I knew it was going to happen, but I still couldn’t stop myself. It was the fourth time in my career I had prepared a playoff analysis, the fourth time I had spent days diving deep into our data to try and find any kind of edge, and the fourth time I could feel my perspective begin to shift. Objectively, I knew we were underdogs — the Rockets had home court advantage, and very few outside observers were picking us to win — but this research process always made me overestimate our odds.
I wouldn’t start the process that way, of course. In fact, quite the opposite. I’d initially see only our opponent’s strengths: only Houston’s lockdown defense in 2009, only the red-hot Suns’ offense in 2010, only the impossibility of guarding 2011 Dirk in any action he was involved in. And in 2014, did we have a prayer to stop the Rockets’ open-court combination of speed and power?
But every team has relative weaknesses. Maybe they’re vulnerable on the defensive glass, or have trouble guarding side pick-and-rolls with their power forward, or don’t defend the post well. And as I’d start to uncover those weaknesses, they’d stick with me. I’d temporarily forget that actually translating all of this to the court was a major challenge in-and-of-itself, and that the opposition would have their own counters to what we’d try. Instead I’d feel optimistic, probably irrationally so, that the work we did as an analytics group, as a coaching staff, and as an organization would pay off.
So it was, that, in 2014, despite having my bubble of irrational confidence quickly burst in each of my previous three playoff series, I was still optimistic. This time, though, my optimism was rewarded. That series against the Rockets was the last I was involved in (I would leave for Philadelphia in the offseason), and it was one for the ages: three overtime games, each of the six contests close near the end, and (who could forget?) Damian Lillard’s walk-off shot. Though it was unique in some ways, that series still illustrates three important ways that the playoffs are different from the regular season: they are more intense, they are more strategic, and they are more exhausting.
The Playoffs Are Intense
I don’t think I’m going to create controversy by claiming that the playoffs are intense. In fact, I’d imagine most people think that seems pretty obvious. But obvious doesn’t mean uninteresting. As Benjamin Morris of FiveThirtyEight wrote last year in a fascinating analysis of the playoffs: “…[T]he NBA playoffs are a different beast entirely. There, many of the traditional ways of thinking — like the philosophy that the playoffs are where winners win and where champions win championships — turn out to be true.”
The playoffs are different enough from the regular season that they allow us to quantify aspects of the game that are usually hard to measure — aspects that are sometimes dismissed because they are hard to measure. Just as “knowing how to win” shows up in the data, so does playoff intensity. And how, exactly, that intensity impacts the play on the court is revealing.
Offense is often described as being about talent, skill, and technique. Defense, by contrast, is talked about in terms of effort, toughness, and grit. While sometimes this gets pushed too far — of course offense requires effort and defense requires skill — that characterization comes from something real: defense is more about effort than offense is. One way to measure this is to see how play changes during back-to-backs, since fatigue should have a big impact on effort. And indeed, it turns out that being on a back-to-back impacts defense more than offense. While a team’s offensive efficiency declines by about 0.5 points per 100 possessions when they are on a back-to-back and their opponent is not, a team’s offensive efficiency increases by about 1.5 points per 100 possessions when they are rested and the defense is on a back-to-back.1
The intensity of the playoffs seems to have the same effect. Effort is ramped up and that means better defense: since 2003, when the NBA switched to a best-of-seven series for all rounds, teams have tended to score about 1 fewer point per 100 possessions in the playoffs than the regular season, after adjusting for the stronger defenses they face in the playoffs. That’s despite the fact that free throw rate has gone the opposite direction.2 The playoffs are so much more intense and physical than the regular season — the concept of “no layups in the playoffs” has been around for a long time — that even though the refs may allow more contact, free throw rates still have gone up.
