An NBA Draft Lottery analysis: Is getting worse than we are really necessary?

[nickagneta]

I have seen a lot of talk on this site regarding tanking and how the Celtics have to get worse than they are to win the lottery. Opinions have varied regarding just how bad the current roster construction is and how it will translate into wins and losses. For people who desperately want to draft Andrew Wiggins, this team just isn't bad enough and they want change so that this team gets even worse.

But just how bad does this team need to be to win the NBA Lottery? How many wins and losses is it going to take? Well, let's take a look at history because, as the saying goes "Those that do not know history are doomed to repeat it."

Since 1994, when the NBA went to the new weighted system for the NBA Lottery, there have been 20 drafts. Let’s look at the results of the top three winners of the drafts.

Teams will be followed by their finish in relation to worse to best record, their win-loss record, and winning percentage. Know that there were years that teams such as Vancouver, Toronto and Charlotte were not eligible to get the first pick and that if 2 teams were tied for a spot the lottery averages their ping pong balls out and then adds one ping pong ball for the tiebreaker but I took the placements directly from basketball-reference.com and so did not figure in ties for placement in regarding to 1st to 14th placement in the lottery. This would, of course, slightly skew the numbers.

1994
1. Milwaukee - 3rd 20-62   0.244
2. Dallas - 1st 13-69   0.159
3. Detroit - 4th 20-62   0.244

1995
1. Golden State - 5th 26-56   0.317
2. LA Clippers(traded to Denver) - 1st 17-65    0.207
3. Philadelphia - 4th 24-58    0.293

1996
1. Philadelphia - 2nd 18-64   0.220
2. Toronto - 3rd - 21-61   0.228
3. Vancouver - 1st 15-67   0.183
 
1997
1. San Antonio - 3rd 20-62   0.244
2. Philadelphia - 5th 22-60   0.268
3. Boston - 2nd 15-67   0.183

1998
1. LA Clippers - 3rd 17-65    0.207
2. Vancouver - 4th 19-63   0.232
3. Denver - 1st 11-71   0.134

1999
1. Chicago - 3rd 13-37   0.260
2. Vancouver(traded to Houston)  -  1st 8-42   0.160
3. Charlotte - 14th 26-24   0.520

2000
1. New Jersey - 7th 31-51   0.378
2. Vancouver - 4th 22-60   0.268
3. LA Clippers - 1st 15-67   0.183

2001
1. Washington - 3rd 19-63   0.232
2. LA Clippers(traded to Chicago) - 8th 31-51   0.378
3. Atlanta(traded to Vancouver) - 5th 25-57   0.305

2002
1. Houston - 5th 28-54   0.341
2. Chicago - 2nd 21-61   0.288
3. Golden State 1st 21-61   0.288

2003
1. Cleveland - 2nd 17-65   0.207
2. Memphis(traded to Detroit) 6th 28-54   0.341
3. Denver - 1st 17-65   0.207

2004
1. Orlando 1st 21-61   0.288
2. LA Clippers(traded to Charlotte) -4th 28-54   0.341
3. Chicago - 2nd 23-59   0.281

2005
1. Milwaukee - 6th 30-52   0.365
2. Atlanta - 1st 13-69   0.159
3. Portland(traded to Utah) 5th 27-55   0.330

2006
1. Toronto - 5th 27-55   0.330
2. New York(traded twice to Portland) - 2nd 23-59   0.281
3. Charlotte - 3rd 26-56   0.317

2007
1. Portland - 6th 32-50   0.390
2. Seattle - 5th 31-51   0.378
3. Atlanta - 4th 30-52   0.365

2008
1. Chicago - 9th 33-49   0.402
2. Miami - 1st 15-67   0.183
3. Minnesota(traded to Memphis) - 3rd 22-60   0.268

2009
1. LA Clippers - 3rd 19-63   0.232
2. Memphis - 6th 24-58   0.293
3. Oklahoma City - 4th 23-59   0.281

2010
1. Washington - 4th 26-56   0.317
2. Philadelphia - 6th 27-55   0.330
3. New Jersey - 1st 12-70   0.146

2011
1. LA Clippers(traded to Cleveland) - 8th 32-50   0.390
2. Minnesota - 1st 17-65   0.207
3. New Jersey(traded to Utah) - 6th 24-58   0.293

2012
1. New Orleans - 3rd 21-45   0.318
2. Charlotte - 1st 7-59   0.106
3. Washington - 2nd 20-46   0.303

2013
1. Cleveland - 3rd 24-58   0.293
2. Orlando - 1st 20-62   0.244
3. Washington - 7th 29-53   0.354

So let’s crunch some numbers and see what we get. I didn’t figure averages based on wins and losses as there were 2 draft lotteries (10% of the sample) that came after lockout years and had either 50 or 66 game seasons instead of the regular 82 game season.  I based the averages on winning percentages and then translated the winning percentages into wins and losses for better translation.

