NBA! NBA! NBA!

[DoctorUnne]

C's don't have a prayer.

No, but they have a ring.

[RichardCranium]

No, but they have a ring.

True, and theirs doesn't even have an asterisk attached to it, unlike some other Boston teams.

[Gan]

At least the crime rate will go down now.

[Latrinsorm]

I didn't bother watching the Heat - Hawks game because no LeBron/Wade, and I happened to turn it on right as Ivan Johnson started stomping people flat. Takeaways:
-Just a pathetic performance from the Hawks.
-After seeing Bosh dominate the offense, it's easy to see why the Raptors never went anywhere.
-Terrel Harris at 6'5" with 14(!!!) rebounds. Big in a season where depth is crucial.
-Dexter Pittman wasn't completely outmatched... but it was close.

Things looking good for the Heat!

[Drew]

The Heat look like the deepest team in the league at this point. They still have Mike Miller coming back, it's crazy. I don't get your Bosh comment though, he looked pretty good?

[Latrinsorm]

Oh sure he looked pretty good, my point was that if your best player is only pretty good you're going to get bounced in the first round every year, at best.

Can you imagine if LeBron could start playing like a big big man, shades of Magic 1980? Every base would be covered for the Heat.

[Drew]

Oh sure he looked pretty good, my point was that if your best player is only pretty good you're going to get bounced in the first round every year, at best.

Yeah even Kobe missed the playoffs when he had bad teams.

[DoctorUnne]

Every base would be covered for the Heat.

Except not wilting in crunch time in the finals

[Latrinsorm]

Except not wilting in crunch time in the finals

He really should have sent someone with an iPhone to the Mavs' practices, huh?

Anyway, Drew's comment reminded me of a project I had been working on recently. I extended Professor Hollinger's PER concept to the team level by weighting individual PERs of everyone on the team by minutes played in any given year. If a person had a 25 PER and played 3000 minutes, he contributed 25 * 3000 / tmMP (where tmMP accounted for overtimes and that teams generally play 5 players at a time). I then applied this to every Finals team since 1980, and found a few interesting things.

The only teams to get above a 17 regular season TER were:
-1997 Bulls (17.44)
-1991 Bulls (17.40)
-1996 Bulls (17.29)
-1992 Bulls (17.24)
-1987 Lakers (17.10)
-1983 Lakers (17.08)
-1997 Jazz (17.07)
All of these teams won the Finals except the 1997 Jazz (obviously) and the 1983 Lakers, who lost to the Moses-Erving 76ers (16.50).

The only teams to get below a 15 TER (noting that a 15 PER is by definition league average) were:
-2007 Cavs (14.92)
-1999 Knicks (14.05)
-1994 Knicks (14.78)
Obviously all these teams lost.

TER predicted the outcome of the Finals 20 of 32 times so far, which is within statistical variance of a coin flip. I also kept track of supporting cast TER, which I defined simply by removing the largest contribution. This largest contribution usually coincided with who seems like the best player on the team. (LeBron on the Cavs, Kidd on the Nets, etc.) There were a few interesting ones: by this metric Garnett was the largest contributer on the 2008 Celtics but Rondo was in 2010, Billups was on the 2005 Pistons but Ben Wallace in 2004 (very strange as PER traditionally underrates defense), the only year for Kareem on the Lakers was 1980. This highlights one of the obvious flaws in the analysis: how to account for injuries, especially those either during the Finals themselves (as Kareem's) or like Jordan's in 1986.

That digression aside, supporting cast TER improved the prediction to 22 of 32 times, right on the cusp of statistical significance. Additionally, 3 of the 10 misses were Jordan years and 1 was 2000 (The Year Shaq Cared, The Year Some Jabrone Voted For AI as MVP). If there were some statistical measure of unstoppability, that could be factored in.

The only times both metrics were wrong were:
-2006 (Heat/Wade over Mavericks/Nowitzki)
-2004 (Pistons/Wallace over Lakers/O'Neal)
-1998 (Bulls/Jordan over Jazz/Malone)
-1995 (Rockets/Olajuwon over Magic/O'Neal, sweep)
-1984 (Celtics/Bird over Lakers/Johnson)
-1983 (76ers/Malone over Lakers/Johnson, sweep)
It is interesting to recall how in some of these cases the star on the losing team had a notoriously poor performance in the finals: you shouldn't have shook Mr. Olajuwon's hand, Tragic Johnson. It is also interesting to recall how some of those cases are perhaps considered poor erroneously: Nowitzki's 2006 is statistically barely distinguishable from his 2011 Finals performance, but the differences in public perception are dramatic; Shaq's performance in 1995 wasn't really all that bad: 28 points 13 rebounds and even 6 assists per game.

As we would expect, the 2007 Cavs had the worst supporting cast TER so far: 10.99. Others below 11.50:
-1994 Rockets (11.07)
-1981 Rockets (11.32)
-1994 Knicks (11.34)
-2000 Lakers (11.46)

Finally, I measured the 2005 Lakers, the only year Kobe's team missed the playoffs, and got 15.04 overall TER, 11.87 supporting cast. Were they really that bad? This was a team with Lamar Odom and Caron Butler, after all.

[Drew]

After reading all of that I somehow feel no less informed. Also I thought Magic fans mostly blamed "Brick" Anderson for 95.