While the reasons home court advantage exists are debated, whatever is causing this advantage gets stronger in the playoffs. Home court advantage is worth about 3.5 points per 100 possessions during the regular season (after accounting for the fact that home teams play fewer back-to-backs), but in the playoffs it’s worth around 5.5 points. I’d argue that this, too, stems from increased playoff intensity: the in-arena atmosphere of a playoff game is markedly different from most regular season contests. It’s something you can feel — the air seems to have a different energy running through it, the anticipation and passion of 20,000 fans is palpable. And the data indicates that the electric crowds change the performance of the players (and perhaps the refs as well).
The playoffs are so intense because of a combination of the stakes being so much higher and the margin for error being so much smaller. A loss in the regular season matters, but there’s still time to make it up. Though a best-of-seven series is the longest we play in the major American professional sports, it’s short enough that one loss can have a major impact on your chance of winning the series. Consider that for two evenly matched teams, at the start of the series the team with home court advantage should have around a 55% chance to win. If they win the first game those odds jump to 2 out of 3; if they lose they fall to 1 out of 3.3
That means one play, even a fluky one, can have an enormous impact on the outcome of a series. For example, take this possession from Game 1 of our 2014 series against the Rockets. After an improbable comeback in regulation, the game had gone to overtime, and Houston had jumped out to a quick six point lead in the first minute. We were in dire straits — according to Inpredictable’s win probability estimate, we had only a 15% chance to win the game at that point. And then, this:
Aldridge, who had hit 3 three-pointers all season, drops the ball off his leg and ends up with a contested, fadeaway three with 4 seconds on the shot clock…and somehow it finds bottom. Inpredictable’s estimate says that after the shot went in we had about a 25% chance to win, while if we had come up empty on that trip it likely would have been below 10%. A circus shot ends up as one of the biggest reasons we won the game and likely changes the entire series. The margins in the playoffs are that small.
If the stakes are high and the margins small in the playoffs in general, Game 7s, truly win or go home, are the playoffs boiled into one game. So we’d expect all of these effects to be enhanced further, and that’s exactly what we see. Offense declines even more in a Game 7: instead of the 1 point dip we see in other playoff games, it falls by around 3 points per 100 possessions. Home court advantage is also much stronger in a Game 7: rather than the 2 point per 100 possession bonus we see in other playoff games, in Game 7s we see around a 5 point bump. That means for two evenly matched teams playing a Game 7, the advantage is so strong that we’d expect the team with home court to be favored by over 8 points!4 Lillard’s series winner was so clutch not only because it sealed the series, but because if it had missed the mark we would have headed to Houston for Game 7, where, as these numbers show, we would have been heavy underdogs.
Effort in the playoffs is also captured in something I call the “desperation effect”. When teams feel their backs up against the wall, they play even harder. After we won Game 1 in Houston, the Rockets knew that losing a second game on their own court would put them in an extremely bad spot. Assuming both teams were evenly matched, a second home loss would drop the Rockets to just a 1-in-7 chance of winning the series. On the other hand, after winning Game 1, we didn’t feel nearly the same sense of desperation. Conventional wisdom said that by taking one on the road we had done our job, and anything else was just a bonus. Knowing these two mentalities, which team would you expect to come out playing harder in Game 2?
It turns out this difference in mindset and intensity can be found in the data. Since 2003, 73% of home teams won Game 1. This is a high number, but it makes sense: home teams not only have home court advantage, they’re also usually the better team, since they earned home court by having a better regular season record. Of the home teams that won Game 1, 73% went on to win Game 2. But of those that lost Game 1, 83% went on to win Game 2. This gap is especially surprising given that there is selection bias at work here: the teams that won Game 1 are generally those teams who have more of an advantage over their opponent, and therefore should be more likely to win Game 2. But instead we see the opposite — and by a large margin.5 Desperation seems to fuel effort, and that gap in effort makes a real difference.6 Because the playoffs are intense.
The Playoffs Are Strategic
I wrote in my introduction to playoff coverage that regular season pre-game preparation “pales in comparison to what can happen when you have the resources of the whole organization behind your game prep. Combine that with facing the same opponent again and again over a best-of-seven series, at least a day off between each game, and multiple practices wholly focused on the specifics of opponent sets and tendencies, and you end up with a style of play that depends more on the strategic elements of the game.”
