The Cleveland Cavaliers got 10 years older by trading Darius Garland for James Harden at the 2026 trade deadline. To many, myself included, this indicated that the Cavs were finally all-in on winning now. There was an assumed subtext to the trade that if things went poorly in the postseason, more all-in, win-now moves could be coming.
And then the NBA Board of Governors approved the “3-2-1 Draft Lottery” reform.
How are these two topics connected? They are linked because the biggest remaining trade chip for the Cavs to further go win now is Evan Mobley. And with that reform, the idea of trading Mobley became much, much riskier for the Cavaliers.
The headlines for the 3-2-1 reform are that it is meant to curb tanking and that it smooths out the odds for the top pick. But if you dig into the math behind the proposal, then you see that it is far more than that. The reform is a complete and total redistribution of where new talent goes when entering the league.
Exploring this topic is a math-heavy endeavor, so for the sake of brevity, we shall start at the conclusion:
The reform does not just address tanking. It completely changes the incentive structure for team building. It does not just smooth out odds; it also greatly increases the variance of the result for all teams. This has two major impacts:
The safest place to rebuild is no longer the bottom of the standings. It’s in the middle.It is highly unlikely that an age-aligned core of All-Star talent can be built purely through the draft
The change is so extreme that the end result leads to these outcome tables:
Target # of All-StarsTeam CategoryDraft SystemYears to Acquire (Unlucky)Years to Acquire (Typical)Years to Acquire (Lucky)2Bad TeamsOld7532Bad TeamsNew 3-2-111752Non-PlayoffOld8642Non-PlayoffNew 3-2-110642Playoff BubbleOld11752Playoff BubbleNew 3-2-110642Play-In TeamsOld211382Play-In TeamsNew 3-2-11395
The table above assumes a team finishes in the same grouping year-over-year. If they do, how long would it take to acquire two All-Stars if you get unlucky (25th percentile outcome), get a typical result (median outcome), or get lucky (75th percentile outcome)? For our purposes, we are broadly grouping teams as such:
GroupTeam Finish PositionBad Teams1-3Non-Playoff Teams4-7Playoff Bubble8-10Play-In Teams11-16
Look at those years carefully and remember that the rookie contract extension comes after year four. The odds aren’t just shifting away from the bad teams in favor of middle finishers. It is now difficult for anybody to acquire two All-Stars on the same timeline without getting very lucky. It gets even worse when looking at the odds of acquiring three All-Stars.
Target # of All-StarsTeam CategoryDraft SystemYears to Acquire (Unlucky)Years to Acquire (Typical)Years to Acquire (Lucky)3Bad TeamsOld10753Bad TeamsNew 3-2-1161283Non-PlayoffOld12963Non-PlayoffNew 3-2-1141073Playoff BubbleOld161183Playoff BubbleNew 3-2-1141073Play-In TeamsOld3021143Play-In TeamsNew 3-2-119139
The best teams are built by hitting on many All-Stars in a tight time window. Think, Kevin Durant – Russell Westbrook – James Harden with the Oklahoma City Thunder or Stephon Castle – Victor Wembanyama – Dylan Harper with the San Antonio Spurs. These team constructions are simply not going to happen under the new system, or at least would require far more luck than it currently does. The tables above are just based on math. They don’t even take into account the hard rules the 3-2-1 reform implements, such as no back-to-back number one overall picks and no picking in the top five three years in-a-row.
The new way to build is going to be through finishing in the middle of the league with an All-Star already on the roster waiting for support.
So how did we get here? Why does the reform have this big of an impact? The key is that the reform doesn’t just do one thing. It does many things. And the largest impact is not on the odds for the top pick you can get. It’s what happens to the worst pick you can get and how many teams can get each pick. The reform:
Evens out the oddsGives odds priority to the middle over the bottomMoves the guaranteed pick floor from five to 12 for the bottomDraws every pick, giving access to picks 5-12 to the playoff bubble and play-in teamsExpands the team pool by two
The Odds Advantage Lies in the Middle
The full odds breakdown based on currently available information can be viewed in the heat maps at the end of this article. Let’s jump straight to looking at the impact of these changes in a slightly aggregated form.
