BCR Fixes the NBA


Fans aren’t asking for perfection when it comes to the NBA, just something a little better. The NBA’s been beaten down this season with concerns over tanking, expansion, scheduling issues, and load management. There’s no one overarching fix for these issues, and the goal isn’t to disrespect or emphasize flaws already identified. But these issues are bringing change. Frankly, I’m not optimistic it’s good change. The regime of Adam Silver and league executives seem determined in making radical changes.

It’s concerning because every idea floated across so far doesn’t resemble a fitting solution. In my opinion, most of the solutions to combat these issues are too narrow minded, unbothered by the cost of everything else that’d be affected: a zero-sum game. That’s a dangerous place to be. It led me down a rabbit hole of potential solutions derived by data. And while everyone has their own fix, I’m certain you haven’t thought of these ones. This is how the BCR would fix the NBA.

BCR Data

The NBA’s been under the microscope in terms of competition, or lack thereof, the last few seasons. My dataset includes over 2300 NBA games, in which I investigated the average margin of victory in that span. Over the sample size, the average margin of victory is 13 points- about 4 possessions difference over the last 3+ years of regular season games. Drilled down further, 34% (779) were decided by 15+ and 20% (468) by 20+. So, on any given night the last 3-4 seasons, your chances to watch a close game weren’t favorable. Actually, if you’d watched a regular season game over the last few seasons, you have a better chance watching a game decided by 15+ points than 7 or less.

Why does this matter? I’d argue it’s the single biggest issue of the league given that tanking is more than likely the culprit, and expansion is on the horizon

Expansion

The data above tells us the NBA is suffering from a basic economics lesson: if supply is less than demand, a shortage occurs. In this case the shortage is measured in competitiveness. Tanking aside, the numbers lead to serious concern that the talent pool of players cannot meet the demand of 30 teams to be competitive. And yet here we are talking about expansion. 

Expansion should be the farthest thing from the NBA’s radar. If we expand an already struggling talent pool from each roster, we’re essentially cutting the crust off a sandwich hoping to make a new one from the scraps. In the end, all we get is a shittier sandwich. The margin of victory data supports this, but so do team records. An average of 7 teams per season struggle to win 33 games (40%-win rate) over the last 8 seasons.

Expansion would dilute the talent level and in return decrease competitiveness and parity. Based solely on the on-court product and competitive concerns, the NBA would probably benefit from a league wide shrinkage to 28 or 26 teams. It will never happen, the league makes too much money, but be warned that expansion will only harm the league in its current state.


Tanking

There’s no right answer, everyone’s got their fixes and opinions, and I am no different.

My initial belief on tanking is there’s no reason to fix it, because it works. For most of my life, the Philadelphia 76ers were nothing more than average and in the 2010s, they were awful. They’re currently on pace to make the playoffs in eight of the last nine seasons. The Pistons and Thunder were no different. Tanking works.

Unfortunately, it doesn’t matter if I believe tanking is fine, because change is inevitable. So, if we’re changing the processes, the BCR can help. With plenty of ideas already expressed, the BCR’s solely analytical approach can offer a different view. There are 3 different options to determine the order that the BCR’s identified. The first step was creating a new, simple set of criteria to determine draft lottery odds. The criteria consist of 3 variables: Wins, avg. point differential, and team salary. The wins are on a “tier-adjusted” scheme to put larger weighted values on the worst team’s vs teams at the end of the lottery. The second variable, point differential is slightly heavier weighted, and the salary is weighted less to act as a differentiating variable vs a deciding variable.

BCR Lottery Formula = (1-W%) * (Avg. Point Differential * 1.1) * ((Salary/100)*.55)

Let’s just flush out all the counter arguments: “by incorporating point differential, you’re rewarding teams for losing by more points? That will incentivize teams to lose by more.”  That is true, but only if draft/lottery odds are in 1 – “x teams” order. I’ll go into that approach later. Point differential overall is aiming to prioritize teams losing by more in situations where number of wins is similar between teams.

