Race to the Bottom thread

[twslam07]

I know everyone has an opinion about how to fix tanking, but I have an idea (more of a theory) on how to fix the issue.

I was too lazy to type up the entire explanation so I dropped in the key bullet points in AI and asked it to build that summary for me. This is how it summarized my idea…

NBA "Games-Back Bonus" Draft Points System
(A creative idea to fix tanking and make draft order more exciting and merit-based)

The Big Motive
Right now, bad teams sometimes lose on purpose late in the season to get better draft picks (the famous "tanking" problem). This system flips that upside down:

It rewards teams for winning when they're far behind in the standings—instead of punishing wins or rewarding losses.

The goal? Use these points to decide the draft order for the next year's NBA Draft. The team with the most points gets the #1 overall pick, then #2, and so on—but only among teams that miss the playoffs. Any team that actually makes the playoffs (top 8 in their conference) gets their normal late picks (15th–30th range).

It gives struggling teams a real reason to compete and pull off upsets, while still letting true contenders focus on winning now.

How the points work, super simple
Every time a team wins a game, they earn points equal to how many games back (GB) they are from the 8th seed (last automatic playoff spot) in their own conference on that day.

- Farther back = bigger reward for a win (because it's harder to win when you're way out of it).
- Close to or above the 8th seed = small points (or even 0 if they're ahead).
- Losses = 0 points added that day.

Half-games count (e.g., 7.5 GB = 7.5 points). Points are only from wins, and the value changes daily as standings shift.

Quick example (Sacramento Kings style)
- Jan 1: Kings are 8.5 games back from West's 8th seed. They win → +8.5 points.
- They lose the rest of January → 0 points those days.
- Feb 1: Now 14 games back after the skid. They win → +14 points.
- Total so far: 22.5 points.

If they keep racking up big-GB wins while staying out of the playoffs, they could climb high in draft points—even leapfrogging other bad teams that just lose quietly.

Why this could be awesome for the draft
- Bottom teams get rewarded for fighting and winning tough games (no more pure tanking incentive).
- A scrappy team that pulls off upsets when hopeless could jump to #1 pick.
- Playoff teams stay untouched—they earn their high seeds by succeeding, so their picks stay late.
- It adds drama: Late-season games between lottery-bound teams become must-watch because every win could mean a huge draft boost.

At season's end, rank the non-playoff teams by total points earned → that's your draft order for picks 1 through (whatever number of lottery teams there are).



The reason I call this a “theory” more than an “idea” is because I haven’t been able to validate this point scoring system against historical seasons to see how it plays out, if any weighting would make it work better, etc. I asked AI to give me the results on the 2024-25 season and it basically said it’s too hard to gather the data points to give me an answer.

Very curious to see how others think this “theory” would work? And if anyone has the skills to put this theory to the test with some real data (I certainly don’t).

 

[sactowndog]

Yes, but I did make clear that this was the outline of a proposal that did not include full details as there are obviously plenty of kinks to work out (as there would be in any radical departure). None of them are too difficult to work out, I don't think.

I don't understand the question about good teams giving up an economic advantage. Once all of the results are tallied, *somebody* has to get the number one pick, *somebody* has to get the #2 pick, etc. Nobody wants to see the #1 pick go to a team that is already competitive, because THAT hurts their own chances to win.

Yes, it would not make sense to allow teams to rank themselves. But that's a minor detail.

Again, don't nitpick on details here. Details can be ironed out. The point is that that we move away from *any* objective metric that takes record into account and change to a subjective and collective ranking, where the ranking is made by the people (NBA front offices) who are the world's biggest experts in how good/bad all of the NBA's teams are, and who each have game-theoretical incentive to rank accurately, because any good team that is misranked higher than they deserve will get a competitive advantage, and no NBA front office is going to want to give any other team a competitive advantage.

I think the saying the devil is in the details comes to mind here. I think you are completely on the right track, but the details will matter.

 

[sactowndog]

I know everyone has an opinion about how to fix tanking, but I have an idea (more of a theory) on how to fix the issue.