The average NBA Draft Lottery winner won 24.5 games and finished with the 4th worst record in the league. The most common placement for a team winning the Lottery was 3rd with 8 wins out of 20 lotteries. Strangely enough, having the absolute worst record in the league has netted only 1 Lottery win. Here is the placement of winners:

1st = 1
2nd = 2
3rd = 8
4th = 1
5th = 3
6th = 2
7th = 1
8th = 1
9th = 1

The average NBA Draft Lottery runner up won 20.6 games and finished with the 3rd worst record in the league. The most common placement for a team finishing second in the Lottery was 1st with 8 wins out of 20 lotteries.  Here is the placement of runner ups in the Lottery:

1st =8
2nd = 2
3rd = 1
4th = 3
5th = 2
6th = 3
8th = 1

The average NBA Draft Lottery 3rd place finisher won 23.1 games and finished with the 3rd and a half place finish in the league. The most common placement for a team finishing third in the Lottery was 1st with 6 wins out of 20 lotteries.  Here is the placement of runner ups in the Lottery:

1st = 6
2nd =3
3rd = 2
4th = 4
5th = 2
6th = 1
7th = 1
14th = 1

Some lottery facts:

- Teams with the worst record did not win any Lottery placement whatsoever 5 times (25% of the time).
- Teams with the 2nd worst record did not win any Lottery placement whatsoever 13 times (65% of the time).
- Teams with the 3rd worst record did not win any Lottery placement whatsoever 9 times (45% of the time).
- Teams finishing in the bottom three in the league did not win any Lottery placement whatsoever 45% of the time.
- Teams with a winning percentage above 30% (or 25 wins or more) have placed in the Lottery 20 times or 33% of the time or an average of once per year.
- Teams with a winning percentage above 34% (or 28 or more wins) have placed in the Lottery 12 times or 20% of the time or 3 out of every 5 years.

So do the Celtics really have to get as bad as possible to get their chance at Andrew Wiggins or Jabari Parker? No, not really. It might help incrementally but in the long run it’s like buying 1000 Powerball tickets instead of 100 Powerball tickets. It helps your chances of landing the jackpot ever, ever so slightly but really has little effect. About the only thing that really helps is finishing in the bottom 6 for win loss record as compared to the rest of the league. Finishing in the bottom 6 in the league, accounts for 70% of all NBA Draft Lottery position winners (1st, 2nd, or 3rd). To give you an idea, the most wins any 6th place team had during the last 20 years was 33 wins.

So judge for yourself how useful it would be to tank but also know this, only one NBA Draft Lottery winner has won a title in the last 20 years with the player they draft at #1, that being San Antonio and Tim Duncan. Yes, San Antonio won 20% of the last 20 titles but the winner of the Lottery winning with the player selected is rare. So even winning the Draft Lottery means very little in regards to winning a title with the player you select at number one because it only happened 5% of the time and that percentage is more than likely only going to drop in the future.

[rondoallaturca]

TP, and quite honestly, this is home page material.

[indeedproceed]

Let's say there is a prize in a thing. And to get the prize, you need to bet on which color (red, blue, yellow, green) ping pong ball will be picked out of a totally random assortment of balls.

In that assortment, there are 25 red balls, 20 blue balls, 16 green balls, 14 red balls, 12 purple balls, and 10 white balls.

You know that in the last 20 draws, a red ball has only come up once.

If you knew that winning the bet would change your life forever, what color do you bet on?

[Neurotic Guy]

Let's say there is a prize in a thing. And to get the prize, you need to bet on which color (red, blue, yellow, green) ping pong ball will be picked out of a totally random assortment of balls.

In that assortment, there are 25 red balls, 20 blue balls, 16 green balls, 14 red balls, 12 purple balls, and 10 white balls.

You know that in the last 20 draws, a red ball has only come up once.

If you knew that winning the bet would change your life forever, what color do you bet on?

OK, I'll state the obvious -- Red.  But if you can't afford to lose, the correct bet is no bet at all. Even with red your chance of losing every time is about 75%.   

If it costs me less to bet on white, I'd go with white since the chance of losing, while greater (90%), isn't that much greater than red.

So if the metaphor is that a tanking C's team is risking the franchise's future by decimating the current team for the sake of more Ping-Pong balls in the lottery, it may well be a risk not worth taking.

[rondoallaturca]

Let's say there is a prize in a thing. And to get the prize, you need to bet on which color (red, blue, yellow, green) ping pong ball will be picked out of a totally random assortment of balls.

In that assortment, there are 25 red balls, 20 blue balls, 16 green balls, 14 red balls, 12 purple balls, and 10 white balls.

You know that in the last 20 draws, a red ball has only come up once.

If you knew that winning the bet would change your life forever, what color do you bet on?

Not red. To bet on the color red, I'd have to sell my car so I can have a 75% chance of losing.

