josh smith to boston? (rumor?)

[Fafnir]

If you're going to be nerdy Jake, you should know that using Point Per Shot is the wrong measure to judge "hack-a" strategies relative efficiency. Point Per shot ignores turnovers and other free throw attempts, which are the other possibilities beyond a make or a miss.

Given you're supposing Rondo/Smith being wrapped up at half-court then by conceit you're ignoring the chance of a turnover or foul to another player during the hack.

Meanwhile comparatively playing traditional defense you want to look at points per possession because you can still get a turnover or foul a free throw shooter.

Portland scores at 1.12 points per possession. A composite Rondo/Smith hack strategy will yield 1.08 points per possession using your 52% figure.

So a 52% make rate renders an output of a top 8 offensive team. A mere 60% free throw make rate yields a 1.2 per possession, which would lead the league.

Now where a "hack-a" strategy can be useful is three-fold.

1. It can be used to get into a players head.
2. It can be used to rest your team from running up and down the floor.
3. It can make sense if you're already down and you wish to gamble because your defense is proving to be a sieve. An all-in risky strategy.

No, it's the proper metric. As I understand the debate, the question is whether the Celtics would ultimately lose more game as a result of poor free throw shooters being fouled more often.

The objection to this line of reasoning is that compounding all free throws, and then looking at the true percentage, ACTUALLY, a free throw is a better percentage bet than letting the player take the shot.

Since the debate centers around: shooting v. free throw shooting, a metric which ignores free throw shooting is PRECISELY the appropriate metric to use. My argument (or at least the part of my blog post subject to debate) is whether teams would more likely foul our bad free throw shooters than allow them to shoot. It IS more likely, but only because Josh Smith is such an outrageously bad free throw shooter. If Smith were even as good as Rondo then there would be some question about the efficacy of this strategy - but, unfortunately, his free throw shooting is so bad, that you would rather be Milwuakee than have a Rondo/Smith Combo shooting your free throws.

Your framing makes zero sense and completely departs from what Bballtim initially was talking about and actual basketball. Break down the problem at its core when is it advantageous to foul Smith/Rondo types?

Fouling is a worse outcome than any shot Rondo/Smith take except if its at the rim. This is true even with their poor free throw shooting! (using Rondo's shooting from last year for sample size)

PPS for 3s for both of them: .69 for Josh, .72 for Rondo
PPS for 2 point jump shots: .63, .87
PPS at the rim:1.428, 1.196

Such a team wouldn't be fouled into oblivion, instead the paint would be packed and they'd attempt to force jump shots. Fouls would happen, but mostly on penetration into the paint to prevent shots at the rim. Just like with every other basketball player.

This is why coaches scream, don't foul jump shooters. Even if a guy makes 40% of his 3s, its better to not foul him if he even shoots 50% from the line.

By aggregating overall points per shot you're missing that crucial element. Beyond that, I still don't see what you're talking about in response to BBalltim's analysis of why "hack-a" strategies are ineffective.

If you're analyzing overall offense, you have to account for turnovers, offensive boards, and free throws.

[Fafnir]

Teams aren't even fouling Detroit all that often, they're only 19th in the league at drawing free throws. Only .201 free throws per field goal attempt.

And teams don't foul Houston to stifle their offense, they foul Houston a lot because Howard will dunk the ball a ton otherwise and James Harden gets a ton of calls. Even if Howard hit 75%+ of his free throws he'd still get fouled a ton.

[jaketwice]

If you're going to be nerdy Jake, you should know that using Point Per Shot is the wrong measure to judge "hack-a" strategies relative efficiency. Point Per shot ignores turnovers and other free throw attempts, which are the other possibilities beyond a make or a miss.

Given you're supposing Rondo/Smith being wrapped up at half-court then by conceit you're ignoring the chance of a turnover or foul to another player during the hack.

Meanwhile comparatively playing traditional defense you want to look at points per possession because you can still get a turnover or foul a free throw shooter.

Portland scores at 1.12 points per possession. A composite Rondo/Smith hack strategy will yield 1.08 points per possession using your 52% figure.