I was too lazy to type up the entire explanation so I dropped in the key bullet points in AI and asked it to build that summary for me. This is how it summarized my idea…

NBA "Games-Back Bonus" Draft Points System
(A creative idea to fix tanking and make draft order more exciting and merit-based)

The Big Motive
Right now, bad teams sometimes lose on purpose late in the season to get better draft picks (the famous "tanking" problem). This system flips that upside down:

It rewards teams for winning when they're far behind in the standings—instead of punishing wins or rewarding losses.

The goal? Use these points to decide the draft order for the next year's NBA Draft. The team with the most points gets the #1 overall pick, then #2, and so on—but only among teams that miss the playoffs. Any team that actually makes the playoffs (top 8 in their conference) gets their normal late picks (15th–30th range).

It gives struggling teams a real reason to compete and pull off upsets, while still letting true contenders focus on winning now.

How the points work, super simple
Every time a team wins a game, they earn points equal to how many games back (GB) they are from the 8th seed (last automatic playoff spot) in their own conference on that day.

- Farther back = bigger reward for a win (because it's harder to win when you're way out of it).
- Close to or above the 8th seed = small points (or even 0 if they're ahead).
- Losses = 0 points added that day.

Half-games count (e.g., 7.5 GB = 7.5 points). Points are only from wins, and the value changes daily as standings shift.

Quick example (Sacramento Kings style)
- Jan 1: Kings are 8.5 games back from West's 8th seed. They win → +8.5 points.
- They lose the rest of January → 0 points those days.
- Feb 1: Now 14 games back after the skid. They win → +14 points.
- Total so far: 22.5 points.

If they keep racking up big-GB wins while staying out of the playoffs, they could climb high in draft points—even leapfrogging other bad teams that just lose quietly.

Why this could be awesome for the draft
- Bottom teams get rewarded for fighting and winning tough games (no more pure tanking incentive).
- A scrappy team that pulls off upsets when hopeless could jump to #1 pick.
- Playoff teams stay untouched—they earn their high seeds by succeeding, so their picks stay late.
- It adds drama: Late-season games between lottery-bound teams become must-watch because every win could mean a huge draft boost.

At season's end, rank the non-playoff teams by total points earned → that's your draft order for picks 1 through (whatever number of lottery teams there are).



The reason I call this a “theory” more than an “idea” is because I haven’t been able to validate this point scoring system against historical seasons to see how it plays out, if any weighting would make it work better, etc. I asked AI to give me the results on the 2024-25 season and it basically said it’s too hard to gather the data points to give me an answer.

Very curious to see how others think this “theory” would work? And if anyone has the skills to put this theory to the test with some real data (I certainly don’t).

The more complex you make a system the more teams will manipulate it to their advantage to get a generational player. Imagine the Wemby draft. 2/3’s of the teams would be losing at the start of the season to position themselves to win on the back-end so they can draft Wemby.

If the NBA does this wrong, which a high likelyhood exists they will, they risk permanently and significantly damaging the league.

 

[twslam07]

The more complex you make a system the more teams will manipulate it to their advantage to get a generational player. Imagine the Wemby draft. 2/3’s of the teams would be losing at the start of the season to position themselves to win on the back-end so they can draft Wemby.

Is it complex? Can you demonstrate your POV under the rules I proposed? Let’s play it through

 

[Insomniacal Fan]

I know everyone has an opinion about how to fix tanking, but I have an idea (more of a theory) on how to fix the issue.

I was too lazy to type up the entire explanation so I dropped in the key bullet points in AI and asked it to build that summary for me. This is how it summarized my idea…

NBA "Games-Back Bonus" Draft Points System
(A creative idea to fix tanking and make draft order more exciting and merit-based)

The Big Motive
Right now, bad teams sometimes lose on purpose late in the season to get better draft picks (the famous "tanking" problem). This system flips that upside down:

It rewards teams for winning when they're far behind in the standings—instead of punishing wins or rewarding losses.

The goal? Use these points to decide the draft order for the next year's NBA Draft. The team with the most points gets the #1 overall pick, then #2, and so on—but only among teams that miss the playoffs. Any team that actually makes the playoffs (top 8 in their conference) gets their normal late picks (15th–30th range).

It gives struggling teams a real reason to compete and pull off upsets, while still letting true contenders focus on winning now.

How the points work, super simple
Every time a team wins a game, they earn points equal to how many games back (GB) they are from the 8th seed (last automatic playoff spot) in their own conference on that day.

