Statistical Analysis of Post-Trade Team: There is Hope!

[Nivek]

I agree that it's kind of a weird result. That's probably the main reason I finally wrote about it. I guess the point of it is that there's an actual basis for the extreme reactions game to game. I like "optical illusion" to describe the effect. In a manner of speaking, each individual game is a kind of illusion. Even bad teams look awesome when they win. Even good teams look bad when they lose.

[Zonkerbl]

So I guess the lesson to take from this is that when teams play well, they win, and when they play like crap, they lose.

Deep!

[Nivek]

So I guess the lesson to take from this is that when teams play well, they win, and when they play like crap, they lose.

Deep!

That's actually NOT the lesson I took from it. I sorta already knew that one. The lesson I got from this was that even bad teams look good when they win.

[WizarDynasty]

the problem is that coaches have not coach our starting players how to think 3 seconds in advanced at every moment in the basketball game. That is all based on coaching and the gm.

[fugop]

I would like to see the same analysis done on wins and losses, but broken down by quarters of the game. Do teams play better in wins throughout the entirety of the game, or only in the 4th quarter? Is there performance consistently better from tip off? Is the margin of victory inflated by garbage time/4th quarter fouls etc?

[pancakes3]

Illuminaire wins the reading comprehension prize. It started a few years ago when I was wondering why reactions were so strong whether win or lose. And it seemed to be across the board -- didn't matter the team. So, I took a look and found that teams look GREAT when they win and look AWFUL when they lose. Shocker, right?

and of course, the curve balls thrown in when we play our butts off and STILL lose (OKC 2OT game), and other times where we look like crap and manage to win (error: game cite not found).

[montestewart]

I would like to see the same analysis done on wins and losses, but broken down by quarters of the game. Do teams play better in wins throughout the entirety of the game, or only in the 4th quarter? Is there performance consistently better from tip off? Is the margin of victory inflated by garbage time/4th quarter fouls etc?

I'm guessing it comes down to margin of victory and that can come anywhere in the game, thus averaging out across the league and, to a lesser extent, over the course of the season for a team, producing other seemingly intuitive results like teams really come out and build up big leads tend to win and teams that do well in the 4th quarter look good.

San Antonio being grouped with teams that looked the worst in losses seemed odd, but the Spurs are 6-1 in close (3 or less) games. It seems that they win the close ones, which results in a worse margin in both losses and victories, but a better overall record for the Spurs. The Heat are 2-5 in such games, and the Wizards are 6-2. By that measure, nearly half their victories could easily have been defeats, speaking to the original point, because it makes both their victories and their defeats look weaker (like the Spurs) but they have many many many more defeats than the Spurs do.

When I was younger, I'd read or hear post game interviews, and some player or coach, summarizing a victory, would say something like:

We came out strong, rebounded well the entire game, especially on the defensive end, turned up the defense in the second half, shutting down (Star Player X), and hit our fouls shots.

And I'd think, "They did all that and they only won by one point?"

Very interesting article Nivek. Thanks.

[Nivek]

I would like to see the same analysis done on wins and losses, but broken down by quarters of the game. Do teams play better in wins throughout the entirety of the game, or only in the 4th quarter? Is there performance consistently better from tip off? Is the margin of victory inflated by garbage time/4th quarter fouls etc?

Not quite the same thing, but Roland at 82games did a quarter-by-quarter analysis a couple years ago. In his study, the decisive quarters appeared to be the 1st and 3rd -- those were the ones that good teams "won" most frequently and by the biggest margin. Winning percentage in those quarters correlated best with overall winning percentage. The lesson was that good teams usually get a lead relatively early in the game and then hang on the rest of the way. It's obviously not any kind of a "rule."

In "Basketball On Paper," Dean Oliver wrote some about the "slacking off" effect -- basically that margin of victory is in many cases artificially reduced by the team that's ahead letting up when they have a comfortable margin.

I agree it'd be interesting to look at teams quarter by quarter in wins vs. losses. My guess is that it would echo Roland's study.