Move or die

October 14, 2015 11:21 AM

Don't feel like using Nate Silver's new statistical prediction model CARMELO to figure out if your NBA team will be any good this season? Maybe this fact will help instead: The most important contribution an NBA basketball player can make to their team is no longer thought to be scoring points. Like, at all.

In the past, even the recent past, assists might have been thought mere stat padding, or not as important as they sound. Some might argue the steal or rebound is actually the most important. But in recent times the death of hero ball has been widely proclaimed.Of course, if indeed assists are that important, perhaps they should be incentivized better?

posted by Potomac Avenue (49 comments total)

I haven't fully digested it yet, but that is a beautiful graphic.
posted by benito.strauss at 11:28 AM on October 14, 2015 [4 favorites]

My god when did basketball turn into baseball?
posted by GuyZero at 11:35 AM on October 14, 2015 [3 favorites]

Anybody have a clue WTF an assist is? Scorekeepers don't seem to.
posted by srboisvert at 11:36 AM on October 14, 2015

All the analysis in the world doesn't matter for me, because my team's destiny is written in stone: the Wolves will be exciting on paper for a few weeks until an unlikely combination of injuries takes all of the major pieces off the board and the team then spends the rest of the season tanking. Hard not to be excited about a guaranteed thrill ride like that.
posted by the phlegmatic king at 11:37 AM on October 14, 2015 [2 favorites]

Anybody have a clue WTF an assist is? Scorekeepers don't seem to.

NBA.com: "Basketball U on Assists"

An assist is a pass that directly leads to a basket. This can be a pass to the low post that leads to a direct score, a long pass for a layup, a fast break pass to a teammate for a layup, and/or a pass that results in an open perimeter shot for a teammate. In basketball, an assist is awarded only if, in the judgement of the statistician, the last player's pass contributed directly to a made basket. An assist can be awarded for a basket scored after the ball has been dribbled if the player's pass led to the field goal being made.

posted by Celsius1414 at 11:41 AM on October 14, 2015

My god when did basketball turn into baseball?

Advanced stats have been a major thing in basketball for almost a decade now. (To be a basketball fan is to be the nerd version of a sports fan anyway - you like sport #3, you like the RC Cola of American major league sports, you like the sport that *gasp* black people like. So basketball fans mostly embrace this.)

And it's good that basketball has embraced advanced stats, because what advanced stats have helped players and coaches learn is that teams with a strong passing game and a good inside/out division of offense, a combination of rim-attacking dunks and three-point shooting, have stronger offenses overall, and it just happens that this sort of play is much more aesthetically pleasing to watch than grind-it-out post play and midrange shooting heroball.

Without analytics, we don't get the Spurs playing their beautiful game. So yay analytics.
posted by mightygodking at 11:42 AM on October 14, 2015 [24 favorites]

My god when did basketball turn into baseball?

It's really a natural sport for it; lots of events, including scoring events, means lots of data to collect and analyze.
posted by Bulgaroktonos at 11:43 AM on October 14, 2015 [2 favorites]

Actually this analysis counts assists as any pass that leads to a basket within 2 seconds. "Shots coming within two seconds of a pass attempt are considered potentially assisted, while shots not preceded by a qualifying pass are considered unassisted." It's less strict because he's looking not at individual player rating but the more broad positive effects of passing.
posted by Potomac Avenue at 11:51 AM on October 14, 2015

And it's good that basketball has embraced advanced stats, because what advanced stats have helped players and coaches learn is that teams with a strong passing game and a good inside/out division of offense, a combination of rim-attacking dunks and three-point shooting, have stronger offenses overall, and it just happens that this sort of play is much more aesthetically pleasing to watch than grind-it-out post play and midrange shooting heroball.

This has been my favorite thing about the current analytics consensus in basketball. The fact that pretty basketball is also statistically superior basketball just makes me happy.
posted by protocoach at 11:54 AM on October 14, 2015

it just happens that this sort of play is much more aesthetically pleasing to watch than grind-it-out post play and midrange shooting heroball.