Heading into the series against Houston we had identified a few different areas we thought were particularly important. The clearest one was our transition defense. All season long transition defense had been a major weakness, as well as one of the most important indicators for whether we’d win or lose a game, and here we were facing the best transition offense in the league.
As an analytics staff we wanted to get a little bit deeper into answering the question of how much to emphasize transition defense and in what situations. We were the 4th best offensive rebounding team in the league that season, and there’s a tradeoff between transition defense and offensive rebounding: the more you pursue the ball on the offensive glass the more compromised a position you’re in if you don’t get it. Would we be taking away one of our strengths by being too concerned with getting back on defense, and on balance end up hurting ourselves?
These are the kinds of in-depth, team-level strategic questions you can dig into before a playoff series that you might not have time to examine before each game in the regular season. In this case we analyzed the data from our games against Houston that season, from all of our games and from all of Houston’s games on the year, and everything seemed to point in the same direction: we needed to get back in transition.7
We’d do the same thing on the player level, attempting to tweak coverages and sets to minimize each opposing player’s strengths and expose their weaknesses. James Harden was obviously a focal point of ours, and so we spent a lot of time trying to determine what we could do to slow him down. Again, this is something you can do during the regular season, but not to the same extent. With the extra practice time to prepare, we could drill certain situations we thought we’d see frequently and the players could get used to what we wanted to do.
One small example of this was something one of our players had observed: part of Harden’s success at drawing fouls is his incredible hand strength. He often exposes the ball on drives, luring defenders into instinctively swiping down on the ball to get the steal, but has the grip to maintain control and use the swipe to draw the whistle. So as part of practice we drilled retreating and absorbing the drive with hands straight up, resisting any urge to come down on the ball, something Wesley Matthews executes perfectly here:
Houston didn’t have many complicated sets or a particularly deep playbook — their offensive success was based on spacing, timing, and talent. That didn’t mean it was easy to guard, by any means, but it did give our staff the opportunity to spend time determining the best way to guard their common sets and our players the ability to thoroughly prepare for these sets. Meanwhile, we knew Houston would know all of our sets and would have their own creative ways to try to throw us off. Take another look at that three Aldridge hit in overtime of Game 1, and notice how the Rockets knew exactly what play was coming:
Beverley sees the play call, signals it out to his teammates, and Lin then knows that Aldridge is heading to set a down screen for Batum, his man, so he jumps into a deny position. Howard, meanwhile, anticipates the post up and does his best to push Aldridge’s post position further out. This happens in the regular season too — sometimes defenders will see the play signal, mimic the motion, and the coaches will yell out what’s about to happen (“down screen to post up!”). But in the playoffs, with the extra prep time and the frequency with which you see the same sets, the players know what’s coming much more often. That means that offensive execution is more difficult, but it also means there’s real opportunity for the coaching staff to add in little wrinkles and catch the opposition off guard.
The Rockets had come out in Game 1 mainly defending Aldridge with smaller players, rather than putting Dwight Howard or Omer Asik on him. Aldridge had responded by aggressively powering his way to the rim, to the tune of 12/16 shooting around the basket and 46 total points. He took advantage of the ability to back his man deep into the paint on plays like this:
Heading into Game 2 we thought the Rockets would adjust, so our staff was prepared. When Houston came out with Howard guarding Aldridge, we put LaMarcus in pick-and-pops, forcing Howard to guard on the perimeter.
Watch these three consecutive possessions from early in Game 2. Each time we ran the same set, each time the team read the defense and took what was open. Notice that this is all initially set up by the threat of Aldridge’s pick-and-pop jumper.
The adjustment worked. Aldridge’s shot was falling, and he ended up scoring 43 in the game: only 5 for 8 around the rim, but 13 for 20 on jumpers.
