Predicting playoff success?

[True Warrior Zo]

As the playoffs begin, I was wondering if it would be possible to predict winners (or justify them) based on their schedule difficulty. We all know that for many different reasons, the NBA schedule is not fair/equal to all teams.

Can anyone here show a statistical analysis of either previous "underdogs" winning a series, or at least monitor this year's to see how it plays out? I think this would be particularly interesting in some of the 4-5 seed match-ups.

Now obviously injuries, fatigue, and match-ups all play large factors in the equation as well, but I am wondering if there is a correlation to this undertone.

I assume posts like this are not well respected on this board. I assure you I am not trying to get rich quick off others hard work, I'm just curious to see if my hunch is right. Feel free to flame away or let me know what you think.

[Silver Bullet]

Sorry dude, I don't understand what you are saying. Are you saying predicting playoff success based on regular season schedule ?

[True Warrior Zo]

Yeah. When teams have easier schedules, their overall win total is inflated compared to the rest of the league. If there was a way to equalize the bias in the schedule, would it help to predict playoff success. Maybe something like win percentage divided schedule difficulty (combined winning percentage of opponents).

As I said above, I am not mathematically inclined, so I am not sure that is is even possible to do, but it sounds interesting. For example, if Dallas and Phoenix played each other in the second round. Dallas would be the favourite in the series. However, what if Dallas played one more game against the Nets and Wolves (compared to Phoenix), then maybe Phoenix should be the favourite in the series. Even though they ahve less wins, maybe their wins came against a higher quality of opponents; in essence making them the better team.

Does that help?

[ElGee]

Yeah. When teams have easier schedules, their overall win total is inflated compared to the rest of the league. If there was a way to equalize the bias in the schedule, would it help to predict playoff success. Maybe something like win percentage divided schedule difficulty (combined winning percentage of opponents).

As I said above, I am not mathematically inclined, so I am not sure that is is even possible to do, but it sounds interesting. For example, if Dallas and Phoenix played each other in the second round. Dallas would be the favourite in the series. However, what if Dallas played one more game against the Nets and Wolves (compared to Phoenix), then maybe Phoenix should be the favourite in the series. Even though they ahve less wins, maybe their wins came against a higher quality of opponents; in essence making them the better team.

Does that help?

It sounds like you are talking about ratings systems used frequently by the computer power rankings group (eg Jeff Sagarin). I think that's what you're asking about - it's quite common. And yes, there is a strong correlation between regular season performance and playoff success.

Not sure if that's what you're looking for...

[True Warrior Zo]

It sounds like you are talking about ratings systems used frequently by the computer power rankings group (eg Jeff Sagarin). I think that's what you're asking about - it's quite common. And yes, there is a strong correlation between regular season performance and playoff success.

Not sure if that's what you're looking for...

Yes, that is largely what I had in mind. I've never seen that before. So according to his rating system Dallas shouldn't make it out of the first round, and will definitely lose to Phoenix. I'm interested to see if this holds true.

Has he been fairly accurate?

[sp6r=underrated]

True

I've never evaluated his system so I can't rate out successful it is at making predictions. That said I would ignore Dallas. There point differential picked up to about + 5 (I believe) after the trade. The computer system doesn't realize that Dallas made a trade, so they are undersold.

[Silver Bullet]

Yeah. When teams have easier schedules, their overall win total is inflated compared to the rest of the league. If there was a way to equalize the bias in the schedule, would it help to predict playoff success. Maybe something like win percentage divided schedule difficulty (combined winning percentage of opponents).

As I said above, I am not mathematically inclined, so I am not sure that is is even possible to do, but it sounds interesting. For example, if Dallas and Phoenix played each other in the second round. Dallas would be the favourite in the series. However, what if Dallas played one more game against the Nets and Wolves (compared to Phoenix), then maybe Phoenix should be the favourite in the series. Even though they ahve less wins, maybe their wins came against a higher quality of opponents; in essence making them the better team.

Does that help?

It sounds like you are talking about ratings systems used frequently by the computer power rankings group (eg Jeff Sagarin). I think that's what you're asking about - it's quite common. And yes, there is a strong correlation between regular season performance and playoff success.

Not sure if that's what you're looking for...

Sorry I'm not familiar with Jeff Sagarin, but wouldn't strength of schedule stuff only work during the season. At the end of the season, hasn't everybody played the exact same teams the exact same number of times (within conferences).

[ElGee]

Yeah. When teams have easier schedules, their overall win total is inflated compared to the rest of the league. If there was a way to equalize the bias in the schedule, would it help to predict playoff success. Maybe something like win percentage divided schedule difficulty (combined winning percentage of opponents).

As I said above, I am not mathematically inclined, so I am not sure that is is even possible to do, but it sounds interesting. For example, if Dallas and Phoenix played each other in the second round. Dallas would be the favourite in the series. However, what if Dallas played one more game against the Nets and Wolves (compared to Phoenix), then maybe Phoenix should be the favourite in the series. Even though they ahve less wins, maybe their wins came against a higher quality of opponents; in essence making them the better team.

Does that help?

It sounds like you are talking about ratings systems used frequently by the computer power rankings group (eg Jeff Sagarin). I think that's what you're asking about - it's quite common. And yes, there is a strong correlation between regular season performance and playoff success.

Not sure if that's what you're looking for...

Sorry I'm not familiar with Jeff Sagarin, but wouldn't strength of schedule stuff only work during the season. At the end of the season, hasn't everybody played the exact same teams the exact same number of times (within conferences).

No: http://www.eightpointsnineseconds.com/2 ... -schedule/

[jicama]

... At the end of the season, hasn't everybody played the exact same teams the exact same number of times (within conferences).

Almost, but not quite. The toughest schedule was had by the worst team in the better conference: Minny didn't get to play themselves, and the majority of their games were against other Western teams.
Orlando, ranked the best in the East, played lots of bad teams and didn't have to face themselves.

The difference between the easiest and hardest schedules is about 1 point per game.
http://www.usatoday.com/sports/sagarin/nba0910.htm

[True Warrior Zo]

True

I've never evaluated his system so I can't rate out successful it is at making predictions. That said I would ignore Dallas. There point differential picked up to about + 5 (I believe) after the trade. The computer system doesn't realize that Dallas made a trade, so they are undersold.

Interesting; we might be onto something here. I am interested to see how the rest of the playoffs go.