It's actually kind of astounding that of the major US sports, only baseball ended up worse aesthetically (well, at least to me) for all of the analytics over the last 10 years. You mentioned basketball, but analytical football has a lot more quick passing and going for it on 4th downs and the hockey teams that have dipped their toes into analytics are generally leaning toward smart, fast, and skilled guys even if they don't score and away from giant dudes lumbering down the ice hoping to stand in front of the net for a while. Three true outcomes (i.e. strikeouts, walks, and home runs) baseball is less fun than watching guys hit triples and steal bases, but at least the home runs are still fun.

I'm not saying I've never enjoyed a low-scoring slugfest, but in general I'd rather watch games with a lot of ball/puck and player movement, and analytics has so far been encouraging that, which is awesome.
posted by Copronymus at 11:56 AM on October 14, 2015

Counterpoint: DINGERS!
posted by protocoach at 11:59 AM on October 14, 2015 [1 favorite]

Why do you thing Gregg Popovich has been jumping up and down on the sidelines screaming "MOVE THE FUCKING BALL!" for the last umpteen years?

He is a serious potty mouth. But he's been way ahead of the curve on the assist thing. You swing it side-to-side enough, or you move it inside/out, you're gonna catch a defender leaning the wrong way sooner or later.
posted by Devils Rancher at 12:05 PM on October 14, 2015

It's actually kind of astounding that of the major US sports, only baseball ended up worse aesthetically (well, at least to me) for all of the analytics over the last 10 year

I guess hockey isn't really major, but in some ways it ended up a lot worse. I mean, "low-scoring slugfest" is in part what analytics has done to hockey. Soooo much board-grinding and high-percentage shots from as close to the crease as possible, etc. We are seeing the slow death of the stay-at-home defenseman and the useless punchman who can't skate because of it, and that's great, but the Kings are even more stats darlings than the Hawks (who have the advantage of generational talent), and the Kings hockey is unbearably dull. I kind of love them, but my god that's boring hockey. The Bruins were dull as hell, too. I think the Hawks wouldn't be nearly as interesting if they hadn't had as good luck with the draft as they did; they miss out on Keith and Kane and they'd be dull defensive grinders, too. So basically every Cup winner since 2010 has been either incredibly dull, defensive grind hockey, or just really lucky they actually won the rights to Patrick Kane instead of the Flyers. I'm not buying it. We're basically back to dead puck, just with faster, better-skating dudes.
posted by The Master and Margarita Mix at 12:05 PM on October 14, 2015 [2 favorites]

When shooters create shots on their own, especially midrange jumpers, they usually produce low-value attempts.

Did the Knicks know this when they signed Anthony for a billion dollars?
posted by bukvich at 12:09 PM on October 14, 2015

There are two (to me) vastly different narratives here. One is about strategy (who shoots, when/how/where, passing etc), and a second trying to calculate a WAR for basketball. I'm ignoring the WAR, but am interested in the strategery part.

So if stats say "do this play/take these steps, and based on historical/modern stats, you'll do better", where's the next disruptive coach to come in and break that strategy ?

(I'm thinking this is smoke/mirrors, Jordan was one era, the "big men" another, this is current, and I'm pretty sure there will be another one)
posted by k5.user at 12:10 PM on October 14, 2015

(Bringing hockey up reminds me that the way the neutral zone trap - booooooring hockey - was eliminated through rules changes. So there's that avenue as well, besides a disruptive coach)
posted by k5.user at 12:15 PM on October 14, 2015

Analytics are mixed on Carmelo; he's inefficient and likes to take midrange jump shots, but there's evidence that he makes his teammates better, despite using a lot of possession and not really passing much.
posted by Bulgaroktonos at 12:17 PM on October 14, 2015

CARMELO has neither a creamy caramel center nor beautiful views of Big Sur. Not for me.
posted by oneswellfoop at 12:18 PM on October 14, 2015

IMO they should move the 3pt line closer. Fk it just make more 3s. Everyone loves 3s more than dunks at this point even.
posted by Potomac Avenue at 12:21 PM on October 14, 2015

Hockey's main problem is that the players are so much bigger (seriously, check out game footage from the '80s; they look like high school kids compared to players these days), but the rinks are the same size. There's no real solution to this, other than going to 4-on-4 all the time (unlikely, the players union would fight it) or making the rinks larger (unlikely, seats would have to be removed, which would cost the owners money).
posted by The Card Cheat at 12:29 PM on October 14, 2015 [1 favorite]

IMO they should move the 3pt line closer. Fk it just make more 3s. Everyone loves 3s more than dunks at this point even.

