Reforming The NBA Draft Lottery

Frequently, when individuals make proposals on reforming the NBA draft lottery, they get too caught up on solutions rather than diagnosing the problems.

In order to determine what the ideal alternative should be, it is important to understand what the problems are with the current system and what features need to be highlighted in a new system. In order to do this, we must understand what the NBA draft lottery is trying to accomplish on a conceptual basis.

an ideal system would lead to the following:

– it diminishes the incentivize for losing
– it allows teams to build through the draft
– it does not penalize teams for being bad; so they can play their youth and grow as a team (aka it does not incentivize winning)
– it prevents opportunistic one year tanking
– it allows for developing teams to not peak too early
– it prevents teams from perpetually tanking or benefitting too much from the draft
– it must be equitable
– it must be simple enough to follow along

Basketball fans know that there are two types of tanking. Sure, some teams are so bad and committed that they start losing in November, but for the majority of team’s in the league, tanking is a delayed reaction. Most non-playoff team’s try at the beginning of the year, but once playoffs are clearly out of reach, then losing becomes a set strategy.

To mitigate the problems of this, I would set the deadline for determining the lottery order at the trade deadline*. This allows for the remaining part of the season to be played without any motivation or incentives for losing. By taking away the benefit of losing, the weaker teams in the league would have a more competitive end of season. While this does not completely mitigate the benefits of losing, it helps. While the very worst teams in the league will still be bad from day one regardless, it prevents the teams that originally set out to make the playoffs to be able to tank. Additionally, it allows the truly bad teams to have at least a 30 game period where management would actually prefer the team to win.

There is a big difference between a franchise that is rebuilding and one that is bad for just one year. It is important to aid the franchises that have been bad and are truly rebuilding. At the same time, the league does not want to be rewarding the teams that have had a history of success and are merely bad for one year, hoping to land an additional superstar. This is why its important to view the lottery not as a one season formula, but reflecting a period of time
In regards to this, I propose to take into account not only the standings of the current seasons, but also the previous two seasons.
I would weigh the current season 3X, the previous season 2X, and two seasons ago as being worth X.

This prevents otherwise good teams or teams in good situations from tanking for one year in order to get a superstar. It also accomplishes a more important goal. It allows developing teams to continue their growth. This will allow teams to “go for it”, and start adding talent to their roster without shutting the door of getting more talented young players.

Presently, the NBA lottery does not actual reward teams for how bad they are, but merely how bad they are in comparison to the rest of the league. This leads to a few quixotic results. Firstly, the draft order is skewed because some weak teams that make the playoffs receive worse picks than far better teams that miss the playoffs. Further, teams are encouraged to out tank other teams when there are several teams clustered around certain picks. If 5 lottery teams are tied in the standings, they not only want to lose, but they want to lose more than the 4 other teams they are competing with.

To help ensure a more equitable result, instead of basing the lottery odds on standings, it is preferable to do so by total losses.

lets use the New York Knicks as an example. At the 2015 trade deadline, they had 43 losses. In 2014, 33 and in 2013, 18.
129 + 99 + 18 = 246 points.
All of the teams in the league would then be ranked by their total amount of losses/points under this formula to determine the draft order.

Additionally, because we don’t want teams to luck into a dynasty through the draft, nor do we want to put teams in a position where they want to stay perpetually bad, teams would be forbidden to receive multiple number number one draft picks in any three year cycle.

Moreover, as a measure to further reduce a team’s desire to lose in order to receive a top pick in the draft (and because fans really enjoy the lottery process), I would keep a lottery for only the number one pick in the draft. I don’t think a lottery is still necessary for the entire selection process because of all of the stop-gaps put in place to mitigate teams from tanking, but I think its beneficial because of how coveted the number one pick is due to its ability to land a superstar.

I’m not sure what the best set of odds would be – if people would want to use the current NBA lottery odds, the NHL lottery odds, or a new set of numbers.

Lastly, because we are reformatting the lottery, we might as well reformat the way the actual lottery is done.

The current lottery system is very frustrating because not only is it an incredibly boring show, but it still leaves a large amount of NBA fans thinking there is some sort of conspiracy going on.

My proposed solution would be to have a live lottery system where there are 1000 balls** in a balls in a hopper. Each team gets to include the amount of balls corresponding with what their odds of winning are. Then, a point system would be created where in order to “win” the lottery, it would be the first team to have their ball picked 10*** times.
Like rapid fire, balls could be drawn and teams would be racing to get to 10 points first. This would provide for an entertaining broadcast, while also ensuring fans trust the process.

*because teams do not all have the same amount of games played by the trade deadline, it should be set to playing something like 53 games, instead of by the actual date.
** this number actually depends on the odds put in place to make the math work – it might be something like 1003 balls in actuality.
*** once again, it depends on what the odds used are, but however many times is necessary to have an appropriate sample.