Showing posts with label march madness. Show all posts
Showing posts with label march madness. Show all posts

Wednesday, June 25, 2014

Fixing the NBA Draft (Revisited)

On May 20th, almost three million people tuned in to watch the NBA Draft Lottery. That's right -- more people watched this year's NBA Draft Lottery than have attended Miami Heat games since they signed LeBron James.

The draft lottery consists of twelve ping-pong balls getting pulled from a spinning plastic drum. But the really pathetic thing is that we don't even get to see the balls bounce around. The ping-pong balls are drawn off-camera, and then the results are put in an envelope. And then three million Americans turn on their TV to watch the envelopes get opened.

It's like bingo, but worse. It's like waiting for your grandmother to go play bingo. Then, when she gets home, asking her what happened. As an American, that makes me sad. We have nothing better to do than watch a TV show that reveals, second-hand, the results of "Bingo For Billionaires".

The good news is that we can fix it. We don't have to allocate draft picks like a church pastor calling out bingo numbers. As we first proposed back in 2007, we can replace the NBA Draft Lottery with the NBA Draft Tournament.

The NBA Draft Tournament

Instead of putting all 14 non-playoff teams into a big bucket and playing bingo, we put all 14 teams into a single-elimination tournament bracket. All the excitement of March Madness, but with the future of your favorite NBA team resting in the balance.

Here's how it works:

1) Divide the 14 non-playoff teams into two brackets containing 7 teams each (one bracket for each conference).

2) In each bracket, the teams with the most regular-season wins play each other in a 1-game playoff. The loser goes home and the winner advances to play the team with the next best regular-season record.

3) Continue until you have one winner from each conference. These two teams play head-to-head in a "Championship Game" for the #1 pick.

4) Award the remaining picks according to how far each team advanced in the NBA Draft Tournament.

For example, these are what the brackets would have been for the 2014 NBA Draft Tournament:

Western Conference
Eastern Conference
Game #1
Phoenix Suns (48-34)
at
Minnesota Timberwolves (40-42)
Game #2
New York Knicks (37-45)
at
Cleveland Cavaliers (33-49)
Game #3
Winner of Game #1 (above)
at
  Denver Nuggets (36-46)
Game #4
Winner of Game #2 (above)
at
Detroit Pistons (29-53)
Game #5
Winner of Game #3 (above)
at
 New Orleans Pelicans (33-49)
Game #6
Winner of Game #4 (above)
at
Boston Celtics (25-57)
Game #7
Winner of Game #5 (above)
at
 Sacramento Kings (28-54)
Game #8
Winner of Game #6 (above)
at
Orlando Magic (23-59)
Game #9
Winner of Game #7 (above)
at
Los Angeles Lakers (27-55)
Game #10
Winner of Game #8 (above)
at
Philadelphia 76ers (19-63)
Game #11
Winner of Game #9 (above)
at
 Utah Jazz (25-57)
Game #12
Winner of Game #10 (above)
at
Milwaukee Bucks (15-67)
Championship Game
Winner of Game #11
vs.
Winner of Game #12

But Is It "Fair"?

This tournament model has the advantage of maintaining the "integrity" of current system. In other words, the worst teams still have the best chance to earn the #1 pick. But they actually have to earn it -- on the basketball court.

This table shows the chance of each team getting the #1 pick using this tournament format, compared to the chance currently given to them by the NBA in the lottery:

(Unlike our previous analysis, we are using each team's regular-season winning percentage to determine their strength in the tournament.)

Lottery Tournament Western Conference Eastern Conference Tournament Lottery
0.5% 0.8%Suns (48-34) Knicks (37-45) 0.9% 0.7%
0.6% 1.1%Timberwolves (40-42) Cavaliers (33-49) 1.4% 1.7%
0.8% 1.7%  Nuggets (36-46) Pistons (29-53) 2.2% 2.8%
1.1% 2.9% Pelicans (33-49) Celtics (25-57) 4.2% 10.3%
4.3% 5.4%Kings (28-54) Magic (23-59) 7.9% 15.6%
6.3% 12.7%Lakers (27-55) 76ers (19-63) 15.6% 19.9%
10.4% 19.3%Jazz (25-57) Bucks (15-67) 22.4% 25.0%

The End of Tanking?

Because the tournament requires that you win at least two consecutive games in order to win the #1 pick, you still have to put a decent team on the court. A bad record gives you a better position in the tournament, but you still need a team good enough to win.

For example, the Bucks never won two consecutive games in the entire season. Yet they would be required to go two-for-two in the NBA Draft Tournament to secure the #1 pick.