Secondly, the salary numbers are included to “reward” a team for at least tying a financial amount to losing. Teams spending more are at least trying to win in terms of financial commitment. If a team decides to tank by employing lesser players, at smaller contracts, they may get penalized. A GM could sign a bunch of lofty contracts on players less deserving in attempt to “loophole” this issue, but good luck getting the owner’s satisfaction in paying more just to get less. And what about teams with numerous rookie contracts? If a team of mostly rookie contracts is already performing that poorly, the salary won’t keep them from a high pick because the other two variables outweigh the salary impact. The main concern is, “does this salary variable help the big market teams too much, given they usually spend more than other franchises?” A lot of the teams with the much higher payrolls aren’t near the lottery. The 14 teams in the lottery’s payroll are much closer valued than teams on the outside. 

We have our criteria and the lottery system still needs to be determined. There are 3 scenarios to pick from. In all 3 scenarios, the lottery shrinks from 14 teams to 11 in hopes to deter more teams from mid-season tanking efforts.

Draft Lottery Scenarios

  1. Only the top 6 teams in the BCR draft order get a shot at the number 1 pick, and the highest odds go to team #6 in the order, followed by 5-1.
  2. Give “tiered odds” to teams 1-9 for chances at the number 1 pick.
  3. Take the median of the bottom 11 teams in the BCR lottery order, those 3 teams get the best chances at the #3 pick, followed then by 1-4 & 8-11.

Scenario 1

Only the top 6 teams in the draft order get a shot at the number 1 pick, and the highest odds go to team #6 in the order, followed by 5-1. Teams 7-11 ceiling is the #2 pick, but their odds are scaled proportionately after 1-6. 

This option is best suited if the NBA is absolutely convinced that teams are tanking and not just falling short due to a talent gap, in terms of competitiveness. Because in this case, instead of fighting to be the worst, you’re fighting to be the 6th worst.

It gives teams an incentive not to just roll over every game to hope for the highest odds. The difference between the worst team and 6th worst team currently in NBA standings is 8 wins. The question is: if this proposal were implemented and teams were aware pre season, would that gap of 8 games shrink? The hope is yes. The second benefit of this option is limiting the chances at the number 1 pick to just 6 teams. It may seem like too small of a group, but doing so deters the teams in the 8-12 range from midseason tanking because the reward in doing so is lesser.

The NBA would need to be extremely confident that teams are tanking, not just naturally uncompetitive, to enable a lottery system like this. This option becomes dangerous, if the teams at the bottom truly are so uncompetitive because taking away odds to have the #1 pick does real harm. The draft odds could look like this:

*Note: It is difficult to create a mathematically perfect lottery matrix. The specific odds are just to give an idea under this scenario, the important piece is the proposed draft order

It’s clear the incentive to finish 6th worst is incredibly enticing, because under this scenario the team’s odds at the #1 pick would nearly double the highest odds for #1 as the worst ranked team under the current lottery system.

Scenario 2

Give scaled odds to teams 1-9 for chances at the number 1 pick.

Under this option, it’s pretty straight forward: the NBA would just utilize the BCR’s weighted formula for lottery odds to decide order 1-9. The main edge is cutting out 5 teams from the lottery, again in hopes to limit the enticing aspect of tanking for a team somewhere in the middle of the standings. The other main factor is just letting the BCR draft formula penalize the “tankers.” How it achieves this is through the point differential variable. It’s not 100% “fool proof,” but let’s say for example: a team with 20 wins has a point differential of -7.5, but a team with 17 wins has a point differential of -4.2. The team with 17 wins would be penalized. The variable is built on the assumption that the team with 17 wins likely had more chances to win, given the point differential, but maybe made a decision to tank and lose a few more games.

Although it’s an aggressive approach, that seems to be how Adam Silver views tanking. Here’s an idea of what the odds could look like under this option:

*Note: It is difficult to create a mathematically perfect lottery matrix. The specific odds are just to give an idea under this scenario, the important piece is the proposed draft order

Scenario 3

Take the median of the bottom 11 teams in the BCR lottery order, those 3 teams get the best chances at the #1 pick, followed then by 1-4 & 8-11.