So a 52% make rate renders an output of a top 8 offensive team. A mere 60% free throw make rate yields a 1.2 per possession, which would lead the league.

Now where a "hack-a" strategy can be useful is three-fold.

1. It can be used to get into a players head.
2. It can be used to rest your team from running up and down the floor.
3. It can make sense if you're already down and you wish to gamble because your defense is proving to be a sieve. An all-in risky strategy.

No, it's the proper metric. As I understand the debate, the question is whether the Celtics would ultimately lose more game as a result of poor free throw shooters being fouled more often.

The objection to this line of reasoning is that compounding all free throws, and then looking at the true percentage, ACTUALLY, a free throw is a better percentage bet than letting the player take the shot.

Since the debate centers around: shooting v. free throw shooting, a metric which ignores free throw shooting is PRECISELY the appropriate metric to use. My argument (or at least the part of my blog post subject to debate) is whether teams would more likely foul our bad free throw shooters than allow them to shoot. It IS more likely, but only because Josh Smith is such an outrageously bad free throw shooter. If Smith were even as good as Rondo then there would be some question about the efficacy of this strategy - but, unfortunately, his free throw shooting is so bad, that you would rather be Milwuakee than have a Rondo/Smith Combo shooting your free throws.

Your framing makes zero sense and completely departs from what Bballtim initially was talking about and actual basketball. Break down the problem at its core when is it advantageous to foul Smith/Rondo types?

Fouling is a worse outcome than any shot Rondo/Smith take except if its at the rim. This is true even with their poor free throw shooting! (using Rondo's shooting from last year for sample size)

PPS for 3s for both of them: .69 for Josh, .72 for Rondo
PPS for 2 point jump shots: .63, .87
PPS at the rim:1.428, 1.196

Such a team wouldn't be fouled into oblivion, instead the paint would be packed and they'd attempt to force jump shots. Fouls would happen, but mostly on penetration into the paint to prevent shots at the rim. Just like with every other basketball player.

This is why coaches scream, don't foul jump shooters. Even if a guy makes 40% of his 3s, its better to not foul him if he even shoots 50% from the line.

By aggregating overall points per shot you're missing that crucial element. Beyond that, I still don't see what you're talking about in response to BBalltim's analysis of why "hack-a" strategies are ineffective.

If you're analyzing overall offense, you have to account for turnovers, offensive boards, and free throws.

If the question is why "hack-a" strategies are ineffective, the answer is, "because a team will get fewer points per free throw attempt than they will by taking a shot." Each free throw attempt (barring "and 1s") is worth .54 points, by Smith and Rondo, using their 2012-2013 numbers adjusted for the amounts made and attempted by both. A shot is worth 1.3 points, or something to Miami. Obviously, if a better free throw shooter is shooting, it's worth more. While turn-overs happen, they're beyond the scope of this analysis - as are offensive rebounds.

I think what you are ignoring is that "hack-a" strategies allow the opposing team to dictate what shot, precisely, is taken: a pair of free throws.

We are dealing in averages - so we can't really discuss whether the alternative would be a jump shot, or whatever - once you're in the penalty, maybe it would be a shot in the rim. You have to evaluate averages aganist averages. And the data in this instance is blind.

Using YOUR OWN MATH, the PPS from the FT line for Smith is .52, worse than any other alternative you cite above.

Perhaps my error was overstating the issues associated with pairing Smith and Rondo (although Rondo's shaky free throw shooting means that he can't reliably hold the ball to protect Smith).

The larger issue is just that Smith is a truly HORRIBLE free throw shooter. He is a liability at the line for Detroit, and would be a liability at the line for Boston.

Josh Smith to Boston? No thanks!

[Fafnir]

If you're going to be nerdy Jake, you should know that using Point Per Shot is the wrong measure to judge "hack-a" strategies relative efficiency. Point Per shot ignores turnovers and other free throw attempts, which are the other possibilities beyond a make or a miss.

