X points on X shots... split from Cavs-Kings game thread

[6th]

See, that's the thing. Obviously, he uses a possession to score the points. That fact SHOULD be obvious enough to go without saying. I have yet to see any scorer in the NBA score anything besides the one or two points afforded by illegal defense calls, technical fouls etc. WITHOUT using a possession.
Now, if we could figure out how to get him to the line without using a possession, we'd be in great shape

I have an idea. Think we could put Kevin at the line when any Kings player gets fouled? Think they'd notice.

 

[Kingsgurl]

First, I'm fairly certain they would. Second, that would still be using one of OUR possessions. Now, if we could send Kevin to the line each time one of the OPPOSING players was fouled (and both keep the points and get the ball back) THEN he could score without using a possession. (I'm fairly certain they would also notice this)

 

[thelonegunman]

On a side note, does anybody recall that Kevin Martin set an NBA record last season by scoring 20 points while only making 1 FG?

He was 1-8 from the floor and 17-20 from the FT line. Crazy.



Also, putting in my 2 cents on the subject at hand....

if a player takes 10 shots, gets fouled on all 10 attempts, and makes 16 of the 20 free throws, it's as if he made 8 of the 10 field goal attempts. That's damn efficient.

Taking it a step further, if a player got fouled each time he took a shot and only made 50% of his free throws (which we agree would be a horrible FT%) it would still equate to shooting 50% from the field, which is very good. Essentially, he'd score 1 point for every shot attempt. Do that consistently and you'll win a ton of games.

 

[Bricklayer]

Copied from the game thread so as not to disrupt its flow:

Kevin has been the subject of a slight dispute over whether or not his efficiency is really efficient. While I agree that taking 25 possessions to get 30 points is in all reality what he does, but saying this discredits his abilities and must be looked at a little bit deeper. It isn't just 32 points on 17 shots, because that is absurd for anyone to think, but he is one of the more efficient players in the league. I only say this because he shoots a high percentage from the field as well, so regardless of his FT attempts the man should have the ball in his hands often. In effect the possession ends when the foul is called, so the argument that he takes a full possession is frustrating, because often Kevin goes to the line early in the shot clock anyway. All I am saying is Fts does not equal a possession.

First of all, that its not a possession because it happens early in the shot clock is ridiculous. Phoenix, at that rate, does not actually have any possessions, and the great key to efficiency is shooting as soon as you cross halfcourt (or drawing a foul). Last I checked shooting the ball 4 seconds into a possession was still a possession just as much as shooting it 18 seconds into a possession. There is no difference. Never has been. Never will be. The possession starts when you get the ball in your hands, and ends 1-24 seconds later when you shoot the ball, are fouled, or turn it over (occasionally longer if you grab an offensive rebound). There is no "full possession" or "half possession". It is always just a possession -- and only 1 player shoots/gets to score. (And that's without even getting to the issue of whether Kevin is drawing fouls earlier in the possesssion than shots go up or not).

Secondly, while I do not think it is terribly relevant to the issue, Kevin is not shooting very well this year -- .437 from the field. And there the increased defensive pressure is perhaps showing. The foul shots ARE his effectiveness this year in the face of the pressure -- as long as he can keep going to the line at a record setting pace, he can keep on scoring a ton of points.

Thirdly, and once again, accurately counting Kevin's efficiency is not discrediting Kevin. It is in fact crediting him correctly. And the issue has never been that him getting to the line is not efficient, it has been that it is not nearly so efficient as certain members of the sales team work very hard to make you believe. And it may be more important in assessing Kevin Martin than it is any other player in the NBA, because he takes (at least among the major scorers) a higher percentage of FTs than just about any other player. And the worst parts are that a) its unnecessary, since a rational/reasonable/correct accounting of Kevin's efficiency still leaves him looking quite good on most nights (the last two nights the correct numbers are 32pts on 25 possessions and 29pts on 24 possessions); and b) that the argument is obvious on its face and people continue to so desperately try to make excuses for ignoring it.*


*The best/only on point argument was the one advanced by 1kingszfan that technical FTs (from which Kevin gets a free point or two every game) and continuations raise the number of FTs without taking possessions. Which is true, but a minor distortion certainly fairly easy to account for each game (indeed I have done so on those 32 on 25 possessions etc. stats above) , and likely largely accounted for by a simple rough multiplier of let's say .85 for the whole season (meaning 85% of Kevin;s Fts are "normal", and the other 15% are technical FTs, continuations, or a third FT because he got fouled ona three point shot. Strongly susepct that would be within 5% either way.

 

[Cal-Kings]

Jerry is just using the common way to refer to these stats.