- Farther back = bigger reward for a win (because it's harder to win when you're way out of it).
- Close to or above the 8th seed = small points (or even 0 if they're ahead).
- Losses = 0 points added that day.

Half-games count (e.g., 7.5 GB = 7.5 points). Points are only from wins, and the value changes daily as standings shift.

Quick example (Sacramento Kings style)
- Jan 1: Kings are 8.5 games back from West's 8th seed. They win → +8.5 points.
- They lose the rest of January → 0 points those days.
- Feb 1: Now 14 games back after the skid. They win → +14 points.
- Total so far: 22.5 points.

If they keep racking up big-GB wins while staying out of the playoffs, they could climb high in draft points—even leapfrogging other bad teams that just lose quietly.

Why this could be awesome for the draft
- Bottom teams get rewarded for fighting and winning tough games (no more pure tanking incentive).
- A scrappy team that pulls off upsets when hopeless could jump to #1 pick.
- Playoff teams stay untouched—they earn their high seeds by succeeding, so their picks stay late.
- It adds drama: Late-season games between lottery-bound teams become must-watch because every win could mean a huge draft boost.

At season's end, rank the non-playoff teams by total points earned → that's your draft order for picks 1 through (whatever number of lottery teams there are).



The reason I call this a “theory” more than an “idea” is because I haven’t been able to validate this point scoring system against historical seasons to see how it plays out, if any weighting would make it work better, etc. I asked AI to give me the results on the 2024-25 season and it basically said it’s too hard to gather the data points to give me an answer.

Very curious to see how others think this “theory” would work? And if anyone has the skills to put this theory to the test with some real data (I certainly don’t).

Without running this through simulations, it seems like this would incentivize teams to tank at the beginning of the season to build up a large GB differential, then start competing in the second half of the season to maximize the value of their wins. Which isn't necessarily worse than standard practice today, but it's still a little perverse.

I do appreciate the creativity of the solution though; I think it's fun for everyone to share their game design ideas.

 

[Turgenev]

Maybe adjust lottery odds to roster availability as an additional measure.

 

[twslam07]

Without running this through simulations, it seems like this would incentivize teams to tank at the beginning of the season to build up a large GB differential, then start competing in the second half of the season to maximize the value of their wins. Which isn't necessarily worse than standard practice today, but it's still a little perverse.

I do appreciate the creativity of the solution though; I think it's fun for everyone to share their game design ideas.

I’m not so sure.

It’s the promise of a new season. Do you think the Kings would come out “tanking” to start this past season or do you think they actually had a belief that they had a shot at the the playoffs? I think a lot of teams would second guess tanking from the get go with the idea of seeing where they have and if the team the GMs have assembled is good enough to compete.

Also, I think it’s likely self explanatory that the points are more skewed towards the end of the year. It’s hard to be 10 games back from the 8th seed after the first month but it’s much easier to be 10 games back from the 8th seed with 1 month left in the season.

The potential point total toward the end of the season is higher than earlier in the season. And if a team is trying to tank out of the gate, they’re not going to be picking up any points while losing. And if they do start winning (a fair amount) after losing, each win will likely be worth less points as they close in on the 8th seed.

But again, we’re taking in theory right now. I’d be very interested to see the data behind this idea.

 

[sactowndog]

Is it complex? Can you demonstrate your POV under the rules I proposed? Let’s play it through

You don’t provide enough detail but I assume at some point in the season you start to switch to counting points. If I’m a very good team, I have my best player like Sabonis get surgery in Nov so he misses maximum games up to your counting point. Other players I sit for teams in rest to try to lose just like they do today. Once the counting point is reached I bring all my best players back and compete for Wemby.

The NBA Cup would be destroyed because Wemby is a much better prize than the NBA Cup.

 

[sactowndog]

I’m not so sure.

It’s the promise of a new season. Do you think the Kings would come out “tanking” to start this past season or do you think they actually had a belief that they had a shot at the the playoffs? I think a lot of teams would second guess tanking from the get go with the idea of seeing where they have and if the team the GMs have assembled is good enough to compete.

Also, I think it’s likely self explanatory that the points are more skewed towards the end of the year. It’s hard to be 10 games back from the 8th seed after the first month but it’s much easier to be 10 games back from the 8th seed with 1 month left in the season.