Given you're supposing Rondo/Smith being wrapped up at half-court then by conceit you're ignoring the chance of a turnover or foul to another player during the hack.

Meanwhile comparatively playing traditional defense you want to look at points per possession because you can still get a turnover or foul a free throw shooter.

Portland scores at 1.12 points per possession. A composite Rondo/Smith hack strategy will yield 1.08 points per possession using your 52% figure.

So a 52% make rate renders an output of a top 8 offensive team. A mere 60% free throw make rate yields a 1.2 per possession, which would lead the league.

Now where a "hack-a" strategy can be useful is three-fold.

1. It can be used to get into a players head.
2. It can be used to rest your team from running up and down the floor.
3. It can make sense if you're already down and you wish to gamble because your defense is proving to be a sieve. An all-in risky strategy.

No, it's the proper metric. As I understand the debate, the question is whether the Celtics would ultimately lose more game as a result of poor free throw shooters being fouled more often.

The objection to this line of reasoning is that compounding all free throws, and then looking at the true percentage, ACTUALLY, a free throw is a better percentage bet than letting the player take the shot.

Since the debate centers around: shooting v. free throw shooting, a metric which ignores free throw shooting is PRECISELY the appropriate metric to use. My argument (or at least the part of my blog post subject to debate) is whether teams would more likely foul our bad free throw shooters than allow them to shoot. It IS more likely, but only because Josh Smith is such an outrageously bad free throw shooter. If Smith were even as good as Rondo then there would be some question about the efficacy of this strategy - but, unfortunately, his free throw shooting is so bad, that you would rather be Milwuakee than have a Rondo/Smith Combo shooting your free throws.

Your framing makes zero sense and completely departs from what Bballtim initially was talking about and actual basketball. Break down the problem at its core when is it advantageous to foul Smith/Rondo types?

Fouling is a worse outcome than any shot Rondo/Smith take except if its at the rim. This is true even with their poor free throw shooting! (using Rondo's shooting from last year for sample size)

PPS for 3s for both of them: .69 for Josh, .72 for Rondo
PPS for 2 point jump shots: .63, .87
PPS at the rim:1.428, 1.196

Such a team wouldn't be fouled into oblivion, instead the paint would be packed and they'd attempt to force jump shots. Fouls would happen, but mostly on penetration into the paint to prevent shots at the rim. Just like with every other basketball player.

This is why coaches scream, don't foul jump shooters. Even if a guy makes 40% of his 3s, its better to not foul him if he even shoots 50% from the line.

By aggregating overall points per shot you're missing that crucial element. Beyond that, I still don't see what you're talking about in response to BBalltim's analysis of why "hack-a" strategies are ineffective.

If you're analyzing overall offense, you have to account for turnovers, offensive boards, and free throws.

If the question is why "hack-a" strategies are ineffective, the answer is, "because a team will get fewer points per free throw attempt than they will by taking a shot." Each free throw attempt (barring "and 1s") is worth 1.08 points, by Smith and Rondo, using their 2012-2013 numbers. A shot is worth 1.3 points, or something. Obviously, if a better free throw shooter is shooting, it's worth more. While turn-overs happen, they're beyond the scope of this analysis - as are offensive rebounds.

This is wrong a shot isn't worth 1.3 points. Because points per shot includes all the points scored on the free throw line. Furthermore if you're comparing "regular offense" to "hack a Rondo/Smith" you must account for turnovers/fta/offensive boards that occur with regular offense.

Teams don't get a FGA every possession in the course of a normal offensive trip. You have to use points per possession instead of points per shot.

Points per possesion is where its at for advanced look at NBA offense.

I think what you are ignoring is that "hack-a" strategies allow the opposing team to dictate what shot, precisely, is taken: a pair of free throws.

We are dealing in averages - so we can't really discuss whether the alternative would be a jump shot, or whatever - once you're in the penalty, maybe it would be a shot in the rim. You have to evaluate averages aganist averages. And the data in this instance is blind.

Using YOUR OWN MATH, the PPS from the FT line for Smith is .52, worse than any other alternative you cite above.

