and when this shit burns us in the Playoffs this sub will act suprised.
This team is way too talented for its only offensive gameplan to be “Chuck 3s” but the Mazzulla fanboys don’t want to hear it.
Velynedra
Hello everyone! I had a little bit of time this weekend so I decided to run this analysis on Celtics’ shot selection. The question is whether we lose games due to shooting too many inefficient 3s or else.
**The idea.**
The idea is to separate out the 2s and 3s of each game and see how the running points-per-possession (PPPs) of each shot selection progress over the course of the game. The PPPs of 2s and 3s will be compared to each other, as well as to the PPP threshold of the game (i.e., PPP needed to beat the opponent by 1 point). I plotted such running-PPP analyses of the last 10 games (20231117-20231208), 7W-3L. The first quarter data are skipped because they are variable (not enough shots attempted to reliably estimate PPP). The data are generated by parsing the Bball ref play-by-play page (e.g., https://www.basketball-reference.com/boxscores/pbp/202312080BOS.html).
**The plots.**
Upper panel.
The blue line shows the PPP progression for 2s, with closed and open circles denoting made* and missed shots, respectively. The red line shows the same thing for 3s. The Orange dashed line shows the PPP threshold to win this game. Vertical dashed lines separate quarters. Quarters and timeouts are labeled on the top row.
*If a player makes 1/2 free throws, the circle is filled for plotting purposes but only 1 point is counted towards PPP calculations.
Lower panel.
The thick gray line shows the true PPP progression of the game (2s+3s+turnovers). Blue, red, and black dots denote 2s, 3s, and turnover events. The black zigzig dashed line shows PPPs without turnovers, which tells us how efficient our shot selections are on their own. The Orange dashed line shows the PPP threshold to win this game. Vertical dashed lines separate quarters. Quarters and timeouts are labeled on the top row.
Because the possessions are parsed from the play-by-play data, the numbers might be a bit off from official sources.
**The impression.**
My impression based on the last 10 games is that, no, we mostly did **not** lose games due to shooting too many 3s. But, our 3s are generally less efficient than our 2s. Here are some game-by-game notes.
1. Nov 17, 108-105 W vs Raptors. Both 2s and 3s were above the win threshold and fairly equal in efficiency. Our TOVs pulled down our PPP by ~0.1, which almost lost us the game. 2. Nov 19, 102-100 W vs Grizzlies. Both 2s and 3s were similar in efficiency and above threshold. The 17 TOVs pulled the PPPs down by 0.2, almost lost us the game again. 3. Nov 20, 118-121 L vs Hornets. 2s were 0.2 PPP more efficient by 3s (32%). Both PPPs took a dive after the half time probably because our guys were tired in a B2B. The 13 turnovers pulled us below the win threshold. 4. Nov 22, 119-116 W vs Bucks. Both 2s and 3s were similarly above threshold. We lost a lot of 3 efficiencies after the end of Q2, but it didn’t end too badly (~1.3 PPP). The turnover fest from the second half of Q3 to the start of Q4 almost lost us the game. 5. Nov 24, 96-113 L vs Magic. Our 3s were horrible, below 0.8 PPP. That just lost us the game. 2s were massively more efficient than 3s (~0.5 PPP diff). We couldn’t hit anything on the top of Q3. Crazy. 6. Nov 26, 113-103 W vs Hawks. Our 2s were massively more efficient than 3s again, but our 3s weren’t that much below the threshold, merely 0.05 PPP. The 2s won us the game. 7. Nov 28, 124-97 W vs Bulls. Both 2s and 3s were equally efficient (~1.4 PPP) and above threshold. We can’t lose when we shoot like this. 8. Dec 1, 125-119 W vs Sixers. Our 3s were 0.2 PPP more efficient than 2s, which is good to see. Both shot selections were above threshold and the 20 turnovers almost lost us the game. 9. Dec 4, 112-122 L vs Pacers. Our 2s were a lot more efficient than 3s (~0.4 PPP diff). Oh well. 10. Dec 8, 133-123 W vs Knicks. Our 2s and 3s were roughly similar at the beginning of Q4. After that, 3s took a nose dive but not enough to lose us the game.
Anyway, feel free to draw your own conclusions and, because these plots are pretty easy to make once the code is set up, I’m happy to make more after each game.
Let’s go Cs!
verossiraptors
Your 2s being more efficient than your 3s is not a good place to be in as an offense
3 Comments
and when this shit burns us in the Playoffs this sub will act suprised.
