Looking back at the 2017 NBA playoffs, I still get chills thinking about how wildly unpredictable the championship race felt that year. As someone who’s spent over a decade analyzing basketball statistics and playoff odds, I can confidently say that season was one of the most challenging—and thrilling—to forecast. Most analysts, myself included, went into the postseason with what we thought were solid predictions, only to watch the narrative flip on its head time and again. And you know what? That’s exactly what makes playoff basketball so compelling.
When I first ran the numbers in April 2017, the Golden State Warriors were the obvious favorites, sitting at what most models projected as a staggering 92% probability to win the Western Conference and around 65% to take the title. But here’s where things got interesting: the San Antonio Spurs were quietly sitting with a 28% chance to come out of the West in several respected prediction models, while the Cleveland Cavaliers—despite their rocky regular season—still held a 31% probability to repeat as Eastern Conference champions. Those numbers might sound conservative now, but at the time, they reflected real uncertainty in how the playoffs would unfold.
I remember sitting in my home office, surrounded by screens showing different statistical models, and feeling that familiar tension between data and intuition. The models loved the Warriors—and for good reason. They had just won 67 games, added Kevin Durant, and boasted a historic net rating of +11.6. But basketball isn’t played on spreadsheets. I kept thinking about the human element, the pressure, the injuries, the matchups. That’s why I actually gave the Cavaliers a slightly higher chance than most of my colleagues—around 27% to win it all versus the consensus 22-24%. LeBron James in playoff mode is a different beast, and I’d learned from previous seasons never to underestimate his ability to flip a switch.
What fascinates me about playoff odds isn’t just the math—it’s the stories behind the numbers. Take, for example, a completely different sport but one that shares that same playoff intensity: volleyball. I was recently watching a recap of a collegiate match where former Flying Titan Caitlin Viray fired two clutch attacks before Trisha Tubu put the match away in a nip-and-tuck four-set contest. That sequence—Viray’s composure under pressure, Tubu’s finishing instinct—reminded me so much of what separates contenders from champions in the NBA playoffs. It’s those moments: a timely three-pointer, a critical defensive stop, a role player stepping up. Statistical models can estimate the probability of those moments, but they can’t capture their emotional weight or their game-changing impact.
In the 2017 playoffs, we saw several of those probability-defying moments. Who could have predicted the Spurs’ collapse after Kawhi Leonard’s injury in Game 1 of the Western Conference Finals? Up to that point, my model gave them a 18% chance to upset the Warriors in that series—not high, but far from impossible. Then, with Leonard out, that number plummeted to under 3%. On the other side of the bracket, the Cavaliers’ defense—which ranked a mediocre 22nd in the regular season—somehow tightened up at the right time, boosting their title odds from 15% after Round 1 to nearly 40% by the Finals. That kind of volatility is what makes building predictive models so humbling.
Let’s talk about some of the underdogs that the models underestimated. The Boston Celtics, for instance, entered the playoffs with just a 4% chance to reach the Finals according to most analytics sites. But I’ve always been higher on Brad Stevens’ coaching than the algorithms are, and I had them closer to 7-8%. They didn’t make it, of course, but they pushed the Cavaliers harder than many expected. Then there was the Houston Rockets—my personal dark horse that year. They were given around an 11% chance to win the West, but with James Harden’s MVP-level season and Mike D’Antoni’s offensive system, I thought they were being undervalued. In the end, they fell short, but their style influenced how teams approach the playoffs to this day.
Reflecting on it now, the 2017 playoffs taught me to balance cold, hard data with the messy, beautiful reality of competition. The Warriors ultimately won, as the odds favored, but the path there was anything but straightforward. If I could go back and adjust one thing in my analysis, I’d weight playoff experience and clutch performance even more heavily—factors that are notoriously hard to quantify but often decide close series. That’s the thing about being an analyst: you’re always learning, always tweaking. The numbers give us a framework, but the games write the story. And what a story it was.