The potential point total toward the end of the season is higher than earlier in the season. And if a team is trying to tank out of the gate, they’re not going to be picking up any points while losing. And if they do start winning (a fair amount) after losing, each win will likely be worth less points as they close in on the 8th seed.

But again, we’re taking in theory right now. I’d be very interested to see the data behind this idea.

The Kings wouldn’t because Vivek is an idiot but many smart teams for sure would do so.

 

[Capt. Factorial]

My problem with this approach, is that once you go away from a "closed" system based on an objective metric, then it stops being a simple game, it becomes just another marketplace. You might as well just be selling lottery tickets.

Maybe you could get away with all the owners voting for a czar to assign picks. Maybe that just becomes the commish's job; Silver decides who gets the top 5 picks, how would you like that?

A system based on an objective metric can be gamed. That's what "tanking" is, gaming the metric. Name the metric, and teams that want to tank will adjust accordingly.

I don't understand your "marketplace" comment.

As some may have internalized, I'm not at all a fan of Silver so giving him the keys to the draft order would not be on the top of my list. But the suggestion of having a czar or the commish subjectively determine the draft order is of the same approach that I have suggested. I think that using "the wisdom of crowds" (with the crowd being 30 NBA front offices) would be a better idea than assigning one person. It's harder to corrupt 30 people all with different self-serving agendas than it is to corrupt one. But that approach is in what I believe to be the right direction.

 

[sactowndog]

Maybe adjust lottery odds to roster availability as an additional measure.

So teams force players who are really hurt to play.

 

[twslam07]

You don’t provide enough detail but I assume at some point in the season you start to switch to counting points. If I’m a very good team, I have my best player like Sabonis get surgery in Nov so he misses maximum games up to your counting point. Other players I sit for teams in rest to try to lose just like they do today. Once the counting point is reached I bring all my best players back and compete for Wemby.

The NBA Cup would be destroyed because Wemby is a much better prize than the NBA Cup.

The counting point starts from the beginning of the season. How do you demonstrate your point from there?

 

[twslam07]

The Kings wouldn’t because Vivek is an idiot but many smart teams for sure would do so.

Again, demonstrate your point. Play it through so we can assess your criticism properly.

 

[Capt. Factorial]

Why this could be awesome for the draft
- Bottom teams get rewarded for fighting and winning tough games (no more pure tanking incentive).

The problem I immediately have with this suggestion is essentially the same problem I have with all of the others. It is trying to address the symptom (tanking) rather than address the underlying problem (draft order is open to manipulation via game results).

My two primary criteria for a preferred draft ordering system are (probably) both violated here:
1) The best picks should go to the worst teams (here winning - namely losing then winning - is a component; truly bad teams have a hard time winning, probably violated)
2) The draft order should not be dependent on a team's win/loss record, or any other metric the team can independently manipulate (definitely violated)

Very curious to see how others think this “theory” would work? And if anyone has the skills to put this theory to the test with some real data (I certainly don’t).

I don't have a great feel for whether this would give the most points to bad teams or mediocre teams (assuming no deliberate manipulation on top of that). I or somebody else would have to simulate it, and I may not have an opportunity tonight. It would be a minor task.

 

[sactowndog]

The counting point starts from the beginning of the season. How do you demonstrate your point from there?

So the best team gets the first draft pick? Again, as the Captain said, if you have rules of any sort Teams will figure out how to manipulate them.

 

[sactowndog]

Again, demonstrate your point. Play it through so we can assess your criticism properly.

If @Capt. Factorial has a good handle on it I will let him do it as we have the almost exact same perception and you will accept his critique much more likely than mine

 

[sactowndog]

Back to the point of this thread. Washington winning tonight is a good thing for us.

 

[twslam07]

The problem I immediately have with this suggestion is essentially the same problem I have with all of the others. It is trying to address the symptom (tanking) rather than address the underlying problem (draft order is open to manipulation via game results).