For example, Kobe averages 30.4 points on 21 shots a game.

It's how the stats are calculated and discussed and compared.

To count Kevin's free throws is fine, but to compare that to another scorer, one would also have to count their free throws.

Kobe averages 10.4 free throws a game, so divided by 2 and added to 21, Kobe averages 30.4 points on 26.2 shots a game.

Interestingly, that comes out to about the same relationship between the two. Kobe is a tad more efficient.

I expect all scorers who have a high average take a lot of free throws. Kevin is no different.

So I don't understand the fuss. Jerry is just using the typical way this number is talked about. Kobe gets 30.4 points on 21 shots, Kevin gets 28 points on 17 shots.

I think the problem was noticed by Jerry last year because Kevin averaged 20.2 points on 13.3 attempts (7.1 free thow attempts). A scorer of that proficiency should be taking "more shots".

I think we can all agree Kevin is indeed getting more shots this year, especially if he's in Kobe range.

I just think that the reason Kevin is getting more is largely because Bibby and Artest are not there.

I'm hoping that when they return, Kevin is still takes the majority of shots.

There are two players shooting a higher percentage than Kevin on the team, and that's Salmons and Garcia. But Kevin is below his career average, while they are above their career averages.

Is the focus on Kevin helping Salmons and Garcia? It appears so.

The other factor is, we've only played 6 games... it's way too early for stats to mean all that much.

So, Brick, are you saying that the Kings are hyping Kevin unjustifiably? I don't know. If he's managing to score on a level of Kobe and McGrady, I think that's worth mentioning when we're desperately looking for positives on this team.

 

[fnordius]

To count Kevin's free throws is fine, but to compare that to another scorer, one would also have to count their free throws.

Yeah. Getting high points per shot is good, but there's a lot of difference in how players get there.

The best way is how Dwight Howard does it. He's only a 61% FT shooter, but he also hit over 60% of his FG attempts last year, and got fouled while shooting a whole lot. So he had the highest PPS in the league last year, and did it at no cost to his team; he had a better chance of hitting the shot than any of his teammates, and the frequent +1s added up. Unless you're going for a Muss 5-point play, getting 3 points out of a possession (with the added bonus of maybe getting an opponent into foul trouble) can be considered perfect.

The second best way is to be a deadly 3-point shooter. This is how Steve Nash made it into the top 10 for PPS in the NBA last year, with over 45% three point shooting and 50% FG shooting. You don't have the chance at getting your opponents into foul trouble, but if your team can make 3 points off a possession, there's really nothing you can be criticized for.

The last, and least good way, is to up your PPS by going to the FT line a lot without hitting a shot. It still makes your personal stat line look great, but it gives your team only 2/3 as many points per possession as the other ways. Maybe you will get someone into foul trouble, but in most games it won't have much impact.

But is points per possession even that significant?

Probably. The top teams for PPP last year were: Phoenix, Seattle, Washington, Miami, Denver and Toronto. The bottom were Portland, Houston, Utah, Memphis, Minnesota, NJ and Indiana. So of the top PPP teams, 5/6 went to the playoffs, and they averaged 54.7% wins. The low PPP teams mostly went to the lottery, and they averaged 46.1% wins. You can win as a low PPP team, but only if you have someone like AK or Yao who can keep your opponent's score even lower than yours.

 

[GoGoGadget]

So, Brick, are you saying that the Kings are hyping Kevin unjustifiably? I don't know. If he's managing to score on a level of Kobe and McGrady, I think that's worth mentioning when we're desperately looking for positives on this team.

I was under the impression that this conversation was brought up not to argue against Kevin's efficiency, but to counter the thought that Kevin is not getting his fair share of shots in a game.

 

[uolj]

Points per shot is just a rough estimate of points per possession.

Points per possession is the more accurate indicator of efficiency and accomplishment. However, it is also harder to calculate based on the data in current box scores.

Therefore, people often use points per shot instead of points per possession.

I'm sure there are people who don't take the time to recognize the difference, which can lead to confusion. I personally wish possession data was more readily available for my stat-crunching pleasure. However, since it's not then it isn't that big of a deal to continue using points per shot as a substitute.

If somebody could do some analysis that showed that points per shot doesn't correlate well to points per possession, then it might be a bigger deal.

