
www.blazerguy.com
by Blazer Guy on Nov 2, 2009 9:53 AM PST up reply actions 0 recs
Yup.
If a few of either bounce our way, the post-Denver or -Houston stories take a narrative of: “They were sloppy but showed the resilience/tenacity that makes them an elite team, etc.” rather than “what’s happening to the Blazers?”.
by Benjamanic on Nov 2, 2009 9:56 AM PST up reply actions 0 recs
I think the whole Ginobli catches a bat thing was total luck
"And in the end
The love you take
Is equal to the love you make." -The Beatles
by 92wastheyear on Nov 2, 2009 10:29 AM PST reply actions 0 recs
I don't know
He looked like it was something he’d done a thousand times before. Like it was second nature. He didn’t look at all surprised when he succeeded.
www.blazerguy.com
by Blazer Guy on Nov 2, 2009 10:37 AM PST up reply actions 0 recs
Maybe bat catching is big in South America. hmmm
"And in the end
The love you take
Is equal to the love you make." -The Beatles
by 92wastheyear on Nov 2, 2009 10:38 AM PST up reply actions 0 recs
elementluck
never forgotten
Enjoy the Ride
by DigitalDaggers on Nov 2, 2009 3:38 PM PST reply actions 0 recs
Excellent post...
Good points about luck having a factor in a small sample size, and also about the effect on the opponent’s D resulting from hot or cold outside shooting …
I’d like to expand on your these to say that luck has a large factor to play over a larger sample size as well.
I mean, last year, if the end of that last SA game of the year hadn’t gone against us, we wouldn’t have played Houston. That was about as unlikely a break as you can get. Last year, we got lucky with a lot of back-to-backs. As teams vary in how well they are playing, you can be fortunate playing good teams when they are cold. You can get a ref who likes your team, or maybe one who doesn’t.
The timing of your and your opponent injuries have a huge effect, and are largely attributable to luck: we were lucky JR was injured the other night, just couldn’t capitalize. We may even be lucky that Nic is injured early, so Martell can show his stuff, to reduce our minutes logjam, and hopefully so he can be available for good D in the postseason.
In short, I’d rather be lucky than good… but good teams can also make a lot of their own good fortune… And the good usually beat the lucky and not so good… So hopefully, the Blazers can become good and lucky…
Timing is everything, and a lot of that is luck.
by Visionary2 on Nov 2, 2009 4:31 PM PST reply actions 0 recs
I agreed
BTW JR is suspended..
"And in the end
The love you take
Is equal to the love you make." -The Beatles
by 92wastheyear on Nov 2, 2009 4:35 PM PST up reply actions 0 recs
I'd rather be good than lucky
I can tell you how to make an Excel spreadsheet that proves Portland wins 62 games this year.
by jscot on Nov 2, 2009 10:54 PM PST up reply actions 0 recs
Luck usually evens out over time
I can tell you how to make an Excel spreadsheet that proves Portland wins 62 games this year.
by jscot on Nov 2, 2009 10:54 PM PST up reply actions 0 recs
As the old saying goes, "The harder I work, the luckier I get."
The best teams usually overcome bad luck, which we quickly forget, but we remember their good luck because they won.
by BlazerFanSince1970 on Nov 2, 2009 4:46 PM PST reply actions 0 recs
This is a good post but it leaves a question or two
Attributing deviations from season average to luck alone would be statistical insufficiency I think. For example what are the standard deviations each for good teams and bad teams?
If a team hits 67% of its shots against losing teams and 33% of its shots against winning teams its average will be 50% but would individual games within teams in these categories establish a different expectation for them? What about home and away?
Luck to one person might mean hot/cold to another. Is this the same thing? Teams go through hot and cold streaks. Are all of these luck?
Your well-reasoned post simply suggests that some of it is luck and I have no argument. But my read of the implication made me want to subtract the low average from the low shooting and call the rest luck and the same for the high. In actuality good defenders when motivated get statistically better results. The averages include “playing down to opponents” by good teams aptly documented elsewhere. Portland was desperate for the win last night and was working really hard. So I want to give credit to that as well as find the comments by Martell about not working hard enough in Houston as some evidence of the bad defense contributing, Am I off track?
"Either way we have two phenomenal units. I'm excited to play with either one." - Martell Webster
by lee3022 on Nov 3, 2009 1:14 AM PST reply actions 0 recs
whether or not "luck" is the right term...
…the reality is that much of the other team’s performance on the shooting end is beyond the control of your defense. It may not be “luck” in the sense of random occurrence, but it is luck in the sense that it is beyond the your control. You’re right that averages can mean different things. For instance, if a team routinely shoots 50% from long range against bad teams and 30% from long range against good teams, the average will be 40%, but the variation would not be random. But, as far as I can tell, this is not the way shooting percentages fluctuate. Even the worst teams in the league generally hold opponents under 40% from downtown. While the level of defensive play is a factor, it does not appear to be the only or even primary factor causing variation in 3-point shooting from night to night.
www.blazerguy.com
by Blazer Guy on Nov 3, 2009 6:42 AM PST up reply actions 0 recs
I dont think the evidence you are giving supports your claim. I would say that you would want to look at the standard deviation of the 3ptFG% rather than just looking at averages, as that might tell you about luck i.e. a team or player that shoots 30% over their career, but then shoots 50% for the game. Also, the percent difference between 33.3% and 40.6% is nearly 18%, which is a non trivial difference and shouldnt be dismissed as merely 7%.