My two primary criteria for a preferred draft ordering system are (probably) both violated here:
1) The best picks should go to the worst teams (here winning - namely losing then winning - is a component; truly bad teams have a hard time winning, probably violated)
2) The draft order should not be dependent on a team's win/loss record, or any other metric the team can independently manipulate (definitely violated)

I don't have a great feel for whether this would give the most points to bad teams or mediocre teams (assuming no deliberate manipulation on top of that). I or somebody else would have to simulate it, and I may not have an opportunity tonight. It would be a minor task.

And you’re entitled to that opinion but your two principles are preferences at the end of the day.

Truth be told, I don’t know how my recommendation would play out, but it does (on the surface) seem like an idea that would reward weaker teams while also still incentivizing them to win.

 

[twslam07]

So the best team gets the first draft pick? Again, as the Captain said, if you have rules of any sort Teams will figure out how to manipulate them.

What? No. That’s not at all what my proposal outlined.

My proposal assigns points based on wins, but wins only get points if you are outside of the 1st-8th seed. And the points are weighted based on how far back you are from the 8th seed.

So no, the “best team” doesn’t get the first pick since…

1.) the teams that make the playoffs are eliminated from the #1-#14 picks
2.) if you’re the best team, you’re not racking up points with your wins because you are not any games back from the 8th seed

 

[Insomniacal Fan]

A system based on an objective metric can be gamed. That's what "tanking" is, gaming the metric. Name the metric, and teams that want to tank will adjust accordingly.

I don't understand your "marketplace" comment.

As some may have internalized, I'm not at all a fan of Silver so giving him the keys to the draft order would not be on the top of my list. But the suggestion of having a czar or the commish subjectively determine the draft order is of the same approach that I have suggested. I think that using "the wisdom of crowds" (with the crowd being 30 NBA front offices) would be a better idea than assigning one person. It's harder to corrupt 30 people all with different self-serving agendas than it is to corrupt one. But that approach is in what I believe to be the right direction.

On the 1 czar vs 30 front-offices; I am on the same page with you that this is fundamentally a similar proposition. I think they'd end up in roughly the same place, some of the 30 teams' individual motivations would cancel each other out, but other parts would resonate I don't assume that competitive balance would be something would be a resonant interest amongst the owners. I can imagine a priority for everybody with money in the league would be to direct picks to large media markets, and/or to make sure that highly marketable stars always receive talent infusions.

To explain the "marketplace" criticism, once you make the process of assigning picks completely arbitrary, you're suddenly allowing the real world to decide how talent is allocated. And if you were annoyed at gambling companies influencing NBA policy, just wait until picks are lobbied for by major corporations, religious organizations and nation-states. Why wouldn't the owners just start selling these picks?

Your "game-theory" based approach is naive in that you presume that all parties are playing the same game, while it simultaneously exposes the process to every form of external influence that exists. I don't think you can just handwave this away by saying the NBA could figure out a way to police corruption somehow. That's actually a load-bearing part of the whole scheme.

 

[sactowndog]

What? No. That’s not at all what my proposal outlined.

My proposal assigns points based on wins, but wins only get points if you are outside of the 1st-8th seed. And the points are weighted based on how far back you are from the 8th seed.

So no, the “best team” doesn’t get the first pick since…

1.) the teams that make the playoffs are eliminated from the #1-#14 picks
2.) if you’re the best team, you’re not racking up points with your wins because you are not any games back from the 8th seed

Again it can be manipulated and most likely by the team that finishes the highest in the lottery. As long as you have control you have a path to maximizing your draft results. And trust each team employs enough mathematicians to figure it out.

 

[iowamcnabb]

Number 2 is the most interesting. The thing is though you know teams like the Kings would exploit the heck out of that. Suck all the way to the deadline then go out and sign/trade for 10 Westbrooks and try to make the play in, lol.

Do it over 2 years like the W does.

 

[sactowndog]

The problem I immediately have with this suggestion is essentially the same problem I have with all of the others. It is trying to address the symptom (tanking) rather than address the underlying problem (draft order is open to manipulation via game results).

My two primary criteria for a preferred draft ordering system are (probably) both violated here:
1) The best picks should go to the worst teams (here winning - namely losing then winning - is a component; truly bad teams have a hard time winning, probably violated)
2) The draft order should not be dependent on a team's win/loss record, or any other metric the team can independently manipulate (definitely violated)

I don't have a great feel for whether this would give the most points to bad teams or mediocre teams (assuming no deliberate manipulation on top of that). I or somebody else would have to simulate it, and I may not have an opportunity tonight. It would be a minor task.