Perhaps my error was overstating the issues associated with pairing Smith and Rondo (although Rondo's shaky free throw shooting means that he can't reliably hold the ball to protect Smith).

You are right that for the given FTAs you ignore turnovers because the terms are given, 2 FTAs. (as an aside 2 FTAs are more valuable than indicated as OReb do occur on second free throw misses but the effect is slight(

But on the other side of the comparison you can't ignore turnovers, offensive boards, and free throws.

Also its not .52 pps for 2 Josh Smith FTAs. 2 FTAs = 1 possesion = 1 shot. So that means its 1.04 PPP

[LooseCannon]

According to one estimate, the break-even point for a hack-a-shaq strategy to work is if the player's FT% is 46.41%.

This season, four regulars are below that mark: Andrew Bogut, Andre Drummond, DeAndre Jordan, and Gerald Wallace.  Bogut and Wallace have career averages that are significantly higher.  Drummond is still under 40% for his career.

[jaketwice]

Also its not .52 pps for 2 Josh Smith FTAs. 2 FTAs = 1 possesion = 1 shot. So that means its 1.04 PPP

No, this is quite wrong. Your metric is pps, above, and now you are changing to ppp.

If you're going to be nerdy Jake, you should know that using Point Per Shot is the wrong measure to judge "hack-a" strategies relative efficiency. Point Per shot ignores turnovers and other free throw attempts, which are the other possibilities beyond a make or a miss.

Given you're supposing Rondo/Smith being wrapped up at half-court then by conceit you're ignoring the chance of a turnover or foul to another player during the hack.

Meanwhile comparatively playing traditional defense you want to look at points per possession because you can still get a turnover or foul a free throw shooter.

Portland scores at 1.12 points per possession. A composite Rondo/Smith hack strategy will yield 1.08 points per possession using your 52% figure.

So a 52% make rate renders an output of a top 8 offensive team. A mere 60% free throw make rate yields a 1.2 per possession, which would lead the league.

Now where a "hack-a" strategy can be useful is three-fold.

1. It can be used to get into a players head.
2. It can be used to rest your team from running up and down the floor.
3. It can make sense if you're already down and you wish to gamble because your defense is proving to be a sieve. An all-in risky strategy.

No, it's the proper metric. As I understand the debate, the question is whether the Celtics would ultimately lose more game as a result of poor free throw shooters being fouled more often.

The objection to this line of reasoning is that compounding all free throws, and then looking at the true percentage, ACTUALLY, a free throw is a better percentage bet than letting the player take the shot.

Since the debate centers around: shooting v. free throw shooting, a metric which ignores free throw shooting is PRECISELY the appropriate metric to use. My argument (or at least the part of my blog post subject to debate) is whether teams would more likely foul our bad free throw shooters than allow them to shoot. It IS more likely, but only because Josh Smith is such an outrageously bad free throw shooter. If Smith were even as good as Rondo then there would be some question about the efficacy of this strategy - but, unfortunately, his free throw shooting is so bad, that you would rather be Milwuakee than have a Rondo/Smith Combo shooting your free throws.

Your framing makes zero sense and completely departs from what Bballtim initially was talking about and actual basketball. Break down the problem at its core when is it advantageous to foul Smith/Rondo types?

Fouling is a worse outcome than any shot Rondo/Smith take except if its at the rim. This is true even with their poor free throw shooting! (using Rondo's shooting from last year for sample size)

PPS for 3s for both of them: .69 for Josh, .72 for Rondo
PPS for 2 point jump shots: .63, .87
PPS at the rim:1.428, 1.196

Such a team wouldn't be fouled into oblivion, instead the paint would be packed and they'd attempt to force jump shots. Fouls would happen, but mostly on penetration into the paint to prevent shots at the rim. Just like with every other basketball player.

This is why coaches scream, don't foul jump shooters. Even if a guy makes 40% of his 3s, its better to not foul him if he even shoots 50% from the line.

By aggregating overall points per shot you're missing that crucial element. Beyond that, I still don't see what you're talking about in response to BBalltim's analysis of why "hack-a" strategies are ineffective.