This team is way too talented for its only offensive gameplan to be “Chuck 3s” but the Mazzulla fanboys don’t want to hear it.
Hello everyone! I had a little bit of time this weekend so I decided to run this analysis on Celtics’ shot selection. The question is whether we lose games due to shooting too many inefficient 3s or else.
**The idea.**
The idea is to separate out the 2s and 3s of each game and see how the running points-per-possession (PPPs) of each shot selection progress over the course of the game. The PPPs of 2s and 3s will be compared to each other, as well as to the PPP threshold of the game (i.e., PPP needed to beat the opponent by 1 point). I plotted such running-PPP analyses of the last 10 games (20231117-20231208), 7W-3L. The first quarter data are skipped because they are variable (not enough shots attempted to reliably estimate PPP). The data are generated by parsing the Bball ref play-by-play page (e.g., https://www.basketball-reference.com/boxscores/pbp/202312080BOS.html).
**The plots.**
Upper panel.
The blue line shows the PPP progression for 2s, with closed and open circles denoting made* and missed shots, respectively. The red line shows the same thing for 3s. The Orange dashed line shows the PPP threshold to win this game. Vertical dashed lines separate quarters. Quarters and timeouts are labeled on the top row.
*If a player makes 1/2 free throws, the circle is filled for plotting purposes but only 1 point is counted towards PPP calculations.
Lower panel.
The thick gray line shows the true PPP progression of the game (2s+3s+turnovers). Blue, red, and black dots denote 2s, 3s, and turnover events. The black zigzig dashed line shows PPPs without turnovers, which tells us how efficient our shot selections are on their own. The Orange dashed line shows the PPP threshold to win this game. Vertical dashed lines separate quarters. Quarters and timeouts are labeled on the top row.
Because the possessions are parsed from the play-by-play data, the numbers might be a bit off from official sources.
**The impression.**
My impression based on the last 10 games is that, no, we mostly did **not** lose games due to shooting too many 3s. But, our 3s are generally less efficient than our 2s. Here are some game-by-game notes.
1. Nov 17, 108-105 W vs Raptors. Both 2s and 3s were above the win threshold and fairly equal in efficiency. Our TOVs pulled down our PPP by ~0.1, which almost lost us the game.
2. Nov 19, 102-100 W vs Grizzlies. Both 2s and 3s were similar in efficiency and above threshold. The 17 TOVs pulled the PPPs down by 0.2, almost lost us the game again.
3. Nov 20, 118-121 L vs Hornets. 2s were 0.2 PPP more efficient by 3s (32%). Both PPPs took a dive after the half time probably because our guys were tired in a B2B. The 13 turnovers pulled us below the win threshold.
4. Nov 22, 119-116 W vs Bucks. Both 2s and 3s were similarly above threshold. We lost a lot of 3 efficiencies after the end of Q2, but it didn’t end too badly (~1.3 PPP). The turnover fest from the second half of Q3 to the start of Q4 almost lost us the game.
5. Nov 24, 96-113 L vs Magic. Our 3s were horrible, below 0.8 PPP. That just lost us the game. 2s were massively more efficient than 3s (~0.5 PPP diff). We couldn’t hit anything on the top of Q3. Crazy.
6. Nov 26, 113-103 W vs Hawks. Our 2s were massively more efficient than 3s again, but our 3s weren’t that much below the threshold, merely 0.05 PPP. The 2s won us the game.
7. Nov 28, 124-97 W vs Bulls. Both 2s and 3s were equally efficient (~1.4 PPP) and above threshold. We can’t lose when we shoot like this.
8. Dec 1, 125-119 W vs Sixers. Our 3s were 0.2 PPP more efficient than 2s, which is good to see. Both shot selections were above threshold and the 20 turnovers almost lost us the game.
9. Dec 4, 112-122 L vs Pacers. Our 2s were a lot more efficient than 3s (~0.4 PPP diff). Oh well.
10. Dec 8, 133-123 W vs Knicks. Our 2s and 3s were roughly similar at the beginning of Q4. After that, 3s took a nose dive but not enough to lose us the game.
Anyway, feel free to draw your own conclusions and, because these plots are pretty easy to make once the code is set up, I’m happy to make more after each game.
Let’s go Cs!
Your 2s being more efficient than your 3s is not a good place to be in as an offense