This season, ESPN Insider writer John Hollinger announced a fantastic new tool for NBA rankings and predictions.
This tool, the Hollinger Playoff Predictor, predicts the outcomes of the various NBA teams' seasons, their likelihood of making the playoffs, winning their division, being the conference’s No. 1 seed, making it to the Finals, being the Champs and even winning the lottery.
The Predictor simulates the remainder of each teams' season 5,000 times, based on their current spot on the Hollinger Power Rankings.
For example, the Boston Celtics, number two in the power rankings, are expected to go 63-19, have a 100 percent chance of making it to the playoffs (in all 5,000 simulations, the Celtics made it), and have a 25.1 pecent chance of being NBA Champions (they were champs in 1,255 of the simulations).
Meanwhile, my Cavs, thirteenth in the Power Rankings, have a projected record of 47-35, a 99.2 percent chance of making it to the playoffs and a 1.3 percent chance of being NBA Champs (only 65 simulations).
The only problem with the predictor stems from its greatest attribute. The simulations are completely computerized, so there are no human biases. The simulations cannot adjust for injuries.
For example, when LeBron James was out for a while and the Cavs lost several straight games, they had low predictions because the predictor bases its simulations on the team’s then-current level of play.
There is no doubt that this will be a tool used extensively in the future to evaluate a team's play.