Here’s the average draft pick produced by a team’s finish position under both systems, put side by side.
The plot above shows the expected pick a team finishing in a given position should get under the old and new system. This is done using a weighted average based on every pick they might get under the lottery odds. This view makes it clear how severe the odds smoothing is. The worst teams expected pick goes way up, and the playoff bubble teams expected pick goes way down.
Under the old system, the average outcome for the worst-place finisher is the 3.7 pick. Under the new, it is the 8.1 pick. But this aggregation hides the real story. The story of what is happening to the worst pick a team can get.
The plot above only shows the odds for only the worst-place finishing team. These are the cumulative odds at getting the draft pick listed on the x-axis or better. You can see the floor of the pick outcome clearly. The worst team has a 100% chance of getting the fifth pick or better. That floor is what teams were really buying when trading for a pick from a team they project to be bad. That floor is the “incredibly valuable pick.” The new system kills that floor. In its place is a stairstep function up to the 12th pick. The very concept of the “incredibly valuable pick” is gone.
What about the team that finished 10th worst?
The opposite effect. They used to have small odds at picks one through four and were highly likely to get pick 10, 11, or 12. Eighty percent likely, in fact. This is also replaced by a stairstep function.
The value of picks is effectively compressed under the new system. Picks from any finish position are far more equal. This has a double impact on the trade value of picks. The idea of a “super valuable” pick is gone. But any pick is now “more valuable.” GMs will need to figure out this new world.
Any pick has upside. No pick has a guarantee.
The story gets even more complicated when you consider the non-linear return for a high draft pick. In order to assign “value” to each draft pick, data on the number of All-Stars produced by that pick slot will be used. While this is not perfect, it is a good enough indicator to assign a “hit rate” value to a draft pick. Getting an All-Star is a “hit.”
To reduce noise from small-sample outcomes at individual draft slots, a monotone smoothing is applied to the published All-Star rates. The method preserves the overall ordering of draft value while making the smallest possible adjustment to the underlying data. This prevents the numbers from overly fitting historical GM performance. Intuitively, we know earlier draft picks should not have lower expected value than later draft picks. In effect, just because lots of GMs picking at eight in our 30-year sample picked the wrong player does not mean pick nine is superior to pick eight.
You can quibble with the exact methodology, but many studies over many years show a roughly similar shape. It is good enough for our purposes of examining the trends to come rather than projecting exact outcomes. Now that we have picked value numbers, let’s apply them to our odds of getting a certain pick based on team finish position.
This now shows the odds that a finish position results in acquiring an All-Star-level talent. As expected, the old system concentrated All-Star outcomes on the worst finishing teams. The new system smooths this out evenly.
The size of this shift in the distribution of new All-Star talent throughout the league stands out even more when look at the difference between the two systems.
Now let’s get back to our grouping of teams from before.
GroupTeam Finish PositionDescriptionBad Teams1-3The worst of the worstNon-Playoff Teams4-7Teams that cannot truly competePlayoff Bubble8-10Teams on the outside looking in of the play-inPlay-in Teams11-16Play-in tournament teams
The plot above examines where new talent is expected to land in the league under the new system. In this view, the number of teams in each grouping does matter (and is not even per group, which influences the graph). The question is, “If a single All-Star is in the draft, which category of team does he get drafted by?” In this light, each finish position effectively flips a weighted coin for its chance to acquire that All-Star. So the “play-in teams” group gets six flips because there are six play-in teams. The “bad teams” group gets three flips because it is the three teams at the bottom of the standings. The important part is that both systems are evaluated using the same groupings.
This plot is devastating. This is the one that shows you that the worst teams are about to be in a dire situation. It’s the reason why people in NBA circles are calling the bottom three the “relegation zone”.
Teams used to think that it was easy to climb out of the bottom three. That’s because you had a draft capital advantage in the bottom three AND teams were actively trying to get down there to get that advantage. Now, not only are the teams that are better than you improving faster than you, those that are much better than you have the highest odds of landing new All-Stars in aggregate. The middle now has a draft capital advantage over the worst teams for both trading and drafting, and other teams will no longer attempt to reach the bottom.