Full disclosure….. I thought your articulation was so well stated I copied it for a couple threads on RealGM. I did switch it up and put the criteria first and then the problem.

But I could not articulate it more succinctly and clearly so I shamelessly copied it.

 

[Capt. Factorial]

And you’re entitled to that opinion but your two principles are preferences at the end of the day.

Truth be told, I don’t know how my recommendation would play out, but it does (on the surface) seem like an idea that would reward weaker teams while also still incentivizing them to win.

OK, it took some time, but I know my role around here, so...

I did several simulations of 1000 "full seasons". The reason I did several is that it's a bit tough to get simulated win totals to match up with realistic-looking league win totals. But by doing this, I think I have a handle on what this proposal does.

First, I want to list my simplifications. I don't think they should really make much of a difference, but for full disclosure:
1) I modeled 30 teams, but only ONE conference, thus, the playoff line was 16th place, not 8th place
2) I modeled 82 games, but there were no off days. Thus, every team played every day. This also means that the Games Behind measure was never fractional (e.g. never 1.5 or 2.5 etc., it was always 1 or 2 or 3...)
3) The schedule was truly random, not modeled on an actual NBA schedule
4) Each team was given an integer "true strength" value N, which could be thought of as N lotto balls. For each simulated game, Team A's lotto balls and Team B's lotto balls were combined, and one random ball was selected to determine the winner. Teams' lotto balls did not change over the course of a season - if Team A started with 30 lotto balls, then Team A had 30 lotto balls in every game of the season. Team A would have had a 33% chance of beating a team with 60 lotto balls, a 50% chance of beating a team with 30 lotto balls, a 67% chance of beating a team with 15 lotto balls, etc.
5) No team was doing any manipulation of any sort, no tanking, no nothing.

The gist of the results: For teams with 21 season wins or more, the long term average draft position exactly matched the team standings. In the long term, for teams with 21 wins or more, this proposal does not produce a different draft order than "reverse standings". Teams with fewer than 21 wins are penalized in the draft order. A team with 18 wins ends up about 2 spots in the draft order BELOW a 21-win team, a team with 13 wins ends up about 5 spots in the draft order BELOW a 21-win team, a team with 7 wins ends up about 8 spots in the draft order BELOW a 21-win team.

This proposal seems to favor about 21 wins, in the long term.

 

[Capt. Factorial]

And you’re entitled to that opinion but your two principles are preferences at the end of the day.

Well, I think they are sensible principles.

For #1, the point of the "competitive balance draft" is to help competitive balance. The way you do that is to give a leg up to the bad teams, not to give a leg up to the teams in the middle. This seems common sense. Doing anything else will hurt competitive balance because bad teams (e.g. teams who cannot sign good free agents based on geography) will stay bad.

For #2, if you want to eliminate tanking, you cannot use an objective metric because teams will manipulate the metric (which is basically the definition of tanking).

So if the goal is to eliminate tanking AND to improve competitive balance, I think both of those principles are NECESSARY, not preferences.

 

[Capt. Factorial]

On the 1 czar vs 30 front-offices; I am on the same page with you that this is fundamentally a similar proposition. I think they'd end up in roughly the same place, some of the 30 teams' individual motivations would cancel each other out, but other parts would resonate I don't assume that competitive balance would be something would be a resonant interest amongst the owners. I can imagine a priority for everybody with money in the league would be to direct picks to large media markets, and/or to make sure that highly marketable stars always receive talent infusions.

I don't think the owners would be thinking in terms of competitive balance. They would be thinking in terms of not improving opposing teams. the point is that owners would want the #1 pick this year to go to Sacramento, or to New Orleans, or to Utah, or to Washington. They wouldn't really want the #1 pick to go to Indiana, because they will be getting Hali back next year. They definitely wouldn't want the #1 pick to go to OKC! It is fundamentally in everybody's interest to make sure that the best players go to the least threatening teams, and nobody has to care about "competitive balance" for that to shake out.

You seem to suggest that, for instance, the Celtics would prefer a high pick go to the Lakers (a large market) than to the Grizzlies (a small market). But making the Lakers into a super team is suicidal for the Celtics. The Celtics want to win. Why would they help the Lakers, who can sign premium free agents on the regular? I don't see it happening.