If you're analyzing overall offense, you have to account for turnovers, offensive boards, and free throws.

If the question is why "hack-a" strategies are ineffective, the answer is, "because a team will get fewer points per free throw attempt than they will by taking a shot." Each free throw attempt (barring "and 1s") is worth 1.08 points, by Smith and Rondo, using their 2012-2013 numbers. A shot is worth 1.3 points, or something. Obviously, if a better free throw shooter is shooting, it's worth more. While turn-overs happen, they're beyond the scope of this analysis - as are offensive rebounds.

This is wrong a shot isn't worth 1.3 points. Because points per shot includes all the points scored on the free throw line. Furthermore if you're comparing "regular offense" to "hack a Rondo/Smith" you must account for turnovers/fta/offensive boards that occur with regular offense.

Teams don't get a FGA every possession in the course of a normal offensive trip. You have to use points per possession instead of points per shot.

Points per possesion is where its at for advanced look at NBA offense.

I think what you are ignoring is that "hack-a" strategies allow the opposing team to dictate what shot, precisely, is taken: a pair of free throws.

We are dealing in averages - so we can't really discuss whether the alternative would be a jump shot, or whatever - once you're in the penalty, maybe it would be a shot in the rim. You have to evaluate averages aganist averages. And the data in this instance is blind.

Using YOUR OWN MATH, the PPS from the FT line for Smith is .52, worse than any other alternative you cite above.

Perhaps my error was overstating the issues associated with pairing Smith and Rondo (although Rondo's shaky free throw shooting means that he can't reliably hold the ball to protect Smith).

You are right that for the given FTAs you ignore turnovers because the terms are given, 2 FTAs. (as an aside 2 FTAs are more valuable than indicated as OReb do occur on second free throw misses but the effect is slight(

But on the other side of the comparison you can't ignore turnovers, offensive boards, and free throws.

Also its not .52 pps for 2 Josh Smith FTAs. 2 FTAs = 1 possesion = 1 shot. So that means its 1.04 PPP

"Points Per Shot (PPS). This statistic measures the number of points a player scores per field goal attempt ((Total Points)/(Field Goal Attempts)).

A free throw is not a field goal attempt. An offensive rebound buttresses an offense, not the reverse. And turn-overs, although they happen, are effectively included in the percentages associated with any shot a player takes. ...we can't venture the debate into passing lanes, and so on - that's too remote.

Above, you argue PPS, and then here, you argue PPP. In determining PPP, you use the player's own poor free throw percentage against him - essentially arguing that, because his poor free throw shooting numbers drag down his PPP, shooting free throws is better than shooting.

If you stick with PPS, then you have to look at the true percentage chance of hitting each shot. One free throw is worth one point, so each individual free throw is worth .52 points when shot by Josh Smith. Otherwise, you need to calculate the true value of each attempt as opposed to each point. The value of a .52 FT shooter shooting two shots is not 1.04/2, it's .52 +(.52*.52) or .79, if you are arguing "per possession." He has to make both in order for it to count. 

...now, you happen to be right, even with those figures, but only because Josh Smith has such an abysmally low shooting percentage (on top of being a terrible free throw shooter). Josh Smith has the worst 3PFG% as measured by ESPN (145/145): http://espn.go.com/nba/statistics/player/_/stat/3-points/order/false

Again - Josh smith to Boston? No!

[Fafnir]

You're mangling per shot and per possession terribly. They're both useful if you understand what they mean. You clearly don't.

You can use a possession three ways in the NBA. All three use up possession of the ball.

1. You can shoot (and then from there one team will rebound the ball, offense gets a do over, or the ball changes hands)
2. You can turn it over
3. You can get to the free throw line typically for 2 FTAs though sometimes for 1 in an "and 1" or 3 for a 3 point foul (if the last free throw is missed there is again a rebounding situation)

You to look at all three outcomes to measure an offenses effectiveness. Your chosen comparison ignores that.

