So, how did BPI do with its projections?

Getty Images

With the NBA Finals three days away, it’s a good time to look back at ESPN’s NBA Basketball Power Index and some of the projections based on the rating system.

Below are three teams that attracted a lot of scrutiny and discussion around NBA BPI when the Stats & Information group launched it, with a recap of how they performed relative to those projections.


The most common question around the time of BPI’s launch in mid-January was, “Why are the Atlanta Hawks so low?” This was in the midst of what wound up being a 19-game win streak for Atlanta.

ESPN’s Marc Stein had the Hawks as the top team in his power rankings, which serve as a good proxy for the at-that-moment public perception of a team, for six straight weeks, never dropping them below No. 6 after January. They were cruising to the No. 1 seed in the Eastern Conference, but BPI never rated them as a top-5 team; they hovered around the 10th spot from midseason on.

The relatively low ranking was mostly the result of the Hawks' vastly outperforming preseason expectations. Those expectations don’t fade entirely from the model, since they do help in measuring a team’s true strength going forward. A weak strength of schedule from playing in the East and a relatively unimpressive scoring margin also took some shine off the Hawks’ great win-loss record to that point in the season.

At the All-Star break, BPI projected the Hawks (43-11 at the time) to come back to earth and finish with a record of about 60-22. Atlanta went 17-11 after the break and finished with a regular-season mark of -- 60-22.

Yes, they made the Eastern Conference finals but did so by winning close six-game series against two of the worst playoff teams (based on BPI at the end of the regular season; the Nets were No. 23, and the Wizards No. 14). BPI projections had the Hawks as fairly significant favorites going into both of those series.

BPI projected the Cavaliers as 57 percent favorites entering the conference finals against the Hawks, partly because of higher BPI and partly because of greater playoff experience. Cleveland swept Atlanta, even without Kevin Love and with Kyrie Irving missing significant time.

Looking back, it sure seems as if BPI was more accurate than public perception about the Hawks.


On the flip side of that Hawks debate were the Cavaliers, who started the season slowly (19-20 in mid-January) despite lofty preseason expectations. Although the Cavs weren’t seen as a top team at that point (ESPN’s Stein had dropped them into the 15 to 17 range), they had a solidly above-average BPI. LeBron James & Co. were rated about the same as the Hawks, despite a very different record, in large part because of those preseason expectations.

Although many analysts were questioning the Cavaliers’ chances of making any sort of run, BPI still gave them a pretty good chance in a weak Eastern Conference. As the Cavaliers started putting together wins against good teams and by bigger margins late in the season, BPI adjusted to that, and Cleveland was ranked fourth by the time the postseason came around.

Before the playoffs, BPI projections saw the Cavaliers as the clear favorite in the Eastern Conference, giving them a 44 percent chance to reach the Finals, compared with 24 percent for the Hawks and 17 percent for the Bulls. Obviously Cleveland made the road look even easier than those projections indicated, losing two games en route to winning the East.


It might seem obvious in hindsight that Golden State was clearly the favorite to emerge out of the West, but it wasn’t too long ago that people were skeptical about the team’s style, lack of playoff pedigree, new coach, etc. BPI has been consistently high on the Warriors since the metric’s debut: They’ve always been a couple of points ahead of all other teams.

Despite there being many good teams in the West, BPI projected that the race for the top seed was basically over at the All-Star break, giving the Warriors a 99 percent chance of securing the No. 1 seed at that point. This felt too high to some observers (Memphis was within a few games of Golden State), but the Warriors clinched the No. 1 seed with nine games left and went on to have the best record in the conference by 11 games.

The Warriors not only won a lot, but (unlike the Hawks) they also played a tough schedule and outscored opponents by a margin that was among the top 10 in NBA history in terms of points-per-game differential. Going into the playoffs, coach Steve Kerr’s team was the clear BPI favorite in the West at 47 percent likely to make the Finals, compared with 27 percent for the Spurs.

Many analysts still liked San Antonio because of the Spurs’ greater playoff experience, but BPI saw the matchup with the Clippers and then another series without home-court advantage as stumbling blocks to the Spurs even getting to the Warriors. BPI projections gave the Spurs a less than 50 percent chance to make the conference finals, and that played itself out, as Golden State was able to reach the Finals without having to face the defending champions.

BPI has the Warriors as a 72 percent favorite in the Finals against the Cavaliers, so we’ll see if they are able to continue their run over the next couple of weeks.

Looking back, a summary of a few of BPI’s best predictions is in the introduction to this March 26 article by Ben Alamar, director of production analytics:

“It's June 4. You've survived late-night overtime games, the bursting of the Hawks' bubble and the endearing but ultimately failed run of the Grizzlies, so now you can sit back and enjoy the compelling matchup between the Cleveland Cavaliers and Golden State Warriors.

"That is the most likely NBA Finals scenario, according to ESPN's Basketball Power Index.”

NBA BPI is not always right, and we don’t suggest that it is -- that’s why there are percentages to everything. But it is good to point out times like this when the system went somewhat against the grain and winds up looking smart in the end.