Even the most ardent, verdant Boston Celtics fans remained skeptical when the C's won 10 of 15 in January after a miserable start.
But the first hint that maybe we were watching something sustainable came from an outlet best known for its political prognostications.
While most of us remained in wait-and-see mode, the models at FiveThirtyEight were positively bullish. On Feb. 5, they gave the Celtics the fifth-highest odds to win a championship, at seven percent. A month later, they led the entire league, at 17 percent. By the start of the playoffs, that number had climbed to 34 percent, nearly double the runner-up Suns.
After dismantling the Warriors with a stunning fourth quarter rally in Thursday's Game 1, the Celtics have seen their odds soar to 92 percent. So, we put the question to Nate Silver, FiveThirtyEight's founder: what did their models see before anyone else?
In an e-mail, Silver laid out the three primary factors, in order of importance, and No. 1 won't remotely surprise you. It's Boston's defense.
"One of the best in the NBA in a long time," Silver wrote. "Defense is still probably underrated relative to offense. And I'd mention that while sometimes defensive stats can be fluky, e.g. if your opponent is missing good looks, I don't think that's the case here.
"In other words, my subjective impression matches the model on the Celtics' D. They have a lot of plus defenders up and down the rotations, they're well-coached, and they rarely concede easy shots, especially in the half-court."
Next up is their net rating. At its simplest level, think of net rating as scoring differential, except as measured over 100 possessions to remove pace from the equation. The Celtics didn't just win a bunch of games in the second half, they blew teams out, and the models reward teams that win a lot by a lot.
"From January 23 through the end of the regular season, they outscored opponents by 15 points per game," Silver wrote. "You just don't do that unless you're very good; the best point differential sustained over a full season is 12.3 points (1971-72). And, although the closeout against Miami was rocky, they faced a much tougher road to the Finals than most teams."
That leaves factor No. 3, and it shouldn't surprise anyone, either: Jayson Tatum. A First-Team All-NBA choice for the first time this season, Tatum has elevated his game to the elite level, and that's a virtual necessity to win it all.
"Our model identified him as a Top 10 player pretty early on," Silver wrote. "That's especially important in the playoffs when rotations get shorter and star talent matters more. So the Celtics are not a team like the 2004 Pistons that is good across the board but with no standout player. They have a great rotation but also an All-NBA star."
Add it all together, and the Celtics find themselves in a position few envisioned four months ago: just three wins from a title.
FiveThirtyEight was the exception, but even now, Silver preaches caution. His outlet just ran a story headlined, "We might be overrating the Celtics, but you're probably underrating them," and he recognizes the limits of analytics in predicting real-world outcomes. That helps explain how the Warriors opened the Finals as slight favorites in Las Vegas.
"I'd keep in mind that all models have flaws, and usually it's better to trust the consensus (e.g. betting markets) than any one model or opinion," he said. "So, while it's been exciting to see the Celtics do so well in the postseason, they still have their work cut out to beat the Warriors."