Life is exhausting when you are this stupid.
by jonestr on Nov 3, 2009 10:06 AM PST reply actions 0 recs
True enough
In order to make my point ironclad, I would need to marshall evidence re: standard deviations to show just how abberrant various shooting percentages are. But, sadly, I don’t have anywhere near the time to do that. Plus, I think most people’s eyes would glaze over. Moreover, I think it is a fairly reasonable assumption that an opposing team’s 3 point shooting percentage is not entirely defense-determinative, i.e., there are factors outside of a defense’s control that influence how well the other team will shoot on any given night (I used the term “luck” to describe these factors). It is undoubtedly the case that sometimes other teams’ shooters are hot or cold for reasons that have nothing to do with how well you are playing them defensively. This can obscure, at least in a small sample size, the true picture of how well you are playing defense.
www.blazerguy.com
by Blazer Guy on Nov 3, 2009 11:27 AM PST up reply actions 0 recs
I forgot to link to the this article about the myth of the hot hand
http://espn.go.com/blog/truehoop/post/_/id/6241/hot-and-heavy-about-nba-shooting
Life is exhausting when you are this stupid.
by jonestr on Nov 3, 2009 5:29 PM PST up reply actions 0 recs
I've read that too
but I’m not sure it’s relevant to what I’m saying. That deals more with the question of whether a player’s shooting within a game is predictive of his continued shooting in the game, i.e., if Durant makes 5 in a row, is he more likely than usual to make his next shot. Regardless of the answer to that question, though, guys have good shooting nights and bad shooting nights. One night Durant will be 12 for 15. The next he’ll be 3 for 20. My point is that the difference between those two nights isn’t determined solely by the quality of the defense. External factors, beyond the control of the defense, affect shooting.
www.blazerguy.com
by Blazer Guy on Nov 3, 2009 6:25 PM PST up reply actions 0 recs
if you have to rely on luck
to decide if you win or lose……. then….well….. Good Luck, You’ll need it….but even though you make a spotty case for luck, it doesn’t hold up over time……
You remarked about the 3 point shot ……it could be said that the 3 point shot is what changed the game into what we see today (we now reward players for taking the easy way out and also a theoretically lower percentage shot)…..When the NBA decided to introduce the 3 point shot, it changed the way teams attack on offense and also more of a need to develop perimeter defense…..So the 3 point shot adds a new dimension to the game of basketball and could also give merit to the degree of luck involved…..
That said, it’s all about skill, good fundamental teamwork, hustle and desire that usually determine a game’s outcome…..you should never put yourself in a gambling mode to begin with….
by WyEast on Nov 3, 2009 11:04 AM PST reply actions 0 recs
I think you missed my point
I’m not saying that basketball is a game of chance. It’s not. But in a small sample of games, luck can play a big enough role to obscure the picture. Over the course of a season, luck evens out and the teams differentiate themselves. My point was only that it’s hard to judge the Blazers after just 4 games, especially when there was some fluky shooting involved in at least two of the games.
www.blazerguy.com
by Blazer Guy on Nov 3, 2009 11:22 AM PST up reply actions 0 recs
I understood your point
I was actually trying to elaborate (on a side note), to try and keep luck at the right spot (in the outline of basketball)…it seemed to me, you may have inadvertently, put luck in the wrong position in the outline….but considering it was your main topic, maybe I’m wrong….most likely, we have the same opinion of where luck fits in……
My judgment of the Blazers never gets concluded with the play of one or two games….it usually rears it’s ugly head as a trend (based on a much larger sample size)…however, I would have to rule luck out entirely, on things that are lacking in their play, that has got them into a situation to have to rely on it for a win………or even, blame it for a loss…. even if it’s just one game……in other words, if it equates into either a win or loss on a game then your surely on borrowed time and the players better realize, quickly, it’s not a comfortable way to make a living…..In other words, it’s the same cliche; you make your own Luck
Anyway, your point is well taken, but is subject to a bit of bashing, due to the fact that no one wants this team to be lucky …..or even unlucky….Better to be good (as someone said earlier) To me, this team has shown it is just not ready for the season, and thus, have left some doubts in people’s mind……luck has little to do with it…
by WyEast on Nov 3, 2009 12:30 PM PST up reply actions 0 recs
We probably should've rolled characters with higher luck.
I say a trade is in order.
--
by CaptainSexyJacob on Nov 3, 2009 4:31 PM PST reply actions 0 recs
There always after me Lucky Charms!
What is the cliche, hard work makes its own luck? No, “luck” is just too hard to quantify. Yes, a player can hit a flurry of 3 pt shots in a row, maybe the defender is right in his face, maybe they still go in. Is that luck? Maybe the converse is happening and the player is wide-open and missing, is that bad luck? You can just never know.
No in competition you can never rely on luck or depend on luck. It’s just like life, luck might step in and help you, any time, any place, but you are best served by ignoring it, and working to not need it.
As far as I go with aknowledging luck is saying I think every season there are a percentage of games inwhich you play well enough to win and should of won, that you lose, and also vice versa. But overall? You don’t play to be lucky. You play to win.
"Mother Nature started this fight, I think it's about time we ended it!"
by Krang on Nov 4, 2009 9:35 AM PST reply actions 0 recs
I'm not giving advice to the team
I’m trying to give perspective to fans. Obviously you can’t control things that are out of your control, so you just play hard and smart and trust that luck will even out over time. My point was only that fans often read too much into a handful of games without appreciating the role that chance played in the outcome.
www.blazerguy.com
by Blazer Guy on Nov 4, 2009 10:56 AM PST up reply actions 0 recs
Last night was sure unlucky
I hope our luck turns here soon
Blazer Fan
by leeroyjenkins on Nov 4, 2009 11:17 AM PST reply actions 0 recs

by 


















