Using Talent Right: Title Contenders Force the Tempo

This is part of a two-part series. For observations on the Spurs  and the Thunder's specific matchup, see 48 Minutes of Hell.

As one of my questions in Monday's Statistical Q&A, I fielded a question from the imitable Tim Varner of 48 Minutes of Hell. His query was whether the Spurs stand to gain in OKC smallball lineups by pushing the pace and playing fast. In short? Yes. I covered that today in detail at 48 Minutes of Hell, but there's a lot of interesting tidbits to be had in this table, enough so that I felt a separate post was necessary analyzing the trends and tendencies of the non-Spurs teams. To examine, I've produced this table that shows the W/L record, the offensive and defensive efficiencies, the eFG%, the efficiency differential, and the free throw rates of our four remaining teams in four distinct buckets of possessions. First bucket includes all games with under 91 possessions; the second includes 91-95; the third is 95-100; and the fourth and final bucket includes super-fast games with over 100 possessions. These are roughly quartiles of possessions. I placed in red a team's "worst" pace and in green a team's best.

Looking at this table, some interesting takeaways after the jump.

• • •

  • BOSTON: Bet you didn't expect this, huh? Out of all the paces the Celtics play at, one stands above all others as the absolute worst they could possibly play at. I refer, of course, to... an absurdly slow game? The Celtics force the pace low by Doc Rivers' own desires -- in the Big Three era, he has always preached a defensive-oriented strategy of keeping as few possessions as possible. This season, though, the Celtics have been bloody awful when they play their slowest. When they have under 91 possessions in a game, the Celtics have a losing record (12-15), a defensive rating WELL above their season average, allow teams to shoot almost 50% from the field, and barely ever draw fouls. On the other hand, when they play to a league-average pace, they're a really excellent team -- a +9 differential, fantastic defense, and a sparkling 16-4 record. Had you shown me Boston's numbers before I did this exercise, I wouldn't have believed it. But it's true. When the Celtics play super-slow, they're a terrible team. Doc Rivers may deserve a bit of blame -- no other team is more inefficient at forcing the tempo that suits the team best, and to some extent, that's on his game plan. Not a full extent, but certainly to some.
  • MIAMI: Little rhyme or reason to the Heat's numbers, though some funny stuff here. They average a differential of +7.1 in games with over 100 possessions, but somehow managed to go 5-4 on those games in the regular season. Which means they won those 5 games by over double the margin they lost by. Absolutely silly. Overall, the Miami defense actually gets a bit better as the pace goes up -- their real problems come on the offensive end. I take back my first statement. This actually makes a lot of sense to me. As a team highly reliant on two players, it stands to reason that there is some sort of upper limit on the number of possessions LeBron and Wade can use up in a single game. The more possessions the Heat use, the more likely that one of those extra possessions is something useless, like a Joel Anthony layup or another bricked Battier three. Thus, their offense gets a bit less efficient as the possessions rack up and they're forced to burn more possessions on their atrocious bench. As a mathematical example, assume LeBron a usage rate of 33%. In a 90 possession game, that's 30 possessions -- in a 100 possession game 33. That means that Non-LeBron players used 60 possessions in the 90 game and 67 possessions in the 100 game. What this means, big picture, is that even if the ratio is the same there are more possessions spent on players you know can't really give you much. In simplified terms... how easy is it for two players to have 50-60 out of 80 points in a slow paced game compared to 70-80 points out of 110 in a fast paced game? It takes more effort, and it takes an increase in a player's usage above and beyond simple extrapolation.
  • OKLAHOMA CITY:  To hearken back to economics, the Thunder are in an odd position of not really having big marginal advantages over anyone in any one area, despite a lot of strengths when averaged across buckets. Their only real weakness is that they simply can't play slow-down, knock out ball the way a team like Miami can -- indeed, the Thunder actually were better than the Spurs at super-slow games, and far better than the Celtics. But against the Heat, that relative strength becomes a massive boondoggle. The, conversely, the Thunder are well above average at a faster-paced game... but still significantly worse than either the Heat or the Spurs! The only decisive advantage the Thunder really have in terms of pace is to play a very normal, league-average 91-95 possessions. My theory is that the Thunder defense gradually breaks down as the game gets faster, but the offense (isolation based and transition-heavy as it already is) doesn't have a second gear that allows it to become more efficient in a fast-paced setting. It's worth noting that the Thunder are fantastic at the league-average play, and in today's 48 Minutes of Hell post, I covered how the Thunder need to get back to their game and a league-average halfcourt pace if they want to get back into the series. I reiterate that here.
  • SAN ANTONIO: This is much like my statement in the Q&A, and by far the simplest relationship. As the Spurs get slower, the Spurs get worse. As the Spurs get faster, the Spurs get unbeatable, improving on both the defensive end AND the offensive end. They also shoot better, which speaks to Chip Engelland's yeoman's work in ensuring the Spurs maintain proper form on quick, set shots and the Spurs added efficiency when they force a transition-heavy, D'Antoni style of play. I discussed this a bit at 48 Minutes of Hell, so I won't belabor the point. But really: the Spurs are great when they play fast, and more than any other team left, they're the best at dictating the tempo and forcing teams to play fast. A deadly combination, that.