You're thinking too small.
posted by tonycpsu at 12:30 PM on October 14, 2015 [3 favorites]

The NBA moved the 3-point line from 23'9" to 22' in the mid-90s. It resulted in a ton of made 3s, but overall play suffered because defenses had less ground to cover and thus clogged the the lane.
posted by Turkey Glue at 12:36 PM on October 14, 2015 [3 favorites]

^Whoa that is a strong point.
More #hoopideas: Send the 2 worst teams to the D league every year. Eliminates tanking. But also just permanently eliminate 2 teams. Fewer games = fewer injuries & more of an advantage to the offense who are less tired.
posted by Potomac Avenue at 12:41 PM on October 14, 2015

Analytics are mixed on Carmelo; he's inefficient and likes to take midrange jump shots, but there's evidence that he makes his teammates better, despite using a lot of possession and not really passing much.

If nothing else, he provides his teammates with more opportunities to get offensive rebounds.
posted by Ufez Jones at 12:43 PM on October 14, 2015 [1 favorite]

Not news to anybody who watched the '77 finals, in which two dreadful losses to the immensely talented 'Sixers (featuring Doctor J in his prime plus other greats) made the Bill Walton led Trailblazers look like the most overmatched finals team in NBA history -- until some kind of switch got flipped at the beginning of game 3, and the 'Blazers went on to absolutely crush the 'Sixers with four straight exhibitions of the most dazzling team basketball anyone will ever see.

But that wasn't where the league wanted to go, and various changes, many of them in the way officials called the game, made team basketball as unrewarding on the scoreboard as the league decided it would have been on the bottom line.
posted by jamjam at 1:15 PM on October 14, 2015 [2 favorites]

Just for the record, Doctor J was my favorite player and the 'Sixers were my favorite team.
posted by jamjam at 1:17 PM on October 14, 2015 [1 favorite]

We're basically back to dead puck, just with faster, better-skating dudes.

It's 100% possible that because the team I watch most is the 'Hawks I have a skewed perspective. And Lord knows those Kings and Bruins teams could indeed be pretty dire. Hockey analysis is pretty young, though, and there are enough up-and-coming teams that look good both aesthetically and analytically like the Lightning and Islanders that I'm at least cautiously optimistic. Plus, just getting more faster, better skating dudes into the league is at least a little bit of a win.
posted by Copronymus at 1:22 PM on October 14, 2015 [1 favorite]

My god when did basketball turn into baseball?

NBA teams have been investing heavily in analytics for years now. The big difference from baseball, and the reason it has snuck up on some fans, is that unlike baseball, basketball analytics were developed internally for the most part. Teams keep their formula secret and proprietary
posted by thecjm at 1:55 PM on October 14, 2015

There's also not as much time for announcers to talk about a basketball player's individual stats since they don't have long periods at bat or on the mound. They also might be more difficult for a layman without a series of cameras and data collection software to gather what is being measured.
posted by Space Coyote at 2:05 PM on October 14, 2015

Team basketball has always been the best, it's just that now it looks like this.
posted by Potomac Avenue at 2:58 PM on October 14, 2015 [1 favorite]

I mean, "low-scoring slugfest" is in part what analytics has done to hockey

It's really not? Analytics have only very recently begun to gain any wide acceptance in hockey, they certainly haven't been around long enough to have had a huge impact on scoring. Nor are they ever likely to, they don't really change the nature of the players on the ice that much, other than to point out how useless talentless goons and "stay at home Dmen" (which is code for slow with no puck moving skills) are, as you mentioned.

The low scoring in hockey is due to the increased size of goalie equipment and goalies, better goalie play and the slow but consistent drop in the number of penalties called since the lockout.