Also your comments about free throws is incoherent:

If you stick with PPS, then you have to look at the true percentage chance of hitting each shot. One free throw is worth one point, so each individual free throw is worth .52 points when shot by Josh Smith. Otherwise, you need to calculate the true value of each attempt as opposed to each point. The value of a .52 FT shooter shooting two shots is not 1.04/2, it's .52 +(.52*.52) or .79, if you are arguing "per possession." He has to make both in order for it to count. 

They don't have 1 and 1s in NBA basketball which is what it appears you are modeling for some reason. Josh Smith gets 2 free throws, he doesn't have to make the first to get the second.

In an intentional foul situation 2 FTA = 1 possession. So you have to measure the value of that possession against the average offensive possession. That means accounting for 1, 2, and 3. You can't just consider the shots if you want an apples to apples comparison.

.52 + .52 = 1.04 points per 2 FTA expected value for Josh Smith. That's one used possession. The full probability tree is:

0 points = .48^2 = .2304
1 point = .52*.48 + .48*.52 = .4992
2 points = .52^2 = .2704

0*.2304 + 1*.4992 + 2*.2704 = 1.04

[BballTim]

Also its not .52 pps for 2 Josh Smith FTAs. 2 FTAs = 1 possesion = 1 shot. So that means its 1.04 PPP

No, this is quite wrong. Your metric is pps, above, and now you are changing to ppp.

If you're going to be nerdy Jake, you should know that using Point Per Shot is the wrong measure to judge "hack-a" strategies relative efficiency. Point Per shot ignores turnovers and other free throw attempts, which are the other possibilities beyond a make or a miss.

Given you're supposing Rondo/Smith being wrapped up at half-court then by conceit you're ignoring the chance of a turnover or foul to another player during the hack.

Meanwhile comparatively playing traditional defense you want to look at points per possession because you can still get a turnover or foul a free throw shooter.

Portland scores at 1.12 points per possession. A composite Rondo/Smith hack strategy will yield 1.08 points per possession using your 52% figure.

So a 52% make rate renders an output of a top 8 offensive team. A mere 60% free throw make rate yields a 1.2 per possession, which would lead the league.

Now where a "hack-a" strategy can be useful is three-fold.

1. It can be used to get into a players head.
2. It can be used to rest your team from running up and down the floor.
3. It can make sense if you're already down and you wish to gamble because your defense is proving to be a sieve. An all-in risky strategy.

No, it's the proper metric. As I understand the debate, the question is whether the Celtics would ultimately lose more game as a result of poor free throw shooters being fouled more often.

The objection to this line of reasoning is that compounding all free throws, and then looking at the true percentage, ACTUALLY, a free throw is a better percentage bet than letting the player take the shot.

Since the debate centers around: shooting v. free throw shooting, a metric which ignores free throw shooting is PRECISELY the appropriate metric to use. My argument (or at least the part of my blog post subject to debate) is whether teams would more likely foul our bad free throw shooters than allow them to shoot. It IS more likely, but only because Josh Smith is such an outrageously bad free throw shooter. If Smith were even as good as Rondo then there would be some question about the efficacy of this strategy - but, unfortunately, his free throw shooting is so bad, that you would rather be Milwuakee than have a Rondo/Smith Combo shooting your free throws.

Your framing makes zero sense and completely departs from what Bballtim initially was talking about and actual basketball. Break down the problem at its core when is it advantageous to foul Smith/Rondo types?

Fouling is a worse outcome than any shot Rondo/Smith take except if its at the rim. This is true even with their poor free throw shooting! (using Rondo's shooting from last year for sample size)

PPS for 3s for both of them: .69 for Josh, .72 for Rondo
PPS for 2 point jump shots: .63, .87
PPS at the rim:1.428, 1.196

Such a team wouldn't be fouled into oblivion, instead the paint would be packed and they'd attempt to force jump shots. Fouls would happen, but mostly on penetration into the paint to prevent shots at the rim. Just like with every other basketball player.

This is why coaches scream, don't foul jump shooters. Even if a guy makes 40% of his 3s, its better to not foul him if he even shoots 50% from the line.