[Boris Badenov]

Nice work, TP. Really sheds some light on how tough it is to implement the "draft a franchise guy" strategy over the last 20 years.

That said, your choice of time period influences the results.

If you go back to the period starting in 1980, over the next 20 years almost every championship was won by a team who drafted a top 3 guy and kept the player: Jordan, Olajuwon, Magic, Bird and Thomas. I count 19 out of 20 with the exception being the 1983 76ers.

(I'm counting Bird as top 3 because if he were not picked under that odd rule in 1978 he would have been top 2 in 1979).

[Casperian]

You've omitted the numbers for "worst possible outcome" for every lottery position in your presentation. For the worst record, it´s fourth. That´s still pretty darn good for an allegedly deep draft, and a fantastic trade chip, at worst.

[nickagneta]

Let's say there is a prize in a thing. And to get the prize, you need to bet on which color (red, blue, yellow, green) ping pong ball will be picked out of a totally random assortment of balls.

In that assortment, there are 25 red balls, 20 blue balls, 16 green balls, 14 red balls, 12 purple balls, and 10 white balls.

You know that in the last 20 draws, a red ball has only come up once.

If you knew that winning the bet would change your life forever, what color do you bet on?

Except there is no guarantee that the prize you win will be a prize to change your life forever. It could very well be a prize that makes your life absolutely miserable for a long time. Let's just call that an Oden/Olowakandi prize.

The math above is extremely solid. I even did the variances involved but didn't publish them. But if I started doing probability calculations involved with winning when the player was a sure fire franchise changer that would remain with the team long enough to win a title, the percentages and numbers would fall off the chart and prove your analogy useless.

[aporel#18]

First of all, TP for Nick, this is a fantastic post and I concur with rondoallaturca, this is front page material.

TP to Neurotic Guy also, because he just nailed it with his post. The Celtics franchise can't afford becoming the next Bobcats, and we can't giftwrap our current roster for a slightly better chance (as Nick has showed) to get a player who could become a super star.

And of course you choose the green balls. Always.

[nickagneta]

Nice work, TP. Really sheds some light on how tough it is to implement the "draft a franchise guy" strategy over the last 20 years.

That said, your choice of time period influences the results.

If you go back to the period starting in 1980, over the next 20 years almost every championship was won by a team who drafted a top 3 guy and kept the player: Jordan, Olajuwon, Magic, Bird and Thomas. I count 19 out of 20 with the exception being the 1983 76ers.

(I'm counting Bird as top 3 because if he were not picked under that odd rule in 1978 he would have been top 2 in 1979).

The time period coincides with the new weighted ping pong lottery system used today. It really is the only relevant time period for these percentages.

For instance, it wouldn't be relevant to compare true shooting percentages of players pre and post three point line as they are two completely different systems. Pre-ping pong lottery first pick analysis just wouldn't be relevant in comparing the chances of landing a top 3 pick in today's ping pong lottery system as the chances of getting the first pick were entirely different or predicated upon solely having a worse record.

[Mazingerz]

But how do you explain Cleveland's fantastic odds of landing several number 1 picks in the past decade?

[nickagneta]

You've omitted the numbers for "worst possible outcome" for every lottery position in your presentation. For the worst record, it´s fourth. That´s still pretty darn good for an allegedly deep draft, and a fantastic trade chip, at worst.

Absolutely true. Its also, mathematically, the most probable outcome. The most probable outcomes for the top picks tend to be lower than the actual finishing spot.

Worst record, 4th pick
2nd worst record, 4th pick
3rd worst record, 5th pick
4th worst record, 5th pick
5th worst record 6th pick.

[CelticConcourse]

You've omitted the numbers for "worst possible outcome" for every lottery position in your presentation. For the worst record, it´s fourth. That´s still pretty darn good for an allegedly deep draft, and a fantastic trade chip, at worst.

Absolutely true. Its also, mathematically, the most probable outcome. The most probable outcomes for the top picks tend to be lower than the actual finishing spot.

Worst record, 4th pick
2nd worst record, 4th pick
3rd worst record, 5th pick
4th worst record, 5th pick
5th worst record 6th pick.

Somebody's gotta get a top 3 pick. Why can't it be us. 

[nickagneta]

But how do you explain Cleveland's fantastic odds of landing several number 1 picks in the past decade?

Fantastic luck.

How do you explain them still sucking?

[Mazingerz]

But how do you explain Cleveland's fantastic odds of landing several number 1 picks in the past decade?

Fantastic luck.

How do you explain them still sucking?

True true. But Landing three number one picks (Lebron, Irving and now Bennett) smells something fishy in my books;

Are there other teams who have had this kind of luck?

And what was Cleveland's record prior to these picks - 2nd, 8th (but this was LAC's pick), and 3rd;

IMHO better not to tank as this may setback the team for many years (think bobcats, cleveland, wizards, pistons)