Houston, meanwhile, knew they had success posting up Howard against Robin Lopez, so they went after that matchup right from the start. Their first five possessions of the game were five straight post ups, and five straight buckets:
Coming into the series we were aiming to take away their three point shooters and make them beat us one-on-one, and so we had decided to play Howard straight up. A start like this, though, makes you question that. Do we change how we’re covering Howard in the post? Or simply try to execute our current coverage better? How a staff decides whether or not to adjust in a situation like this is a fascinating topic in its own right, and something I’ll likely write about soon. But for now suffice it to say that these are the kind of back-and-forth tactical battles you see everywhere in the postseason. Because the playoffs are strategic.
The Playoffs Are Exhausting
Much has been written lately about the toll an NBA regular season takes on players. Wesley Matthews has, for my money, one of the best descriptions of the trials players face over the course of the year:
You reach a point where it’s like, ‘What am I gonna do?'” Matthews said. “It’s the fourth game in five nights. I’ve got to wake up, watch film again. I just had a bad game. My ankle hurts. I’m exhausted. My girlfriend’s blowing my damn phone up. My mom needs this. I’m late on my house payment. This charge didn’t go through. Someone else is hitting me up for money. On top of that, I’m sick, I can’t breathe, I’ve got snot coming out my nose — and I’m in Cleveland.”
That’s one kind of exhaustion: exhaustion born of withstanding the constant grind. But the playoffs create a different kind of exhaustion, exhaustion born of putting everything you have into a game and a series, and then, to advance, having to do it over and over again. If the regular season is a marathon, the first round is a sprint you have to run immediately after completing a marathon — knowing that if you win, you’ll have to run another sprint right after.
When we emerged from the Houston series there wasn’t much time to enjoy it. We immediately had to turn our attention to the #1 seeded Spurs awaiting us in San Antonio. And though the Spurs had just come off a seven game series, the difference in energy between the two teams was clear from the start. There was no doubt they were the better team, but that alone didn’t explain how we ended up facing a 20-point deficit in the second quarter of each of the first three games. Our players just looked like they were moving at a different speed than the Spurs. San Antonio’s execution was sharp, their cuts crisp, their energy apparent, while we slogged through our sets and seemed a step behind.
That experience drove home to me just how much endurance a team needs to continue advancing in the playoffs. That to keep raising your level of play to meet each of the increasingly talented opponents you face is a challenge unique to this time of year. Because the playoffs are exhausting.
Oh, and I’ll add one more descriptor to the list…
The Playoffs Are Exhilarating
Where else can you ride an emotional roller coaster like this? (With apologies to Houston fans.)
Intense, strategic, exhausting and exhilarating. And they start this weekend. Can’t wait.
- This is based on a regression predicting offensive efficiency in a game controlling for each team’s season-long offensive and defensive efficiency marks, home court, rest, and whether the game is a regular season or playoff game. Other numbers in this section are also pulled from this regression (or ones set up similarly). ↩
- Again, after adjusting for how much both the offense and defenses tended to draw and commit fouls during the regular season. ↩
- These numbers are based on plugging probabilities into a simulation, rather than looking at what’s happened in history. It’s much harder to pull this kind of information from the historical data since we don’t actually know the true quality of each team. That said, the actual results might be different since, as we’ll see, additional factors like Game 7s and the “desperation effect” also change these numbers. ↩
- All of these Game 7 effects are strong enough that despite the small sample of Game 7s they still show up as statistically significant. ↩
- We can also measure this through regressions attempting to predict winning percentage or efficiency of teams in Game 2, accounting for the quality of both teams. In each case, whether the team won the previous game shows up as a significant predictor of performance in Game 2. ↩
- Somehow we managed to dodge this effect in 2014, and we headed back to Portland owners of a 2-0 series lead, one of only 10 teams to have won the first two games on the road in the last 13 years. ↩
- It might not be clear from this description, but it’s important to note that these analyses are almost always a dialogue with the coaching staff, rather than a one-way presentation. Whether they start because of a question you have or a question a coach has, they’re best when you can relate the initial findings and get feedback. The staff might think of something you didn’t account for, or have follow-up questions that further refine the conclusions and help with implementation. ↩