Never Tell Me The Odds

Given the absolute craziness to which we have been treated so far in the second round of the playoffs, I thought this would be a good time to take a look at the various noteworthy projection systems and see how they're, well, projecting these series to shake out.

We have a pair of surprising 2-0 leads in the East and a pair of 1-1 series in the West, so I thought it would be interesting to compare and contrast how Basketball-Reference, ESPN's BPI, Dunks and Threes' EPM, and Neil Paine's Estimated RAPTOR see things going both in the second round and through the rest of the playoffs.

Already, there is some massively interesting stuff here. The projection systems all see things WAY differently in the East. (Note that RAPTOR doesn't have second-round projections.)

Despite having a 2-0 lead heading home against a team possibly missing three of its top seven players, the Pacers are only 50-50 to win the series at Basketball-Reference. Now, it's possible that BBall-Ref doesn't factor in those injuries, which is why it's important to look at both BPI and EPM, which have the Pacers as heavier favorites. They're basically equivalent favorites to the Knicks in BPI and heavier favorites than the Knicks according to BPM. I think that makes much more sense than BBall-Ref's projection that the Knicks should be the heavier favorites in their series. Boston seems much more likely to stage a comeback, to my eye, given the current situation in both series.

Meanwhile, the Thunder are absolutely massive favorites in every projection system. Again, this makes sense. They're the much better team and they seemingly restored order in Game 2. I do wonder how the projection systems are each accounting for Steph Curry's absence in the Wolves-Warriors series, though. Golden State appears to be drawing dead offensively without him, as we saw last night. That should lead to the Wolves being more than merely 60-40 favorites, I think.

Things get even more interesting here. Despite the Knicks being pretty heavily favored over the Celtics by BBall-Ref, Boston has a nearly equivalent chance to win the conference. And despite the Cavs and Pacers being a 50-50 proposition to win their second-round series, Cleveland has a much better chance of advancing to the Finals, per BBall-Ref.

We don't have second-round odds from RAPTOR, but boy it feels like it thinks the Celtics are still incredibly likely to beat the Knicks. They are still significant favorites to make it out of the East and to the Finals. I know Neil uses different roster ratings for the regular season and playoffs and adjusts those ratings for injuries (this is how the FiveThirtyEight version of RAPTOR did things as well, and I'm obviously very familiar with that one), so I'd imagine that RAPTOR would have the Cavs as pretty significant favorites to win their series, too, if they were actually healthy.

Again, this is why we have the other projection systems to look at. BPI and EPM seem to see the chances in a more logical way, given the chances they have of each team winning the second-round series, which obviously needs to happen if you are going to win the conference finals. I do wonder why there is such a big gap between those two systems and the other two, but I think it's fascinating that each of the four projection systems favors a different team to come out of the East. BBall-Ref has the Cavs, BPI has the Knicks, EPM has the Pacers, and RAPTOR has the Celtics. That actually seems fitting for the situation we're currently in.

Things are much cleaner in the West, where Oklahoma City is the heavy favorite and a better-than-even bet to make the Finals in three of the four projection systems. The only exception is EPM, which also gives the Thunder the lowest-percentage chance to make it out of the second round. (Albeit still a 78.4% chance, as seen above.) Still, it looks like everyone considers the Thunder incredibly likely to represent the Western Conference in the NBA Finals.

Again, we have a significant disparity between projection systems here.

Which makes sense because there is also a disparity in the previous rounds. But man, OKC is just a massive favorite at both BBall-Ref and BPI. They're implying you should take the Thunder against the field of the other seven teams. Honestly, I think I'm with them.

RAPTOR is pretty close to saying the same, but again has the Celtics with by far the second-best chance to win. That fits with its apparent stance that Boston is still incredibly likely to beat the Knicks, and then be a heavy favorite over whichever team wins the Cavs-Pacers series. Everyone else is lagging so far behind in RAPTOR that they're barely worth mentioning.

EPM is interesting because it gives every single remaining team at least a 5% chance of winning it all, and has Oklahoma City as a heavy but not overwhelming favorite. It apparently feels pretty decent about the Wolves' chances of beating the Thunder, should that be our Western Conference Finals matchup. And then EPM is basically just throwing its hands in the air with the Eastern Conference, which, again, seems somewhat fitting.

All in all, it seems like the takeaway here is: There's the Thunder, and then there's everyone else. That's sort of how the regular season went, at least in the West. Given that the two East teams that won 60-plus games are each down 2-0 in their series, I suppose it makes sense that it looks that way again right now.