Playoff Predictions and Sample Size

[jinxed]

We are now about 10% done with the season and the NBA standings are looking much different than what any of us expected preseason. Lakers in the playoffs? Clips and Hornets in first place? Celtics out of the playoff race?

Is this a case of small sample size theater, or is 10% of the season a large enough sample to get a decent prediction? To test this out I looked at the standings on this date last year.

The first number before the team name is their standing on Nov 11, 2015. The second number after their team name is where they finished to end the RS.

East
1. Cavs (1)
2. Raps (2)
3. Hawks (4)
4. Heat (3)
5. Hornets (6)
6. Celtics (5)
7. Pacers (7)
8. Detroit (8)

West

1. GSW (1)
2. Spurs (2)
3. OKC (3)
4. Clips (4)
5. Blazers (5)
6. Mavs (6)
7. Grizzlies (7)
8. Rockets (8)

As you can see, by Nov 11th of last year, the standings were already almost a flawless match of where things would end up by the end of the season. The only deviation is that 4 way tie we had last year between the 3-6 seeds in the east. The WC played out perfectly.

It seems that 10% of the season is a large enough sample to get a good feel for how things will end up at the end of April. Given this, are you a believer that we we have today is going to be representative of what things will look like at the end of this year? If not, what changes do you see? Which team will fall out of the race? Which team will move up on the standings.

[Raptors1991]

Way too early my friend;

- Some teams have harder/eaasier schedules to start the season
- Team chemistry
- Injuries
- Trades
etc.

[jinxed]

Way too early my friend;

- Some teams have harder/eaasier schedules to start the season
- Team chemistry
- Injuries
- Trades
etc.

I would have thought so too. But by this date last year we already knew 16 out of 16 playoff teams, and the exact seeding for 12 of those 16 teams

[Edrees]

That's pretty crazy that it was a match so early last year. I would bet that it's not nearly a match most seasons.

[baldur]

wow. last year must have been an exception i suppose. it was almost exact standing as the end of the regular season.

[JulesWinnfield]

I give you an and1 for the research and it is kind of eery, but I believe that would be somewhat of an anomaly if you studied more years. There have clearly been a bunch of examples of teams who get off slow and hit their stride, and others who get off to starts that are illusions. This is particularly true of teams who make up the 5-8 seeds

[Tinseltown]

Yep my Lakers are playoff bound.

[Chuck Everett]

You should have also done the 2013-14 season and the 2014-15 season as well to see if a trend developed. Hard to call this effective statistical analysis using only one season as your measurement.

[jackwindham]

10 percent seems a small sample size because so many things can happen from now until next spring.

[Berserk_Raptor]

i agree with some of you saying its still too early to judge but at least we can point out that Lakers have done well thus far, the future is bright for them, they look fresh, young and team up pretty good.

[bwgood77]

We are now about 10% done with the season and the NBA standings are looking much different than what any of us expected preseason. Lakers in the playoffs? Clips and Hornets in first place? Celtics out of the playoff race?

Is this a case of small sample size theater, or is 10% of the season a large enough sample to get a decent prediction? To test this out I looked at the standings on this date last year.

The first number before the team name is their standing on Nov 11, 2015. The second number after their team name is where they finished to end the RS.

East
1. Cavs (1)
2. Raps (2)
3. Hawks (4)
4. Heat (3)
5. Hornets (6)
6. Celtics (5)
7. Pacers (7)
8. Detroit (8)

West

1. GSW (1)
2. Spurs (2)
3. OKC (3)
4. Clips (4)
5. Blazers (5)
6. Mavs (6)
7. Grizzlies (7)
8. Rockets (8)

As you can see, by Nov 11th of last year, the standings were already almost a flawless match of where things would end up by the end of the season. The only deviation is that 4 way tie we had last year between the 3-6 seeds in the east. The WC played out perfectly.

It seems that 10% of the season is a large enough sample to get a good feel for how things will end up at the end of April. Given this, are you a believer that we we have today is going to be representative of what things will look like at the end of this year? If not, what changes do you see? Which team will fall out of the race? Which team will move up on the standings.

That's honestly unbelievable if accurate. Good work.

[Ignitowsky]

We are now about 10% done with the season and the NBA standings are looking much different than what any of us expected preseason. Lakers in the playoffs? Clips and Hornets in first place? Celtics out of the playoff race?

Is this a case of small sample size theater, or is 10% of the season a large enough sample to get a decent prediction? To test this out I looked at the standings on this date last year.

The first number before the team name is their standing on Nov 11, 2015. The second number after their team name is where they finished to end the RS.

East
1. Cavs (1)
2. Raps (2)
3. Hawks (4)
4. Heat (3)
5. Hornets (6)
6. Celtics (5)
7. Pacers (7)
8. Detroit (8)

West

1. GSW (1)
2. Spurs (2)
3. OKC (3)
4. Clips (4)
5. Blazers (5)
6. Mavs (6)
7. Grizzlies (7)
8. Rockets (8)

As you can see, by Nov 11th of last year, the standings were already almost a flawless match of where things would end up by the end of the season. The only deviation is that 4 way tie we had last year between the 3-6 seeds in the east. The WC played out perfectly.

It seems that 10% of the season is a large enough sample to get a good feel for how things will end up at the end of April. Given this, are you a believer that we we have today is going to be representative of what things will look like at the end of this year? If not, what changes do you see? Which team will fall out of the race? Which team will move up on the standings.

Your sample size of looking at just last season is too small.

[puppa bear]

We are now about 10% done with the season and the NBA standings are looking much different than what any of us expected preseason. Lakers in the playoffs? Clips and Hornets in first place? Celtics out of the playoff race?

Is this a case of small sample size theater, or is 10% of the season a large enough sample to get a decent prediction? To test this out I looked at the standings on this date last year.

