ESPN Analytics still believe in Celtics

While the 63 percent chance of winning does favor the Heat, it still seems absurdly supportive of the Celtics. Still, it’s better than the odds ESPN Analytics gave the Heat before the series: 3 percent.

Even Miami winning the first two games of the series on the road didn’t impress ESPN’s computers, who still had Boston as significant series favorites going into Game 2.

ESPN Analytics had Boston as more likely to win the series down 2-0 (65 percent) than the Heat up 3-0 (63 percent). And that’s after a 26-point blowout win in Game 3.

ESPN Analytics did up that number to 69 percent for the Heat, meaning they still view the Celtics as having a 31 percent chance to advance.

Computer models do tend to love this Celtics team. And ESPN’s model has been infatuated with them since last year when their Basketball Power Index gave Boston an 86 percent chance of beating the Warriors in the Finals.

Perhaps there’s a die-hard Celtics fan running the computers in Bristol. Maybe the algorithm assumed that Jayson Tatum would make at least one shot in a fourth quarter of this series. Maybe a programmer spilled clam chowder on the keyboard before running the predictions.

But all Boston can do is cling to its 31 percent prediction, and use it to gain confidence for Game 4 in Miami. We think ESPN’s odds are about 37 percentage points too high.