No one wants to talk about the Sixers right now, so Beckett dives into the divisive tanking debate and explains why the NBA is trying to fix a problem of its own creation.
1) The Women's hockey model. This is, once mathematically eliminated, each win is a point towards draft rank. This makes bad teams want to win after getting eliminated but could doom a bad team to be bad forever if they're THAT bad.
2) The "Your Team Sucks" draft. At the start of each season, each team (in reverse record order, so worst team first) picks a another team's draft pick for that year. You can't pick yourself. This completely eliminates the desire to lose and is my favorite solution.
I also am a fan of the drafting ones for sure, also because they would be incredibly fun to watch and spark some rivalries. The only questions you run into there are around how you trade picks. Definitely like that one though!
"But limiting protections to top-4 or top-14 just creates a new arbitrary cutoff for teams to tank toward."
They aren't arbitrary, but I'm not sure that they work.
Top-4 protection is based on the idea that you can't plan to win the lottery. If you do, good for you, but you can't tank your way into it. You can tank to raise the odds, though, so it does create an incentive to lose. It might even incentivize the Sixers to tank out of the playoffs this year - just to have a very small chance at keeping their pick.
Top-14 protection is based on the idea that teams want to be in the playoffs more than they want a late lottery pick. I'm not sure that's accurate, either.
A more important modification would be to mandate that every first round pick that is successfully protected eventually converts into an unprotected first at some point (even if the protections kick in for multiple years). If a team has to give up a first eventually, there's less incentive to actively try to protect it for a year.
Arbitrary was likely the wrong word in this case. Just making a new cutoff that teams will tank towards was the point which you touch on in the comment here
Consider Carry-Over Lottery Allocation (COLA). If the draft order is based on regular-season standings, tanking is inevitable. COLA uses playoff track record instead. It circumvents the impossibility of helping weak teams while disincentivizing tanking. It is also a better measure of weak team anyway. A team could punt a season or just have an injured star and get a back record while still being a fundamentally strong team. Teams that miss the playoffs year after year are the truly weak teams.
I don’t hate this idea! Couple main issues you’d need to figure out: first, it doesn’t fix the problem of higher end teams tanking to miss the playoffs in order to get into the ticket group (which would be similar to the flattened lottery odds). Second, the difference between a team like the Kings who are truly bad and the Blazers who are not that bad (but have also missed the playoffs for the same amount of time) is large and wouldn’t be accounted for with this method. Finally, it prolongs tanks because a team that becomes bad in one season would need to put together several seasons of bad performances to get the odds they need.
In the current NBA draft system, if a team has playoff success for a while and then blows it up, tanks for a year, and is lucky enough to win a top lottery pick, there are two terrible outcomes there. First, the tanking is simply uncompetitive, unwatchable basketball. Second, the team "jumps the line" and gets an immediate top draft pick that other teams have been hoping for over the span of years.
Under COLA (or any system that measure team weakness by playoff track record), a rebuilding team may in fact take longer to get the draft help they need. That is a feature, not a bug. There are always going to be more bad teams than top draft picks. The idea is (or should be) that all teams have an equal opportunity to compete. Teams that have had neither playoff success nor recent draft help are the ones who should be the ones with the best odds of getting a top draft pick. The teams in earlier stages of rebuilding will still get first-round draft picks. If they draft well and get strong enough for playoff success before winning a top pick, that's great. If not, they will eventually get the top draft odds.
The Carry-Over Lottery Allocation (COLA) framework is:
* Each team has a stockpile of tickets that grows and shrinks over time, but extra losses never give extra tickets. (no tanking)
* Tickets carry over from year to year.
* Stockpiles are reduced when a team wins the lottery or wins in the playoffs. (favoring weak and unlucky teams over time)
Within that framework, there are many variants. One of the main design decisions is the exact set of rules to allocate tickets and reduce stockpiles. As part of that - how do we measure degrees of “winning” - in either the lottery or the playoffs?
That goes directly to the Kings vs Blazers issue. Based on the original version of COLA (in the research paper on arXiv), the Kings would have 7109 tickets vs 3422 for the Blazers. With more emphasis on recent playoff success and less emphasis on lottery wins, their totals would be closer.
I’ll address those three separately. First: teams tanking down into the lottery.
In order to design an incentive-compatible draft system, it’s necessary to know what teams actually value and a rough estimate of how much: playoff success, regular-season wins, high draft picks, money. Money is largely a function of playoff success, and we all know regular-season wins are worth very little. Ultimately, there has to be some level of playoff success that teams value more than high draft picks. Correctly identifying that level is important. (There are a few options to tackle that problem - that’s worth a separate conversation. There are tradeoffs to consider there.) But assume we know what that level of playoff success is. Let all teams that don’t reach that level into the lottery.
If all teams are guaranteed to either reach that level of success or get some draft help, and extra losses don’t earn any extra draft help, then they have no incentive to lose. The only way to get draft help is to miss out on a level of playoff success that they actually prefer, and once they get the draft help extra losses do nothing.
I agree that half the problem is the league thinking they can erase tanking as a whole. It will persist through whatever system is in place, just gotta curb it as much as possible
You didn't cover my two favorite solutions!