By aggregating overall points per shot you're missing that crucial element. Beyond that, I still don't see what you're talking about in response to BBalltim's analysis of why "hack-a" strategies are ineffective.

If you're analyzing overall offense, you have to account for turnovers, offensive boards, and free throws.

If the question is why "hack-a" strategies are ineffective, the answer is, "because a team will get fewer points per free throw attempt than they will by taking a shot." Each free throw attempt (barring "and 1s") is worth 1.08 points, by Smith and Rondo, using their 2012-2013 numbers. A shot is worth 1.3 points, or something. Obviously, if a better free throw shooter is shooting, it's worth more. While turn-overs happen, they're beyond the scope of this analysis - as are offensive rebounds.

This is wrong a shot isn't worth 1.3 points. Because points per shot includes all the points scored on the free throw line. Furthermore if you're comparing "regular offense" to "hack a Rondo/Smith" you must account for turnovers/fta/offensive boards that occur with regular offense.

Teams don't get a FGA every possession in the course of a normal offensive trip. You have to use points per possession instead of points per shot.

Points per possesion is where its at for advanced look at NBA offense.

I think what you are ignoring is that "hack-a" strategies allow the opposing team to dictate what shot, precisely, is taken: a pair of free throws.

We are dealing in averages - so we can't really discuss whether the alternative would be a jump shot, or whatever - once you're in the penalty, maybe it would be a shot in the rim. You have to evaluate averages aganist averages. And the data in this instance is blind.

Using YOUR OWN MATH, the PPS from the FT line for Smith is .52, worse than any other alternative you cite above.

Perhaps my error was overstating the issues associated with pairing Smith and Rondo (although Rondo's shaky free throw shooting means that he can't reliably hold the ball to protect Smith).

You are right that for the given FTAs you ignore turnovers because the terms are given, 2 FTAs. (as an aside 2 FTAs are more valuable than indicated as OReb do occur on second free throw misses but the effect is slight(

But on the other side of the comparison you can't ignore turnovers, offensive boards, and free throws.

Also its not .52 pps for 2 Josh Smith FTAs. 2 FTAs = 1 possesion = 1 shot. So that means its 1.04 PPP

"Points Per Shot (PPS). This statistic measures the number of points a player scores per field goal attempt ((Total Points)/(Field Goal Attempts)).

A free throw is not a field goal attempt. An offensive rebound buttresses an offense, not the reverse. And turn-overs, although they happen, are effectively included in the percentages associated with any shot a player takes. ...we can't venture the debate into passing lanes, and so on - that's too remote.

Above, you argue PPS, and then here, you argue PPP. In determining PPP, you use the player's own poor free throw percentage against him - essentially arguing that, because his poor free throw shooting numbers drag down his PPP, shooting free throws is better than shooting.

If you stick with PPS, then you have to look at the true percentage chance of hitting each shot. One free throw is worth one point, so each individual free throw is worth .52 points when shot by Josh Smith. Otherwise, you need to calculate the true value of each attempt as opposed to each point. The value of a .52 FT shooter shooting two shots is not 1.04/2, it's .52 +(.52*.52) or .79, if you are arguing "per possession." He has to make both in order for it to count. 

...now, you happen to be right, even with those figures, but only because Josh Smith has such an abysmally low shooting percentage (on top of being a terrible free throw shooter). Josh Smith has the worst 3PFG% as measured by ESPN (145/145): http://espn.go.com/nba/statistics/player/_/stat/3-points/order/false

Again - Josh smith to Boston? No!

  What you're saying is just plain wrong. Points per shot doesn't measure anything, they might as well have points per steal or points per turnover. LeBron gets 1.56 points per shot because he scores 26.3 points and he takes 16.8 FGA/game. But he doesn't score 26.3 points on those 16.8 shots, he scores 20.5 points on those 16.8 shots. The rest of his points come from free throws. If you want to find out how many points he gets (on average) when he shoots the ball it would be somewhere between his eFG% (61%) and his TS% (65%). So when he shoots the ball he gets between 1.2 and 1.3 points per shot, not 1.56.