Wednesday, May 23, 2007

Fixing the NBA Draft

I watched the NBA Draft Lottery last night.

I didn't watch the NBA Draft. I watched the NBA Draft Lottery.

The Lottery could be more exciting than it is. They start with twelve ping-pong balls in a spinning plastic drum. But we don't get to see the balls bounce around. The balls are drawn off-camera, and then the results are put in an envelope. And then millions of Americans turn on their TV to watch the envelopes get opened. Live! :)

It's like bingo, but worse. It's like waiting for your grandmother to go play bingo. Then, when she gets home, asking her what happened. As an American, that makes me sad. We have nothing better to do than watch a TV show that reveals -- second-hand -- the results of "Bingo For Billionaires".

The good news is that we can fix it. We don't have to allocate draft picks like a church pastor calling out bingo numbers. We can replace the NBA Draft Lottery with an NBA Draft Tournament.

The NBA Draft Tournament

Instead of putting all 14 non-playoff teams into a big bucket and playing bingo, we put all 14 teams into a single-elimination tournament bracket. We get all the excitement of the NCAA Tournament (aka "March Madness") but with the your team's top draft pick resting in the balance.

Here's how it works:

1) Divide the 14 non-playoff teams into two brackets containing 7 teams each (one bracket for each conference).

2) In each bracket, the teams with the most regular-season wins play each other in a 1-game playoff. The loser goes home and the winner advances to play the next team (ranked by regular-season wins).

3) Continue until you have one winner from each conference. These two teams play head-to-head in a "Championship Game" for the #1 pick.

4) Award the remaining picks according to how far each team advanced in the NBA Draft Tournament.

For example, these are what the brackets would have been for the 2007 NBA Draft Tournament:

Western Conference
Eastern Conference
Game #1
Los Angeles Clippers (48-34)
at
Golden State Warriors (40-42)
Game #2
Indiana Pacers (35-47)
at
Philadelphia 76ers (35-47)
Game #3
Winner of Game #1 (above)
at
Sacramento Kings (33-49)
Game #4
Winner of Game #2 (above)
at
Carlotte Bobcats (33-49)
Game #5
Winner of Game #3 (above)
at
Portland Trail Blazers (32-50)
Game #6
Winner of Game #4 (above)
at
New York Knicks (33-49)
Game #7
Winner of Game #5 (above)
at
Minnesota Timberwolves  (32-50)
Game #8
Winner of Game #6 (above)
at
Atlanta Hawks (30-52)
Game #9
Winner of Game #7 (above)
at
Seattle SuperSonics (31-51)
Game #10
Winner of Game #8 (above)
at
Milwaukee Bucks (28-54)
Game #11
Winner of Game #9 (above)
at
Memphis Grizzlies (22-60)
Game #12
Winner of Game #10 (above)
at
Boston Celtics (24-58)
Championship Game
Winner of Game #11
vs.
Winner of Game #12

But Is It "Fair"?

This tournament model has the advantage of maintaining the "integrity" of current system. In other words, the worst teams still have the best chance to earn the #1 pick. But they actually have to earn it -- on the basketball court.

The team with the worst record (the Grizzlies) has to win 2 games in a row to win the tournament. If each game is a coin flip. this gives them a 1-in-4 chance. This is exactly equal to the 25% chance they have with the current ping-pong ball system.

The team with the 3rd-worst record (the Bucks) would have to win 3 games in a row to win the tournament. This is about a 1-in-8 chance (12.5%). Their chance under the current system: 15.6%.

At the other end, we have the Pacers, with the best record in the Eastern Conference among those hat missed the playoffs. Indiana would to win 7 games in a row. Assuming each contest is evenly matched, that gives them a 1-in-128 chance (about 0.8%). This is just a bit better than their odds under the current lottery system (0.7%).

Scheduling

Another problem is that the NBA Playoffs get more boring with each round. Well, more accurately, playoff games get more sparse with each round. The gap between the Conference Finals and the NBA Finals is about 4 days. And then we stretch one series over two weeks in June. This year we ended the year with 15 games spread over 26 days.

We fix this by scheduling the Draft Tournament during the scheduled off-days, with two games each night, leading up to a "final four" weekend:

  • June 3rd (Sunday): Game #1 and Game #2
  • June 5th (Tuesday): Game #3 and Game #4
  • June 8th (Friday): Game #5 and Game #6
  • June 11th (Monday): Game #7 and Game #8
  • June 13th (Wednesday): Game #9 and Game #10
  • June 16th (Saturday): Game #11 and Game #12 (the semi-finals)
  • June 18th (Monday): Championship Game