By far the most aggressive and most penalizing scenario to combat tanking. It’s a mixture of both the previous 2 options, in which the teams that finish 5th ,6th, & 7th worst have the best odds at the number 1 pick vs the other 8 in the lottery. By giving 5th, 6th, & 7th worst teams the best odds at the top spot, you’re hoping that the play amongst the bottom teams improves. The NBA would need the improvement to be drastic enough that the gap between the worst and 5th worst is minor and getting the 4th highest odds as the worst team isn’t a terrible consolation. Ideally the league would need the bottom 1-7 teams to be within 4 losses of each other. In my opinion, that’s near impossible and assumes tanking is currently in full force.

*Note: It is difficult to create a mathematically perfect lottery matrix. The specific odds are just to give an idea under this scenario, the important piece is the proposed draft order

If I had to pick one of these scenarios, I’d probably pick something similar to scenario 1. There’s clearly tanking occurring league wide and I do think the bottom teams could play better. Scenario 1 could expose just how much of the gap is tanking vs true talent gap. A concern to this approach is: “if a team is 7th (or 8th-10th) worst, what’s stopping them from tanking to 6th?” The BCR formula consisting of 3 variables makes it more difficult to guarantee going from 7th to 6th. Under the current system all the team would have to do is lose games to move positions. Salary is fixed and average point differential wouldn’t be affected in the short term. The tank would have to be perfectly timed and a team would almost have to calculate how much to lose by in order to impact point diff.


Scheduling

The 82-game regular season schedule contributes to multiple issues: load management, injury management, competitiveness, etc. Radical change isn’t necessary to have a significant impact on these topics. There’s plenty of studies and data showing that as the season progresses, more load management techniques are utilized- no duh. Is the issue the number of games, the length of games, or the calendar the season follows? The BCR suggests the NBA adapt to the NFL approach of bye weeks.

I don’t have the perfect example of creating the schedule with the bye weeks, but the NBA has all the resources to figure it out and make it work. Each team could get 2 bye weeks a season, not including All Star Weekend. There’s always that lull in the middle of the season where players mention the “grind” of January-March. If the NBA established bye weeks, it could set them in the offseason and lessen the grueling stretches of the regular season. I have a feeling the players would welcome this.

Preferably, the first bye weeks wouldn’t start until the 15-20 game mark. And from that date going forward, every week, one Eastern Conference team and one Western Conference team would be utilizing the bye. Then once all 30 teams got their first bye week, the cycle would restart. Maybe the second bye week window starts around the 40-50 game range. As the schedule stands, the season is 24 weeks long and you’d need 15 weeks to fully incorporate the 2 teams per bye week proposal. The NBA could chose to do a lot of things to shorten the 15 weeks: send 4 teams on bye weeks not 2, or start a new “bye week” every 5 days. Again, the NBA has the resources to make this schedule change work.

You could argue, players would get burnt out slower, injuries could be rehabbed better- no rushing back for games or trying to play through a nagging injury. Competitiveness, or parity, could remain intact. Imagine a middling team fresh off a bye matching up with a top 5 team playing their 9th game in 20 some days. That’s a good promo for game 57 on a random Tuesday night in February.

The question you have to ask about topics like load management or parity is: is it transitive? Will any solution make an impact, and if so, how much of an impact? If you enable a fix like bye weeks will they actually reduce these issues, or will franchises load manage players before/after bye weeks to ensure longer breaks? I believe the bye week system could work perfectly for the NBA.


Fixes

I’m not a believer that there is a perfect answer to all these issues. But I do believe some solutions are way closer to the correct answer while others don’t even understand the question. There’s so many great things about the NBA but it is an odd time as a fan. The league’s never been more popular and has never been under so much fire. If solutions do come within the next weeks or months, I just hope they aren’t implemented in attempts to retroactively fix the lingering problems. They have to be forward looking solutions.

Posted:

Updated:

Member of The FWAA

*If viewing on mobile device, click icon near URL along bottom of screen. Hit the smaller “A” icon until text fits screen