What Are The Odds: A Statistical Analysis of Tanking in The NBABrayden GerrardBlockedUnblockFollowFollowingMar 11What Is Tanking?Tanking is the art of creating a purposefully bad team with the intention of losing games to gain high draft picks.

Ultimately, almost all people agree you need a truly great player on your team to win a championship in today’s NBA.

Tanking aims to gain this player with top draft pick, and to ultimately win a championship with the core constructed while tanking.

The strategy has become increasingly common in recent years as winning an NBA championship has become increasingly difficult.

The Aim of This AnalysisSubsequently, this analysis aims to determine three different things:What is a reasonable scenario that tanking might look like?What kind of draft picks would that team receive, and what are the odds that one of those draft picks might yield a truly great player?If that team does gain a superstar, how likely are they to win a Championship with that player?For the purpose of this analysis, I will define a great player as a Hall of Fame caliber player.

Part I: The Tanking ScenarioDetermining a reasonable tanking scenario is the most subjective part of this analysis.

Tanking can take many different shapes.

How poorly a team places depends on the tanking “competition”, how well they draft, and how committed to tanking they are.

To get a start, I made a decision that a tanking team has made an obvious decision to rebuild and temporarily forgo competitiveness (i.

e.

, they aren’t just bad).

Since no one scenario could encompass everything, I instead decided to create three, modeled after three recent rebuilds the NBA has seen.

Here they are:Philadelphia 76ers: The Hard Rebuild ScenarioThe 76ers rebuild began in the 2013–2014 NBA season, after trading away marquee player Jrue Holiday.

The 76ers are noted for their extreme commitment to tanking, stripping away any and all assets that could possibly assist in winning.

At their lowest point, they were the second worst team in NBA history, thus making this the Hard rebuild.

Four seasons after it began, the 76ers made the playoffs for the first time since the rebuild started thus marking the end of the rebuild as far as this analysis is concerned.

The 76ers seeding during the rebuild was as follows:Win Percentage: 23%Minnesota Timberwolves: The Moderate Rebuild ScenarioThe Timberwolves rebuild began in the 2014–2015 season after trading away franchise player Kevin Love.

However, they did retain a first overall selection in Andrew Wiggins from the trade, which helped to make the rebound more speedy.

The tank would last for three seasons before they would make the playoffs once again.

During this time period, their seeding was as follows:Win Percentage: 31%Utah Jazz: The Soft Rebuild ScenarioThe Utah Jazz started of their rebuild when they lost both Al Jefferson and Paul Millsap.

However, a 4th year Gordon Hayward helped to make the Jazz more respectable during their rebuild than the other two scenarios.

The 2013–2014 season was also their first season rebuilding, and it lasted three seasons before they made the playoffs again.

Win Percentage: 42%Part II: What Are The Odds Of Drafting A Hall of Fame Player?To determine this, we have two components to consider:What kind of draft picks is the draft lottery likely to produce at each seeding position?What is the likelihood of drafting a HoF player at each draft position?Here, we will tackle the first question since it’s quite easy.

Starting in the 2019 NBA Draft, the new lottery odds are as follows:Now that we have the odds of receiving each draft pick from each seed, we need the odds of drafting a Hall of Famer from each draft position.

To do this, I use Basketball Reference’s Draft Finder in order to analyze some data.

I analyze draft selections at each position from the year 1950 to 1995 (going after 1995 included some active or very recently retired players that haven’t had a chance to make the HoF yet).

The odds of drafting a Hall of Famer at each position were as follows:There is some unusual variance, according to this you would rather draft at 8th overall than 4th overall.

This is an unfortunate consequence of small sample sizes.

To combine these two components, I weighted the likelihood of receiving each draft pick with the likelihood of drafting a Hall of Fame player at that position for each seed ( I only completed the seeds necessary for the three scenarios due to time constraints).

