In my post on Friday about the DiMaggio hit streak, I mentioned the old hack-a-Shaq strategy. With Dwight Howard playing for Shaq’s old team, I’m surprised that teams don’t try the hack-a-Howard (Howard is a 60% career free throw shooter). Friday night’s Bulls-Magic game was a potential opportunity; the Bulls led for the entire fourth quarter. Here are the Bulls possessions inside of 4:30 remaining:
4:26 – 4:18: Luol Deng misses 12-foot jumper, Luol Deng offensive rebound, J.J. Redick foul, Luol Deng makes 1 free throw. 86-77
3:53 – 3:35: Carlos Boozer misses 17-foot jumper. 86-79
3:26 – 3:15: Derrick Rose misses 10-foot jumper, Derrick Rose offensive rebound, Jameer Nelson foul, Derrick Rose makes 2 free throws. 88-80
2:55 – 2:42: Hedo Turkoglu foul, Carlos Boozer makes 2 free throws. 90-80
2:27 – 2:13: Carlos Boozer makes 16-foot jumper. 92-80
2:00 – 1:39: Carlos Boozer misses 17-foot jumper. 92-80
1:23 – 1:09: Kyle Korver makes 27-foot three point jumper. 95-80
1:00 – 0:46: Taj Gibson makes jumper. 97-80
0:33 – 0:00: Carlos Boozer misses 9-foot jumper, Chicago offensive team rebound. 97-83
They fouled Howard a couple of times when he got the ball during this stretch, but never off the ball. You can’t foul away from the ball inside of 2 minutes remaining (the offense retains possession if you do), but hacking Howard at 3:35, 3:15, or 2:42 seems reasonable.
Additionally, the Bulls possessions only lasted for 8, 18, 11, 13, 14, 21, 14, 14, and 33 seconds! I wrote about time management in the NBA last week and concluded that, with a big lead, it’s worth looking for a good shot rather than running the shot clock to zero for much of the fourth quarter. However, I figured that looking for a good shot would take around 16 seconds. The Bulls only used 14 seconds on average, even very late in the game.
Starting with the possession at 1:23 (Kyle Korver’s three pointer), let’s work backwards through the Bulls possessions and consider the hack-a-Howard and clock killing strategies with some simulations. I’ve improved on my simulations from last time by assuming that teams will foul to stop the clock near the end of the game. Here are the parameters:
- When either team is behind, they use 10 seconds per possession and make a basket 50% of the time
- When Magic are ahead, they use 18 seconds per possession and make a basket 50% of the time
- When Bulls are ahead, three options:
- Use 14 seconds, 50% chance of made shot
- Use 20 seconds, 45% chance of made shot
- Use all 24 seconds, 35% chance of made shot
- If Bulls hack-a-Howard, Magic use 8 seconds per possession and Bulls play strategy #1.
- 30% of all made shots are three pointers
- Between 30 seconds and 1 minute remaining, teams foul after 2 seconds if behind by 4 to 8 points
- Under 30 seconds, teams foul if behind by 6 points or less
- Both teams make 70% of their free throws in general
- Howard makes 60% of his free throws
- Probabilities come from 1,000 simulations
And now the results:
As in my last post on clock management, the short story is that the Bulls were almost sure to win no matter what. I believe these results even more than last time, since I’ve added late game fouling (this overrides offensive strategy choice when the finish is close and therefore reduces differences between strategies). Playing a medium speed strategy (20 seconds per possession) has some value between 2 and 4 minutes out, but any gains are small. Killing clock is not so important, at least in this game situation.
What about hacking Howard? That strategy isn’t very impressive either. However, if I decrease Howard’s free throw shooting percentage to 50%, the hack-a-Howard percentages are 98.4, 97.4, 92.6, 95.7, and 95.3%. This is almost uniformly better than playing normal defense. The numbers would be even higher if the Bulls were able to burn more than 8 seconds before hacking.
Why does hacking work? The average points per trip when hacking is 0.5*2 = 1 in this case, while the average points per trip when not hacking is 0.5*0.7*2 + 0.5*0.3*3 = 1.15, and hacking eliminates the risk that the Magic hit a few three pointers in a row (i.e., hacking has a lower variance). Are there other pitfalls to hacking? Of course. Most notably, the coach gets the blame following a loss by hacking — this is surely why more coaches don’t order the hack. It may undermine players’ confidence on defense as well. Still, for a very poor free throw shooter, there are times when it’s a good idea.