 

[venom_7]

Copied from the game thread so as not to disrupt its flow:

[/color]

First of all, that its not a possession because it happens early in the shot clock is ridiculous. Phoenix, at that rate, does not actually have any possessions, and the great key to efficiency is shooting as soon as you cross halfcourt (or drawing a foul). Last I checked shooting the ball 4 seconds into a possession was still a possession just as much as shooting it 18 seconds into a possession. There is no difference. Never has been. Never will be. The possession starts when you get the ball in your hands, and ends 1-24 seconds later when you shoot the ball, are fouled, or turn it over (occasionally longer if you grab an offensive rebound). There is no "full possession" or "half possession". It is always just a possession -- and only 1 player shoots/gets to score. (And that's without even getting to the issue of whether Kevin is drawing fouls earlier in the possesssion than shots go up or not).

Secondly, while I do not think it is terribly relevant to the issue, Kevin is not shooting very well this year -- .437 from the field. And there the increased defensive pressure is perhaps showing. The foul shots ARE his effectiveness this year in the face of the pressure -- as long as he can keep going to the line at a record setting pace, he can keep on scoring a ton of points.

Thirdly, and once again, accurately counting Kevin's efficiency is not discrediting Kevin. It is in fact crediting him correctly. And the issue has never been that him getting to the line is not efficient, it has been that it is not nearly so efficient as certain members of the sales team work very hard to make you believe. And it may be more important in assessing Kevin Martin than it is any other player in the NBA, because he takes (at least among the major scorers) a higher percentage of FTs than just about any other player. And the worst parts are that a) its unnecessary, since a rational/reasonable/correct accounting of Kevin's efficiency still leaves him looking quite good on most nights (the last two nights the correct numbers are 32pts on 25 possessions and 29pts on 24 possessions); and b) that the argument is obvious on its face and people continue to so desperately try to make excuses for ignoring it.*


*The best/only on point argument was the one advanced by 1kingszfan that technical FTs (from which Kevin gets a free point or two every game) and continuations raise the number of FTs without taking possessions. Which is true, but a minor distortion certainly fairly easy to account for each game (indeed I have done so on those 32 on 25 possessions etc. stats above) , and likely largely accounted for by a simple rough multiplier of let's say .85 for the whole season (meaning 85% of Kevin;s Fts are "normal", and the other 15% are technical FTs, continuations, or a third FT because he got fouled ona three point shot. Strongly susepct that would be within 5% either way.

haha ok ok you win. I was just trying to say it isn't the same in my eyes, but your points make total and complete sense. I still don't consider 2 free throws a shot attempt. I guess thats what I was getting at. Not a full shot attempt. But I am on your side too, so yeah...

 

[ZanesvilleFan]

So what people are saying is that you should add his field goal attempts and his free throw attempts up and add his field goal attempts made and his free throw attempts made...and make that his field goal %...

 

[fnordius]

If somebody could do some analysis that showed that points per shot doesn't correlate well to points per possession, then it might be a bigger deal.

(Note: PPP stats are per 100 possessions)

I'll show the stats, people will have to draw their own conclusions about how meaningful they are. While neither one's a perfect predictor of who will win games, I note that the top 3 PPP teams were also 3 of the top 4 winningest teams, and 9 out of 10 teams that broke 105 PPP went to the playoffs, whereas Memphis, the worst team in the NBA, was tied for 2nd highest PPS, but had no more than lukewarm PPP. Detroit kicked butt in PPP, but did quite badly at PPS (tied with Atlanta and Minnesota). And a couple of teams, Chicago and Houston, did well despite being low in both PPS and PPP. Seattle stunk despite doing well at both. So depending on the makeup of the team, the two may go together, but much of the time they don't. I'll leave more serious number crunching to someone else.

Team------W/L-----PPS----PPP
Dal--------.817-----1.27-----109.5
Phx-------.744-----1.32-----110.0
SAS------.707-----1.28-----105.6
Det--------.646-----1.21-----109.4
Hou-------.634-----1.22-----100.3
Uta--------.622-----1.29-----100.1
Cle--------.610-----1.19-----105.3
Chi--------.598-----1.22-----101.0
Tor--------.573-----1.25-----108.0
Den-------.549-----1.27-----103.9
Mia-------.537------1.23-----108.0
LAL-------.512-----1.27-----105.7
GSW-----.512-----1.24-----102.6
NJN------.500-----1.25-----102.1
Was------.500-----1.25-----106.7
Orl--------.488-----1.28-----103.2
LAC------.488-----1.24-----103.5
NO--------.476-----1.17-----101.2
Ind--------.427-----1.20-----102.1
Phi-------.427-----1.21-----104.3
Sac------.402-----1.27-----104.4
Cha------.402-----1.20-----99.3
NY-------.402-----1.26-----101.9
Ptl-------.390-----1.21-----98.9
Min------.390-----1.21-----101.2
Sea-----.378-----1.23-----108.4
Atl-------.366-----1.21-----104.3
Mil-------.341-----1.22-----102.8
Bos------.293-----1.22-----102.9
Mem----.268-----1.29-----103.3

 

[uolj]

So what people are saying is that you should add his field goal attempts and his free throw attempts up and add his field goal attempts made and his free throw attempts made...and make that his field goal %...