As it stands, the league is significantly faster than it has ever been, and an increased emphasis on banning head shots and other outright goonery has allowed smaller, skilled players to enter the league and stick (think Johnny Gaudreau) which has improved the quality of play.
posted by Maugrim at 3:00 PM on October 14, 2015 [2 favorites]

It's actually kind of astounding that of the major US sports, only baseball ended up worse aesthetically (well, at least to me)

I'm struggling to see how baseball has changed at all. Even with the new statistics, better understand of the game, and thus more accurate (in theory) valuing of individual players, the game itself seems remarkably similar to what it's always been. Have the games themselves actually changed because of analytics?
posted by cell divide at 3:22 PM on October 14, 2015

Have the games themselves actually changed because of analytics?

Look up the prevalence of the infield shift for one example, and they're moving outfielders around differently as well. But it's becoming huge in other important ways.
posted by Celsius1414 at 3:32 PM on October 14, 2015 [1 favorite]

Have the games themselves actually changed because of analytics?

I think he is referring to the fact that three true outcome (walk, strikeout, homerun) hitters are being prioritized over the speedsters of the past. However, I think that's actually reversing itself a bit now as baseball people are realizing that speed often means good defense so you don't have to do that much more to be valuable (e.g. Billy Hamilton, who while not an all-star, is a solid contributor).
posted by john-a-dreams at 3:37 PM on October 14, 2015

Not sure if anyone mentioned this but I'll assume not. I've been compiling a list of things to improve all the major sports in America. The one for baseball is extensive and very cool. The one for basketball is proven true by this article. I had an advantage, though, I played competitive basketball through college (not for a college but in very competitive intramurals and pickup games). The one big thing is that extending the shot clock will increase scoring in games. Don't double it but add like 6 seconds. This will allow teams to pass the ball a few more times and thus set up higher percentage shots. It's that simple.

Intuitively people assume that having a shorter clock will increase scoring and excitement but as this article points out indirectly what it really does in increase the number of poor shots. So it might be somewhat counterintuitive but I think it holds.
posted by bfootdav at 4:08 PM on October 14, 2015

However, low shot percentages lead to more fast breaks. Which are fantastic, of course.
posted by oddman at 4:41 PM on October 14, 2015 [1 favorite]

College ball finally cut their shot clock down from 35 to 30 for next season, so that will be a data point.

My intuition is that adding 6 points to the pro clock would not raise scoring. If it makes it easier to score in longer possessions, strategies will shift and more possessions will get longer. The pace will slow and total scoring will go down.

There are tons of college teams who play down-tempo. Their offenses are often efficient, but almost everyone hates the slow pace and low scores, even the ones who appreciate efficiency.
posted by nom de poop at 4:49 PM on October 14, 2015

However, low shot percentages lead to more fast breaks. Which are fantastic, of course.

This is true so I guess there is an aesthetic concern at play here: better quality half-court offense (and maybe defense) or more fast breaks? And of course what would be the optimum combination?

One of the problems in the NBA is that offensive style doesn't seem as prominent of a situation as it was with the Lakers's run'n'shoot style v the Celtics's half-court game. Part of this is the 3-second rule for the lane (also, cf with international ball with its crazy lane size and the effect that has had on diminishing inside play) which I think should be reconsidered (not abolished but maybe extended?)

But the big thing is that given the stats this article discussed and how the mid-range jumper improves with more passing/assists then it seems to me that anything that improves passing will improve all areas of the game and might even lead to more distinct styles of play (which I think makes the game more interesting to watch).

There's something about game balance and the need for play-testing and beta periods.
posted by bfootdav at 4:49 PM on October 14, 2015

My intuition is that adding 6 points to the pro clock would not raise scoring. If it makes it easier to score in longer possessions, strategies will shift and more possessions will get longer. The pace will slow and total scoring will go down.

This might happen and would definitely happen if we lengthened the clock by, like, 12 seconds. But something small is all you need to get in an extra pass or two and that's all you need in order to create a better shot. When I played in school my teams always saw our scores go up when we actually listened to the coach and passed a few extra times. Defensive players cannot physically keep up with the speed of a moving ball therefore space is created. Basketball is entirely about having separation. If you get an extra few inches separation between you and your defender then scoring is easier. That's what Jordan and Iverson did with their insane quickness and athletic ability. An extra two passes gets the same results (roughly).

