Nylon Calculus: Did the Warriors really ruin the NBA?

It’s all starting to feel a little unfair, isn’t it? They already had four All-Stars — six if you count veterans Andre Iguodala and David West — and now they’re adding Boogie Cousins, too? C’mon, what’s the point of even watching, anymore? The Warriors can’t lose.

That’s right, I said it: “The Warriors ruined the NBA”.

I’ve been thinking about ways to test this assertion, about ways to visualize the ruination of the NBA. And then I saw this great viz of in-match win probabilities from the World Cup created by Rachael Dottle at FiveThirtyEight. It’s a great chart because it reveals all the excitement of the tournament — all the twists and turns. In fact, it’s such a great chart that I just HAD to rip it off.

That’s the NBA version of Dottle’s viz for the 2016 playoffs. Instead of charting in-game swings in win probability, my version shows the game-by-game fluctuations in win probabilities over the course of each series. All the pre-game win probabilities come from Nylon Calculus wizkid, Jacob Goldstein, and are based on his multi-year Player Impact Plus-Minus stats, which use luck-adjusted on-off data and boxscores to derive each individual player’s value.

Goldstein used the previous iteration of his win-probability model to predict 2018 postseason outcomes. His new-and-improved model projects playoff team ratings by assuming that only the top-seven players in each team’s rotation will have an impact in the postseason, providing better estimates of a team’s true strength than using regular-season ratings. Using this approach, Goldstein found that the average projected rating for playoff teams since 2001 was plus-2.2 points per 100 possessions, with the highest rating belonging to the 2017 Golden State Warriors, at plus-17.1.

In the chart, the broadest swaths of color equate to series that were never in doubt — ones where the winning team’s chances remained close to 100 percent throughout. Look at Cleveland’s wine-colored path, for example. The Cavaliers victory in the Eastern Conference was inevitable, so those blocks were nearly all red. In contrast, if one of the blocks is split more evenly between two colors — like in the 2016 Finals, for example — that shows that the outcome was less certain. These were the most exciting matchups of the 2016 postseason. We can leverage this feature of the chart to make a graphical representation of playoff fun — a Postseason Excitement Index.

To evaluate the Postseason Excitement Index, we condense the playoff bracket down and re-size the blocks for each round. The first round series are shown as eight blocks that are each one-inch square, the conference semifinals are four blocks that are two-inch square, the conference finals are two blocks that are both four-inch square, and the finals are one eight-inch square block. Next, we re-color all of the winners’ odds as white and all the losers’ odds as black. In the end, the bigger the black shapes, the more excitement. Specifically, the Index is equivalent to the percent of the chart that is covered with black. The Index credits situations when the underdog won, series that went the full seven games, and high-stakes excitement. The Index is basically a weighted average of the four playoff rounds that weighs each round equally and, thus, individual series in later rounds are weighed more heavily than series from earlier rounds.

With the year-by-year chart of the Postseason Excitement Index, you can see how 2016 compares to other recent postseasons — it was uniquely exciting. I think this is a pretty useful visual, but it’s kinda distracting, right? I can’t help thinking that they look a lot like cow hides. Weird. But, I guess we might as well embrace it. You know…milk it.

Now that’s a really moo-ving viz.

You can read this version the same as above — the black shapes are proportional to the Postseason Excitement Index — it’s just that, now, the plots are cow-shaped.

So, what can we conclude? Did the Warriors really ruin the NBA? Let’s hustle our cows out to pasture and see what they show us about the trends in the Postseason Excitement Index over the past decade, from 2009 to 2018.

From 2009 to 2015, the Postseason Excitement Index bounced around a narrow range — from 19 to 28. Then, in 2016, two blown 3-1 leads imbued our playoff cow with large black spots and a resulting Index of 36. Lots of excitement.

Then, after Kevin Durant came to Golden State, the Index plummeted to a ten-year low of 11 as the Warriors dominated the subsequent 2017 postseason, losing only one game on the way to the title. In 2018 — on the back of two tightly contested seven-game Conference Finals — the Index rebounded to a respectable score of 19 (although the ultimate outcome was the same, another Warriors championship).

So, yea, the Warriors DID ruin the NBA, but hopefully only for a year. There should still be plenty of entertainment to go around next year, especially with the post-LeBron East now up for grabs.

On the other hand, if you’re hoping to re-live the feverish postseason excitement of 2016 —  you might find yourself waiting until the cows come home.