Nylon Notebook: Who has been the MVP of the playoffs?
In a week we’ll be spending a lot of time talking about the NBA Finals MVP but first, let’s look at who has been the best player in the playoffs. We can measure this in via a few different ways — there’s Jacob Goldstein’s PIPM, where Draymond has been the best player on both a per-possession basis (PIPM) and a total wins added basis; there’s BPM, where Jokic has been the best player on a per-possession basis and tied with Kawhi on a wins added basis (VORP) and finally, we can look at two other metrics — Kevin Ferrigan’s DRE and Nathan Walker’s SPR– which can help us calculate overall value but also how each player has done in a series or an individual game (it’s worth noting these are based on box score stats and so we’re going to be missing a big chunk of the game — mainly defense).
Full data here.
So that’s a lot of numbers and why are there five different versions?
If you want the detailed explanation, click here. Long story short: there are two different versions that are opponent-adjusted that are calculated differently, there’s another version that’s unadjusted but sum’s up to the team’s net rating and there’s the GameDRE/SPR that is the raw score (and doesn’t sum up to the team’s net rating). Finally, I added another unadjusted version that uses expected assists instead of actual assists (expected assists are calculated using the player’s regular season assist percentage — assists/potential assists — and multiplying that by the number of potential assists within that game).
As we can see, Kawhi is leading in all versions of DRE and SPR. What if we look at who has had the best individual series?
Full data here.
Kawhi’s series vs. the Sixers is the best individual series when including opponent adjustments. But if we don’t include opponent adjustments, Jokic had the best individual series vs. San Antonio.
Finally, let’s look at who has had the best individual games.
Full data here.
Again, we see that Kawhi has also had the best individual game (Game 1 vs. Philly). And in this case, even without the opponent adjustment, it still comes out as the best individual game. He’s also had either the second or third best game as well (Game 2 vs. Orlando).
Simply put, Kawhi has been simply sensational in this playoff run. Whether it’s having the best overall playoffs, the best series or the best game, he’s routinely at the top.
How about Pascal Siakam’s Game 1 in the Finals? He was 14–of-17 and finished with 32 points, 8 rebounds and 5 assists and yet surprisingly, his game one was only a top-25 game in the postseason (which to be clear, is still very good).
The lucky and the unlucky
A large portion of generating assists is hoping your teammate makes the shot. A player can try to put that teammate in an excellent position to make the shot but it’s still out of their control. So which players in the playoffs have gotten unlucky with their assists (this is relative to their regular season assist percent which was defined earlier as assists/potential assists)? We can measure this by looking at the total DRE (or SPR) loss which is simply the unadjusted expected DRE of the player (as mentioned earlier, expected DRE uses expected assists) minus the unadjusted DRE of the player. Here are the five most unlucky players in the playoffs:
Eric Bledsoe had a really rough playoffs but he was also pretty unlucky. If his teammates’ are hitting shots on his assist opportunities at the same rate that they hit them during the regular season, Bledsoe would’ve had about 17 more assists in the playoffs, which translates to slightly over one more assist per game.
Of course, this might not tell the entire story. Perhaps he wasn’t getting unlucky and was actually creating worse shots for his teammates than he did in the regular season. This might show up in a shot quality stat for his potential assists (unfortunately, this stat isn’t publicly available).
How about the five luckiest players in the playoffs:
The Conference Finals were really close
Let’s take a look back at the Conference Finals. While the Toronto-Milwaukee series went six games and was considered very competitive, the Warriors swept the Blazers in four games. In most sweeps, the series is not competitive. However, if we evaluate the series in terms of who controlled it (percent of offensive possessions leading vs. trailing), Portland was actually the better team. They spent the majority of their offensive possessions with the lead while Golden State had the lead for just 38 percent of their offensive possessions while trailing on 55 percent of their offensive possessions. Of course, Golden State was able to consistently erase leads in this series but that doesn’t mean it was easy. Another consistent hallmark of Golden State is they usually spend a good portion of their offensive possessions up by 20+ points (against the Clippers, they spent almost 15 percent of their offensive possessions up by 20+ while they were the regular season leader in percent of offensive possessions up 20+) but against Portland, Golden State spent just two offensive possessions up by 20+ (0.5 percent for the series).
In the Eastern Conference Finals, the series was very close. Toronto did spend a slightly higher percentage of their offensive possessions leading than they did trailing (48.01 percent to 47.34 percent) while Milwaukee’s splits were slightly less even: 43.40 percent leading and 50.83 percent trailing.
Now, you may be wondering why those numbers aren’t evenly aligned. Here’s why: it’s from the offensive teams’ point of view. So if a Team is down one on their offensive possession and scores two points, they spent that offensive possession down one but then the opposing team is now down one on their offensive possession and will spend their offensive possession down one. In this specific instance, both teams spent their offensive possession down one and therefore were trailing. So this is how you end up with unequal totals.
Which of these two teams was more dominant? Milwaukee spent more time up 10+ and 20+ than Toronto did. Specifically, this is due to Game 2, where Milwaukee spent almost a quarter of their offensive possessions up by 20+. Meanwhile, Toronto was up by 20+ in Game 4 where they spent 7.29 percent of their offensive possessions up by 20+. It was the only game in the series where they had a lead of 20+.
Finally, let’s look at Game 1 of the Finals:
Not surprisingly, Toronto was in control for the majority of the game but it never got out of hand to the point where Golden State was out of it. Does this mean anything going forward? Not likely, the game-to-game correlation for the percent leading/trailing stats is negligible in these playoffs (small sample size alert- this needs to be studied more in depth).
We’ll leave you with this: the game-by-game margin frequencies and series margin frequencies.