No.

A better rough estimate of efficiency than points per shot is to calculate possessions used to obtain points as:

Possessions = Field goal attempts + (0.45 * Free throw attempts)

Then look at points per possession (points divided by possessions).

You can round 0.45 to 0.5 to make the math easier for a quick look. That makes the equation:

points / (FGA + (FTA/2))


So this year Kevin Martin's line looks like this:

[B]G	FGM-A  	3PM-A  	FTM-A  	PTS[/B]
6	45-103 	9-26 	69-81 	168

So possessions used for points = 103 + (0.45 * 81) = 139 or using quick math 103 + 81/2 = 143 which is close enough.

So Martin has approximately 168 points in 139 possessions (1.20). By that formula Kobe Bryant has 152 points in 128 possessions or 1.18 (in 5 games instead of 6). Tracy McGrady is 191/177 (1.08), LeBron James is 186/180 (1.03).

 

[uolj]

(Note: PPP stats are per 100 possessions)

I'll show the stats, people will have to draw their own conclusions about how meaningful they are. While neither one's a perfect predictor of who will win games, I note that the top 3 PPP teams were also 3 of the top 4 winningest teams, and 9 out of 10 teams that broke 105 PPP went to the playoffs, whereas Memphis, the worst team in the NBA, was tied for 2nd highest PPS, but had no more than lukewarm PPP. Detroit kicked butt in PPP, but did quite badly at PPS (tied with Atlanta and Minnesota). And a couple of teams, Chicago and Houston, did well despite being low in both PPS and PPP. Seattle stunk despite doing well at both. So depending on the makeup of the team, the two may go together, but much of the time they don't. I'll leave more serious number crunching to someone else.

There should be no question about which is the better stat. Points per possession should be more accurate in general. The question is whether points per shot is usually pretty close. Your data shows that it isn't as accurate as I might like overall, which would lead credence to an argument that people should try not to put as much weight on points per shot. These are the team rankings based on the data in your post. If the two stats measured the same things, the rankings should be pretty close:

Edit: New data in post below...

 

[fnordius]

My stats for PPP came from a posting on nba.com: http://www.nba.com/celtics/stats/inside-the-numbers/032106-efficiency.html

My stats for PPS came from the stat page for each team (2006-7) on the ESPN website.

I trusted the accuracy of the NBA and ESPN websites, and made no attempt to verify the stats elsewhere.

 

[uolj]

I don't have any reason to doubt you, but the source of the stats matter. Using ESPN's data from here, the PPS for Utah and Detroit is 1.29 and 1.21 respectively, but the PPP based on my formula above is 1.098 versus 1.067, which is very different than the numbers in the article you cited (Utah is better in both here, but in your data it switches).

Maybe they use a more accurate measure of possessions, or they use a different definition entirely, but either way the point I made above no longer holds water, so I removed it.

 

[uolj]

Ok, so these are the rankings for the teams in PPS and PPP for 2006. These values were based on Points/FGA and Points/(FGA + (FTA*.45)) gathered from here.

TEAM		PPS RNK	PPP RNK
Phoenix		1	1
San Antonio	4	2
Dallas		6	3
Memphis		2	4
LA Lakers	7	5
Utah		3	6
Toronto		11	7
Denver		8	8
Golden State	14	9
New Jersey	12	10
Sacramento	9	11
Seattle		16	12
Orlando		5	13
Washington	13	14
Milwaukee	18	15
Miami		17	16
Houston		19	17
New York	10	18
Minnesota	22	19
LA Clippers	15	20
Detroit		23	21
Chicago		20	22
Portland	24	23
Boston		21	24
Philadelphia	25	25
Charlotte	27	26
Indiana		28	27
Atlanta		26	28
Cleveland	29	29
New Orleans	30	30

So at least on a large scale, maybe points per shot isn't that much different than points per possession (except when you're trying to argue that someone isn't getting enough shots ).

 

[ZanesvilleFan]

No.

A better rough estimate of efficiency than points per shot is to calculate possessions used to obtain points as:

Possessions = Field goal attempts + (0.45 * Free throw attempts)

Then look at points per possession (points divided by possessions).