College ball finally cut their shot clock down from 35 to 30 for next season, so that will be a data point.

I didn't now this! (I don't actually follow professional or college sports anymore except for La Liga and the CL.) That will make for an interesting data point. Of course there are lots of differences between college and pro ball and the lengthening the shot clock is one of those things of diminishing returns (in my theory) so it won't be conclusive but it will be something.
posted by bfootdav at 4:57 PM on October 14, 2015

The Celtics and I have a relationship almost entirely founded on screaming "pass it, pass it, nononono don't take that shot!" That and wanting to braid Kelly Olynyk's hair.
posted by Biblio at 5:44 PM on October 14, 2015

" There's something about game balance and the need for play-testing and beta periods."

Voila, the D-League. The League and individual teams are using it as a big lab. I bet going to D-League games is a lot of fun for stats-head NBA fans.
posted by oddman at 6:35 PM on October 14, 2015 [1 favorite]

It think if something gives an advantage, strategies will shift to include it more. Maybe there's a shot you can get at 15 seconds that you'd take when there's only 9 seconds left, but you skip when you have a whole 15 seconds more... Maybe standards for what's a "good look" just go up to account for what you can get in 30 seconds instead of 24. So pace goes down and any increase in efficiency might not be enough to compensate.

But yeah, who knows.

College ball has actually been losing pace and scoring since 1991 (graph from kenpom.com). People are getting anxious about it, and that's what motivated the shot clock change: force more possessions to increase pace. But lots of people are worried the shorter clock might reduce efficiency so much that scoring doesn't go up. College basketball has also been getting progressively more physical over the years, and the theory is that offenses have been slowing down simply because they've been having to work harder to find good shots.
posted by nom de poop at 7:02 PM on October 14, 2015

As a Pistons fan (historical), I'd love this since part of what broke the Pistons was the reffing in favor of hero ball — all the way back to Jordan's extra steps.
posted by klangklangston at 7:52 PM on October 14, 2015 [1 favorite]

Without analytics, we don't get the Spurs playing their beautiful game yt . So yay analytics.

Didn't the article say that Popovich had a hunch rather than analytics?
posted by srboisvert at 8:45 PM on October 14, 2015

Analytics have contributed a lot to basketball but they are a *lot* less predictive and conclusive than baseball analytics. The dynamics of team play and chemistry are just too subtle and complex -- I would go so far as to say that the computer geeks have 'cracked' baseball but they are a long way from doing so with basketball if they ever do. The Atlanta Hawks looked like the perfect modern analytics team and they won over 60 games in the East last year, but the Cavaliers with Kevin Love out and with Kyrie Irving barely playing swept them in the ECF with ease. In the final analysis, greatness is still sort of ineffable in basketball (e.g. Lebron James semi-visible but pervasive influence in that series).
posted by zipadee at 4:51 AM on October 15, 2015 [1 favorite]

The Atlanta Hawks looked like the perfect modern analytics team and they won over 60 games in the East last year, but the Cavaliers with Kevin Love out and with Kyrie Irving barely playing swept them in the ECF with ease.

Even a "cracked" game is going to have problems with randomness and small sample sizes in playoff series.
posted by Bulgaroktonos at 5:37 AM on October 15, 2015

Also not having Thabo Sefolosha hurt Atlanta more than missing Love hurt Cleveland.
posted by oddman at 4:14 PM on October 15, 2015

Based on things I've read and other peoples' reactions I personally think that the Spurs are going to dominate this season. Watching LaMarcus Aldridge's highlights from their game against the Heat it's clear that they're offense is going to be massive and that their players already have such great chemistry together.

As for my city's team, the Portland Trail Blazers, I think they're going to surpass all expectations. I am not sure if they'll make the playoffs but I'm not going to count them out.

I am so excited for this season. I feel like every season has been getting better and better.
posted by gucci mane at 5:02 PM on October 15, 2015 [1 favorite]

My god when did basketball turn into baseball?

It should be so lucky.
posted by grubi at 11:16 AM on October 16, 2015 [1 favorite]