This is the true reason why it is so risky to trade Mobley. A full reset should now be avoided at all costs. The play-in purgatory is gone. Those teams have true upward mobility now. The new purgatory is at the bottom. And when you’re at the bottom, it’s not even purgatory anymore. It’s hell.
Team Building is Harder Now
But this is not even the whole story. We’re not to those “building an All-Star core” tables yet. The odds didn’t just shift towards the better teams. The odds were diffused in general by allowing more teams a chance at more pick positions. The impact of this is less obvious without running the numbers, but just as severe: It makes it less probable than ever before to assemble a young core of talent of similar ages through the draft.
For the sake of aggregating results, let’s assume a team finishes in the same grouping position. For example, a team finishes in the “bad teams” group or the “play-in” group many years in-a-row. How many All-Stars do they acquire? For this question, the number of teams in a grouping does not matter. The odds of acquiring an all-star for each group is calculated using a weighted average of the odds for every finish position in the grouping.
These plots make it clear how severe the increase in variance is. Under the old system, if you were bad three years in a row, your average outcome was to draft at least one All-Star. Bad for five years? Good odds at two All-Stars.
Under the new system, no grouping expects to draft an All-Star in a three-year window. Nobody is even close to expecting to draft two All-Stars in five years. So how long does it take?
The problem is obvious. This is what leads to the tables at the beginning.
Every NBA fan knows that not all drafts are created equal and that imperfect talent evaluation leads to players getting drafted in the wrong order. All Cavs fans remember being terrible and picking 4-4-1-1 and getting Tristan Thompson, Dion Waiters, Anthony Bennett, and Andrew Wiggins (remember Cavs fans that the Kyrie Irving pick came from the Clippers).
This new lottery reform takes a system that was already high variance and turns the randomness up to 11. If you can’t reliably draft two All-Stars in five years, then you do not know what to do when a rookie contract is up. You don’t know when to stop rebuilding.
The Risk of Trading Evan Mobley
The league and its 3-2-1 reform have fundamentally changed the incentive structure of team building. Under the previous system, bottoming out provided a substantial pathway back to relevance. Under the new system, much of that advantage has been redistributed upward. The safest place to rebuild is no longer the bottom of the standings. It’s the middle. Voluntarily leaving the middle of the league and entering the bottom carries more risk than ever before. It is actively doing what the math underneath the 3-2-1 Lottery Reform explicitly tells you not to do.
The days of aligning a young core of stars through the draft are mostly over without very good fortune. Yet the teams built that way still exist in many places in the league. Catching up from the bottom will be very difficult as new talent is distributed amongst all teams finishing one through 16 rather than consolidated to those finishing one through five.
Keeping Mobley, competing with him to win now, and then building back around him from the middle after James Harden and Donovan Mitchell age out is the safest choice. It’s in line with the behavior the new system incentivizes. This is why the Cavaliers just sat out the Giannis Antetokounmpo sweepstakes. If they had traded Mobley for Giannis, it would have created a very real scenario where the team could bottom out completely in three years. And recovering from the true bottom, as we have just shown, is harder than ever before. It might even be impossible without very good luck. Drafting is a zero-sum game. Teams in the middle now have both the lead and the probabilistic advantage.
So celebrate Evan Mobley, Cavs fans! Our homegrown All-Star. Our hope. Past, present, and future.
Bonus – Full Odds Breakdown
In the old system, if you were the worst team for three years in a row then you were guaranteed top five picks in three consecutive years. The new system hard bans picking in the top five three years in-a-row. Even ignoring the rules on consecutive picks, the damage to the pick floor is severe. The reason is that all teams now get a chance at every pick. The guaranteed pick floor for the worst three-team finishes has moved from 5/6/7 out to 12 for all three.
This diffusion of odds is very visible in the difference between the two heat maps. Where is the redistribution really coming from? It’s happening because the pick the team finish position was most likely to get. For the worst team, they lost 41.9 percentage points from the odds at the fifth pick which was their former floor. All those points go into chances at lower picks. For the 12th place finisher, they lost 84.1 percentage points from their odds at the 12th pick. The majority of those points go into chances at a higher pick.