The results of this are below:Now, using this information we can easily cross-reference it with each scenario’s seeding results in order to determine the likelihood that rebuild yields a Hall of Fame caliber player.

Those results are stated below:So we have now successfully answered Part II of this analysis.

It is important to note something here; a longer tank period may not indicate a higher likelihood of drafting Hall of Fame talent.

It may instead indicate consistently poor drafting on that team’s behalf, which is why they failed to improve.

However, controlling for that would prove very difficult.

Part III: How Likely Is It To Win An NBA Championship With That Player?This part is considerably more tricky.

It is not a precise science (as I will expand on in the conclusion), and is very time intensive.

In order to attempt to measure this, I went back to Basketball Reference’s Draft Finder and looked at every Hall of Famer drafted since 1950.

There was 122 entries.

I eliminated 18 entries for several reasons; some were doubles (players were occasionally drafted twice in the early history of the league, and both drafts were counted).

Others were players who were inducted into the Hall of Fame as coaches or management, and some were primarily ABA players.

This left 104 entries to work with.

Once I had it down to 104 players, I manually checked every one and determined that 60 of them were drafted to teams that achieved a losing record the year before they got there.

The other 44 were drafted to teams already at .

500 or better before their arrival.

In this example, we will consider “losing” as equivalent to rebuilding/tanking for simplicity.

Of those 60 players, 23 of them never won an NBA championship in their careers.

This leaves 37 Hall of Fame players drafted to losing teams whom won an NBA championship.

However, a further 19 of those players won their championships only with other teams later in their careers.

This leaves a final sample of 18 Hall of Famers who originated with a losing team, and won an NBA championship there during their first stint there.

So if a losing team drafts a Hall of Fame talent, there is historically a 30% chance of them winning an NBA championship there.

However, there is an important adjustment to make here.

The NBA only became a 30-team league in 2004, with the addition of the New Orleans Pelicans.

None of the 18 players mentioned above won an NBA championship in 2004 or after.

This means all of the players won championships when it was — to some degree — less challenging than it is currently.

In order to control for this, I applied an adjustment factor to account for big the league was when that player won their championship.

The adjustment factor was simply calculated by taking the number of teams in the NBA at the time, and dividing it by 30.

For example, a player that won a championship in a 15-team league would get an adjustment factor of 0.

5, for 50% of the difficulty of today.

If a player won multiple championships with that team, I used the one most difficult.

The results are here:The new adjusted total of 11.

433/60, or a 19.

1% of winning an NBA championship.

However, I wanted to take things a step further.

I want to determine if the chances of winning a championship changed based on how bad the team was.

To do this, I created three categories based on each scenario.

I took each scenario’s average winning percentage over the course of their rebuild and applied +/- 5% to it in order to get a category range of 10%.

So the categories ended up as follows:Hard, 23 Win %: 18–28%Moderate, 31 Win %: 26–36%Soft, 42 Win %: 37–47%After sorting through the data, I got the following results:According to this, a moderately bad team is more likely to succeed than a very bad team or a more mediocre team.

This might be meaningful, or it might be the result of small sample sizes.

Sadly, this is simply the best data we have to work with.

LimitationsI thought I would make a quick section dedicated to the the infinite limitations that exist in an analysis like this.

Basically, every rebuild is different and I can’t account for that.

Some rebuilds manage to get many extra draft picks (Celtics), others manage to lose theirs (Nets).

Some rebuilds manage to trade picks and prospects for existing NBA superstars, that wouldn’t be captured here.

Some rebuilds have long tank periods because they just suck at building good basketball teams (the Magic have been “rebuilding” since Dwight left).

Basically, there is too many moving part to account for it all.

I just did the best I could with what I had.

ConclusionsTo bring this together, I had to combine the two factors in order to come up with one final odds for each scenario.

I have put the final results in the table below:How meaningful is this all?.I have no idea.

There are many factors still unaccounted for.

But hopefully it was at least kind of fun or interesting to read if nothing else.

.