  So LeBron, when he shoots the ball, scores about as well as Rondo scored on free throws last year. But, again, no team in the history of the league has scored as well as LeBron. No team in nba history scores as well as Rondo scored from the line last year.

  If you need any more reason to wonder whether hack-a-Rondo is an effective strategy, consider that we've seen little to none of it in the last 7 years.

[BballTim]

I'm old enough to remember Nick Anderson. The end of games, where teams that are behind have to foul is the only area of concern. You could just put Smith on the bench for that. That's obviously far different than the non-stop hacking you were talking about.

And Rondo. And don't forget to not give Rondo the ball in the last 2 minutes, either.

  Rondo's been getting the ball in the last 2 minutes of games for at least the last few years.

[kozlodoev]

I'm old enough to remember Nick Anderson. The end of games, where teams that are behind have to foul is the only area of concern. You could just put Smith on the bench for that. That's obviously far different than the non-stop hacking you were talking about.

And Rondo. And don't forget to not give Rondo the ball in the last 2 minutes, either.

  Rondo's been getting the ball in the last 2 minutes of games for at least the last few years.

Not as much as you think. Pierce has always been doing the overwhelming amount of work in crunch time.

[Boris Badenov]

I'm old enough to remember Nick Anderson. The end of games, where teams that are behind have to foul is the only area of concern. You could just put Smith on the bench for that. That's obviously far different than the non-stop hacking you were talking about.

And Rondo. And don't forget to not give Rondo the ball in the last 2 minutes, either.

  Rondo's been getting the ball in the last 2 minutes of games for at least the last few years.

Not as much as you think. Pierce has always been doing the overwhelming amount of work in crunch time.

The samples are very small but last year Pierce averaged 24fga, 8fta and 7ast per 48 minutes of "clutch" time, according to 82games.

Rondo averaged 15fga, 6fta and 13ast.

http://www.82games.com/1213/CSORT11.HTM

If I'm doing the math right, that's 35 possessions per 48 for Pierce with a shot, foul or assist, and 31 for Rondo. It seems like both have the ball a fair amount.

I took a quick look at 2011-12 and it's more weighted toward PIerce but not by a whole lot.

There may be other measures, too, of course.

Edit: last year Rondo shot 75% on FTs in the "clutch" while Pierce shot 73%. I'm sure these are in laughably small samples, but it's fun trivia.

[BballTim]

I'm old enough to remember Nick Anderson. The end of games, where teams that are behind have to foul is the only area of concern. You could just put Smith on the bench for that. That's obviously far different than the non-stop hacking you were talking about.

And Rondo. And don't forget to not give Rondo the ball in the last 2 minutes, either.

  Rondo's been getting the ball in the last 2 minutes of games for at least the last few years.

Not as much as you think. Pierce has always been doing the overwhelming amount of work in crunch time.

  Not for the last few years. PP might have shot the ball but in 2010 or so as soon as they got down to 2-3 minutes left PP would bring the ball up and running the offense. That stopped a few years ago. Rondo would bring the ball up and initiate the offense all the way up to at least the last possession or two. He wouldn't have the ball as much if the other team needed to commit an intentional foul but that's about it.

[Fafnir]

Rondo's injury would skew last years numbers. Pierce definitely was the primary ball handler in crunch time once Rondo was down.

[Boris Badenov]

Rondo's injury would skew last years numbers. Pierce definitely was the primary ball handler in crunch time once Rondo was down.

That's why I checked numbers from the previous year too - or one of the reasons anyway.

[kozlodoev]

Not for the last few years. PP might have shot the ball but in 2010 or so as soon as they got down to 2-3 minutes left PP would bring the ball up and running the offense. That stopped a few years ago. Rondo would bring the ball up and initiate the offense all the way up to at least the last possession or two. He wouldn't have the ball as much if the other team needed to commit an intentional foul but that's about it.

Fair enough. The thing is, you no longer have Pierce to dump the ball to in half-court, and get out of the way.