1) The Women's hockey model. This is, once mathematically eliminated, each win is a point towards draft rank. This makes bad teams want to win after getting eliminated but could doom a bad team to be bad forever if they're THAT bad.
2) The "Your Team Sucks" draft. At the start of each season, each team (in reverse record order, so worst team first) picks a another team's draft pick for that year. You can't pick yourself. This completely eliminates the desire to lose and is my favorite solution.
I also am a fan of the drafting ones for sure, also because they would be incredibly fun to watch and spark some rivalries. The only questions you run into there are around how you trade picks. Definitely like that one though!
"But limiting protections to top-4 or top-14 just creates a new arbitrary cutoff for teams to tank toward."
They aren't arbitrary, but I'm not sure that they work.
Top-4 protection is based on the idea that you can't plan to win the lottery. If you do, good for you, but you can't tank your way into it. You can tank to raise the odds, though, so it does create an incentive to lose. It might even incentivize the Sixers to tank out of the playoffs this year - just to have a very small chance at keeping their pick.
Top-14 protection is based on the idea that teams want to be in the playoffs more than they want a late lottery pick. I'm not sure that's accurate, either.
A more important modification would be to mandate that every first round pick that is successfully protected eventually converts into an unprotected first at some point (even if the protections kick in for multiple years). If a team has to give up a first eventually, there's less incentive to actively try to protect it for a year.
Arbitrary was likely the wrong word in this case. Just making a new cutoff that teams will tank towards was the point which you touch on in the comment here
Consider Carry-Over Lottery Allocation (COLA). If the draft order is based on regular-season standings, tanking is inevitable. COLA uses playoff track record instead. It circumvents the impossibility of helping weak teams while disincentivizing tanking. It is also a better measure of weak team anyway. A team could punt a season or just have an injured star and get a back record while still being a fundamentally strong team. Teams that miss the playoffs year after year are the truly weak teams.
I don’t hate this idea! Couple main issues you’d need to figure out: first, it doesn’t fix the problem of higher end teams tanking to miss the playoffs in order to get into the ticket group (which would be similar to the flattened lottery odds). Second, the difference between a team like the Kings who are truly bad and the Blazers who are not that bad (but have also missed the playoffs for the same amount of time) is large and wouldn’t be accounted for with this method. Finally, it prolongs tanks because a team that becomes bad in one season would need to put together several seasons of bad performances to get the odds they need.
Thanks for these comments.
Now to the third point: Prolonged tanks.
In the current NBA draft system, if a team has playoff success for a while and then blows it up, tanks for a year, and is lucky enough to win a top lottery pick, there are two terrible outcomes there. First, the tanking is simply uncompetitive, unwatchable basketball. Second, the team "jumps the line" and gets an immediate top draft pick that other teams have been hoping for over the span of years.
Under COLA (or any system that measure team weakness by playoff track record), a rebuilding team may in fact take longer to get the draft help they need. That is a feature, not a bug. There are always going to be more bad teams than top draft picks. The idea is (or should be) that all teams have an equal opportunity to compete. Teams that have had neither playoff success nor recent draft help are the ones who should be the ones with the best odds of getting a top draft pick. The teams in earlier stages of rebuilding will still get first-round draft picks. If they draft well and get strong enough for playoff success before winning a top pick, that's great. If not, they will eventually get the top draft odds.
Second: the Kings vs the Blazers.
The Carry-Over Lottery Allocation (COLA) framework is:
* Each team has a stockpile of tickets that grows and shrinks over time, but extra losses never give extra tickets. (no tanking)
* Tickets carry over from year to year.
* Stockpiles are reduced when a team wins the lottery or wins in the playoffs. (favoring weak and unlucky teams over time)
Within that framework, there are many variants. One of the main design decisions is the exact set of rules to allocate tickets and reduce stockpiles. As part of that - how do we measure degrees of “winning” - in either the lottery or the playoffs?
That goes directly to the Kings vs Blazers issue. Based on the original version of COLA (in the research paper on arXiv), the Kings would have 7109 tickets vs 3422 for the Blazers. With more emphasis on recent playoff success and less emphasis on lottery wins, their totals would be closer.
I’ll address those three separately. First: teams tanking down into the lottery.
In order to design an incentive-compatible draft system, it’s necessary to know what teams actually value and a rough estimate of how much: playoff success, regular-season wins, high draft picks, money. Money is largely a function of playoff success, and we all know regular-season wins are worth very little. Ultimately, there has to be some level of playoff success that teams value more than high draft picks. Correctly identifying that level is important. (There are a few options to tackle that problem - that’s worth a separate conversation. There are tradeoffs to consider there.) But assume we know what that level of playoff success is. Let all teams that don’t reach that level into the lottery.
If all teams are guaranteed to either reach that level of success or get some draft help, and extra losses don’t earn any extra draft help, then they have no incentive to lose. The only way to get draft help is to miss out on a level of playoff success that they actually prefer, and once they get the draft help extra losses do nothing.
I agree that half the problem is the league thinking they can erase tanking as a whole. It will persist through whatever system is in place, just gotta curb it as much as possible
It will persist, as long as losing in some way makes your team better in the future. It doesn't have to be that way