• • •

Long story short, I just mathematically proved why I don't want to watch the Heat-Celtics series! Therefore, I'm going to sleep through it. You know how I do. If that offends you as a reader, then I suggest that you link to YOUR mathematical proof in the comments. I promise I will read it and review it, and perhaps rethink my stance.

... Nah, probably not.

10 comments on “Using Talent Right: Title Contenders Force the Tempo

  1. Pingback: Great Exercises in Internet NBA-Related Postings 5-30-12 | Hardwood Paroxysm

  2. Great post. I know a link from True Hoop is nice, but their wholesale copying of your analysis, even if attributed, is lame.

  3. Really interesting analysis - there's a lot of really cool information there.

    I'm reminded of a study I saw a while back that took the basic statistic about number of possessions in the game and instead broke it down to average possession length for both teams in the game. IIRC, it showed (among other things) that the Spurs were the best team in the league at baiting other teams into taking early shots. Since seeing that study, I've always yearned for pace statistics to be broken down into the two sides of the ball for an individual game - it just makes so much more sense! I'd be interested in seeing what happens when you combine the two methods of analysis - do the Spurs win when they take quick shots, or when their opponents do?

    ST

    • Fantastic point. Okay, so I've built a model that calculates schedule-adjusted two-way four-factor data, and the one-sidedness of pace is actually kind of glaring. You'd need play-by-play data to do it certainly, but even something as simple as "proportion of time with possession" or something would be extremely helpful and powerful for understanding how teams force and are forced into style.

      I suppose it could something as simple as the sum over all time difference terms like "time_(opp's def_reb/TO/team's FG) - time_(team's def_reb/TO/opp's FG)". That is, you put markers at all the changes in possession and count each interval between the markers, and then compare the sums, divide by total time, and mutliply by 100 (so that the result is a percentage).

      So teams that have longer possessions would have percentages like 55-60, and teams that have short possessions more like 45-40. And we could see, for example, if teams that have 55-60 that play one another actually decrease the pace demonstrably, as we'd expect.

  4. Any statistics from this year's condensed and abbreviated schedule are outliers. Any statistics that take into account regular season back-to-back-to-back games in which one of the oldest NBA teams is analayzed (Boston) are meaningless. The truth is, the Celtics do play at their best when Rondo is pushing the tempo, attacking the defense. That doesn't mean it's always possible to achieve this. You can't push tempo with this Celtics team unless the defense is getting stops and Rondo and company can beat the defense down the court. You also can't push tempo when 3 of the 5 players on the court are running on fumes. You also can't push tempo when your roster is so thin that you go 8 deep.

  5. Pingback: Atlanta Team Report, 2013: Don’t Trust the Hawks | The Gothic Ginobili

  6. Pingback: Philadelphia Team Report: What’s the Endgame? | The Gothic Ginobili

  7. Pingback: Spurs Link-And-Roll - 5.31.12 - Air Alamo - A San Antonio Spurs Fan Site - News, Blogs, Opinion and More

  8. Pingback: A Thunder storm is brewing: How do the Spurs stop it? - Air Alamo - A San Antonio Spurs Fan Site - News, Blogs, Opinion and More

  9. Pingback: MySynergySports: How should the Spurs utilize Tony Parker? - Air Alamo - A San Antonio Spurs Fan Site - News, Blogs, Opinion and More

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>