You can round 0.45 to 0.5 to make the math easier for a quick look. That makes the equation:

points / (FGA + (FTA/2))


So this year Kevin Martin's line looks like this:

[B]G    FGM-A      3PM-A      FTM-A      PTS[/B]
6    45-103     9-26     69-81     168

So possessions used for points = 103 + (0.45 * 81) = 139 or using quick math 103 + 81/2 = 143 which is close enough.

So Martin has approximately 168 points in 139 possessions (1.20). By that formula Kobe Bryant has 152 points in 128 possessions or 1.18 (in 5 games instead of 6). Tracy McGrady is 191/177 (1.08), LeBron James is 186/180 (1.03).

Is that the same thing as what they call an OER (offense efficiancy rating) if so we do something like that at zanesville till this day but im not sure how they come up with the numbers. but they do it as a team. Then they do a team DER (defense of efficiancy rating) i will have to double check on that and see how they get the numbers. it seems to me that the math you have done is very similar.

thanks for the info

 

[1kingzfan]

There can be no doubt, none whatsoever, that points per possession is a much more meaningful stat to judge player scoring efficiency (and compare players of different styles) than points per shot.

The problem is getting that darned possessions per game for a player. It can be done, but it's not readily available to you and me. We can get team possessions but that does not necessarily correlate to a single player's possessions. Recall how many games last year (and the year before actually) where Kevin never got a touch until 4, 6, 8 minutes into a quarter, mainly because the O didn't come his way?

And even if we had a count for possessions where KMart got a touch (for instance, by using the play-by-play box scores), how many times does he, or any player for that matter, get a pass in a bad position, in the middle of a double team, or directly under the basket where he can't do anything? So you really have to count QUALITY TOUCHES (POSSESSIONS) to tell the story...anybody out there doing that???

Oh wait, what happens when KMart gets the ball in a reasonable position where he can score in the flow of the offense or drive to score, and then he drives and dishes off to another player for a basket? That's a productive possession for KMart, isn't it, even though he doesn't score?

Oh, wait again, what happens if KMart dishes the ball beautifully to set up a teammate and the teammate blows the lay-up or the little chippie or the wide open 3 or fumbles a great pass out of bounds or travels? Is that an unproductive possession for Kevin that should affect his scoring efficiency?

Oh, wait again again, what if Kevin has the ball and a teammate gets an offensive foul? Should that possession count against Kevin's SCORING efficiency?

I am sure there are a number of other problems with counting possessions, and trying to figure out quality possessions for a player, that are not mentioned here.

The bottom line is that the "truth" we all seek is incredibly difficult to account for and tally up. Therefore, for me, points per shot (PPS) is a nice little summary, very easy to obtain and calculate, and which should work as an INDICATOR (not the end all) of player scoring productivity. We can argue about which indicator is best, but they will all be just that, indicators, and not the optimal all-telling stat.

Folks should not interpret the lower number of shots Kevin gets credited for (by stat rules) as meaning that he is not getting enough touches. However, my opinion is that he could still use a few more plays designed for him each game, given his propensity for getting past his man, drawing fouls, and, well, being productive in putting points on the Kings’ side the scoreboard.

I just love these stat discussions…I learn something new every time.

 

[fnordius]

Some of the anomalies aren't too hard to figure out. Seattle, for example, ruined their nice PPP stats by turning the ball over more often than any other team in the NBA, and letting their opponents shoot almost 47% (3rd worst in the NBA, behind Memphis and Milwaukee). Detroit did very well in the TO department, forcing a lot more out of their opponents than they made themselves, which would give their efficiency a nice boost. They also held their opponents to under 44.5% shooting (6th best in the NBA) which wouldn't hurt. Houston and Chicago also defended very well (42.9 and 43.5% FG allowed, top 2 in the NBA), which would account for their deviating from the expected.

Utah is a bit tougher to figure out. There are stats which explain winning with low PPS/PPP, but it's harder to see why the PPS and PPP stats should be so different. Part of it would be a result of their bad TO differential, since turning the ball over will drop your PPP without changing your PPS.

Which brings up an interesting thought. Last year, Darko Milicic made one TO for every 17 shots he took. Dwyane Wade made one TO for every 3.6 shots he took. A simple statistical model, which assumes that PPS and PPP will be close, will therefore work beautifully on Darko, since there is only a 5.5% difference between his possessions and his shots. With Wade, the difference would be more like 22%, so his PPS might be dandy while his PPP kind of stunk.

I'm not sure what to make of your statistical reconciliation, since that assumes that your model is more accurate than the stats provided by ESPN and the NBA. And, while I can show that a simple model can be considerably off the mark, I can't prove anything about the accuracy of ESPN or the NBA. So I think I'm going to watch the game now.