To explain the "marketplace" criticism, once you make the process of assigning picks completely arbitrary, you're suddenly allowing the real world to decide how talent is allocated. And if you were annoyed at gambling companies influencing NBA policy, just wait until picks are lobbied for by major corporations, religious organizations and nation-states. Why wouldn't the owners just start selling these picks?

I don't understand what you mean by "selling these picks". If you mean that owners would take money to manipulate their rankings, it's possible that some might. And it's possible that the NBA would not be able to effectively police this sort of corruption. But remember that there would be 30 different front offices making rankings, which would be averaged. The maximum amount of "push" that one corrupt front office could conceivably make would be about one slot in the draft. Massive moves would require widely coordinated corruption. Who is going to do this? What nation-state is going to spend two billion dollars to bribe 20 or 25 franchises to put the Lakers at the top of the list so that L.A. can get AJ Dybantsa? This is a laughable thought. And not only that, but in most cases it would be blatantly obvious.

Interestingly, I think that the biggest disincentive to corruption is that the risk/reward ratio is just really bad. Corrupt one team, they're only one thirtieth of the total. Corrupt two teams, they're only one-fifteenth of the total. Corrupt three teams...oh, whoops! The third team had a conscience and tattled! And following the subsequent investigation the first two teams just got nailed!

 

[sactowndog]

OK, it took some time, but I know my role around here, so...

I did several simulations of 1000 "full seasons". The reason I did several is that it's a bit tough to get simulated win totals to match up with realistic-looking league win totals. But by doing this, I think I have a handle on what this proposal does.

First, I want to list my simplifications. I don't think they should really make much of a difference, but for full disclosure:
1) I modeled 30 teams, but only ONE conference, thus, the playoff line was 16th place, not 8th place
2) I modeled 82 games, but there were no off days. Thus, every team played every day. This also means that the Games Behind measure was never fractional (e.g. never 1.5 or 2.5 etc., it was always 1 or 2 or 3...)
3) The schedule was truly random, not modeled on an actual NBA schedule
4) Each team was given an integer "true strength" value N, which could be thought of as N lotto balls. For each simulated game, Team A's lotto balls and Team B's lotto balls were combined, and one random ball was selected to determine the winner. Teams' lotto balls did not change over the course of a season - if Team A started with 30 lotto balls, then Team A had 30 lotto balls in every game of the season. Team A would have had a 33% chance of beating a team with 60 lotto balls, a 50% chance of beating a team with 30 lotto balls, a 67% chance of beating a team with 15 lotto balls, etc.
5) No team was doing any manipulation of any sort, no tanking, no nothing.

The gist of the results: For teams with 21 season wins or more, the long term average draft position exactly matched the team standings. In the long term, for teams with 21 wins or more, this proposal does not produce a different draft order than "reverse standings". Teams with fewer than 21 wins are penalized in the draft order. A team with 18 wins ends up about 2 spots in the draft order BELOW a 21-win team, a team with 13 wins ends up about 5 spots in the draft order BELOW a 21-win team, a team with 7 wins ends up about 8 spots in the draft order BELOW a 21-win team.

This proposal seems to favor about 21 wins, in the long term.

I would like to point out @Capt. Factorial did this in an evening to start to model the ideal strategy. Every team in the NBA has an entire analytics department of mathematical statistical experts.

I guarantee they would run even more simulations to develop the best strategy to maximize their draft results and then seek to execute that strategy.

 

[HndsmCelt]

I am a strong proponent of a game-theoretical approach where (in a simple description) each NBA front office ranks the teams in the order they think teams should pick, and those rankings are combined/averaged to get the final draft order. Teams will be motivated to direct the top picks to the teams with the bleakest future outlooks, and won't be fooled by tanking, because they will take rosters/contracts/ability to sign FAs into consideration. It takes the objective "team record" measure completely out of it, which then gives teams no real incentive to tank. That's the basic gist of what I think would work well. Lots of details to iron out, but from a general point of view the approach should be solid.

I am truly impressed that a person so dedicated to quantitative analysis would essentially try to solve a problem with the qualitative solution. Serious much props.