The first number before the team name is their standing on Nov 11, 2015. The second number after their team name is where they finished to end the RS.

East
1. Cavs (1)
2. Raps (2)
3. Hawks (4)
4. Heat (3)
5. Hornets (6)
6. Celtics (5)
7. Pacers (7)
8. Detroit (8)

West

1. GSW (1)
2. Spurs (2)
3. OKC (3)
4. Clips (4)
5. Blazers (5)
6. Mavs (6)
7. Grizzlies (7)
8. Rockets (8)

As you can see, by Nov 11th of last year, the standings were already almost a flawless match of where things would end up by the end of the season. The only deviation is that 4 way tie we had last year between the 3-6 seeds in the east. The WC played out perfectly.

It seems that 10% of the season is a large enough sample to get a good feel for how things will end up at the end of April. Given this, are you a believer that we we have today is going to be representative of what things will look like at the end of this year? If not, what changes do you see? Which team will fall out of the race? Which team will move up on the standings.

Your sample size of looking at just last season is too small.

Exactly. The work is good, but extend that through the next 9 season and get a sample set of 10 seasons (at least). Then let's see how accurate the standings at Nov. 11 actually is ...

[puppa bear]

Alrighty, got bored and did the work for you:

Over the past 10 seasons where there was a standings on Nov 11 (excluding the last lock-out), here are the stats:

East:
Average of 6.3 out of 8 in top-8 make the PO
Minimum 5, maximum 8, mode 7
Most consistently in PO were 1, 3, 5, 6 (90% each)
Least consistently in PO was 7 (50%)

West:
Average of 5.9 out of 8 in top-8 make the PO
Minimum 4, maximum 8, mode 6
Most consistently in PO were 1, 5 (100% each) & 2 (90%)
Least consistently in PO was 4 (40%)

So, most likely 6 out of the 8 in the PO for West make it & either 6 or 7 from the East.

Clippers and Rockets will definitely make it.

90%ers are: Warriors, Cavs, Hawks, Bulls & Bucks.

This doesn't account for removing teams that didn't make the PO because they had injuries to key players during the rest of the season (eg MIA in 2014-15), nor did it take into account for teams that started the season without key players. So, take this as you will - grain of salt suggested.

[jinxed]

Alrighty, got bored and did the work for you:

Over the past 10 seasons where there was a standings on Nov 11 (excluding the last lock-out), here are the stats:

East:
Average of 6.3 out of 8 in top-8 make the PO
Minimum 5, maximum 8, mode 7
Most consistently in PO were 1, 3, 5, 6 (90% each)
Least consistently in PO was 7 (50%)

West:
Average of 5.9 out of 8 in top-8 make the PO
Minimum 4, maximum 8, mode 6
Most consistently in PO were 1, 5 (100% each) & 2 (90%)
Least consistently in PO was 4 (40%)

So, most likely 6 out of the 8 in the PO for West make it & either 6 or 7 from the East.

Clippers and Rockets will definitely make it.

90%ers are: Warriors, Cavs, Hawks, Bulls & Bucks.

This doesn't account for removing teams that didn't make the PO because they had injuries to key players during the rest of the season (eg MIA in 2014-15), nor did it take into account for teams that started the season without key players. So, take this as you will - grain of salt suggested.

Great work Puppa Bear!

[og15]

This season started on October 25th, so I'm going to adjust the date as necessary.

Nov 11th, 2014

East:
1) Toronto (4)
2) Chicago (3)
3) Washington (5)
4) Miami (10)
5) Brooklyn (8)
6) Atlanta (1)
7) Cleveland (2)
8) Milwaukee (6)
9) Boston (7)

West:
1) Memphis (T5 - 5th)
2) Houston (T2 - 2nd)
3) Golden State (1)
4) Portland (4)
5) Dallas (7)
6) Sacramento (13)
7) LA Clippers (T2 - 3rd)
8) New Orleans (8)
9) San Antonio (T5 - 6th)


Nov 12th 2013

East:
1) Indiana (1)
2) Miami (2)
3) Atlanta (8)
4) Chicago (4)
5) Philadelphia (14)
6) Boston (12)
7) Charlotte (7)
8) Toronto (3)
11) Brooklyn (6)
14) Washington (5)


West:
1) San Antonio (1)
2) Oklahoma (2)
3) Portland (T4)
4) Phoenix (9)
5) LA Clippers (3)
6) Dallas (8)
7) Houston (T4)
8) Golden State (6)
10) Memphis (7)

Nov 15th 2012, started season Oct 30th

East:
1) New York (2)
2) Miami (1)
3) Milwaukee (8)
4) Brooklyn (4)
5) Chicago (5)
6) Boston (7)
7) Philadelphia (9)
8) Charlotte (14)
9) Atlanta (6)
10) Indiana (3)

West:
1) Memphis (5)
2) San Antonio (2)
3) LA Clippers (4)
4) Oklahoma City (1)
5) Minnesota (12)
6) Dallas (10)
7) Houston (8)
9) Golden State (6)
11) Denver (3)
14) LA Lakers (7)

[Stoked]

Jazz are already fighting the injury bug and has a tough opening schedule and are 7-4. Their road will get easier.

[Pennebaker]

The only thing I'm certain of is that some team is going to be playing the Cavaliers in the Finals.

[Warriorfan]

Maybe only 4 teams went through major roster overhaul.

OKC, GS, NY, Mia, Chi

I don't see Atl Horford for Howard as a big changer.

SA Duncan for Gasol.

[Prokorov]

wow. last year must have been an exception i suppose. it was almost exact standing as the end of the regular season.

last year you had clear tanking teams and clear awful teams. so playoff teams were more clear.

this year... its muddled and alot of those bad tanking teams are in the stage were they are trying to win games.