Bonn1997 wrote:

mreinman wrote:

Bonn1997 wrote:

mreinman wrote:

Bonn1997 wrote:

mreinman wrote:Go to the below page and sort by Win Shares and you would probably get the closest advanced stat to reality (there will always be anomalies with every stat)

http://www.basketball-reference.com/leagues/NBA_2014_advanced.html

Top 20:

Rk Player WS ▾
1 Kevin Durant 18.4
2 LeBron James 15.2
3 Kevin Love 13.8
4 Stephen Curry 12.3
5 Blake Griffin 11.8
6 James Harden 11.7
7 Chris Paul 11.7
8 DeAndre Jordan 10.8
9 Kyle Lowry 10.8
10 Paul George 10.6
11 Anthony Davis 10.5
12 Joakim Noah 10.4
13 Carmelo Anthony 10.3
14 Goran Dragic 10.2
15 Dirk Nowitzki 10.2
16 Andre Drummond 9.2
17 Damian Lillard 9.2
18 Robin Lopez 9
19 Serge Ibaka 8.8
20 DeMar DeRozan 8.5

What makes it the closest to reality? At the very least, you can make a good argument for using WS48 because you have to adjust for minutes played but I think there are other useful stats too.

I did not say that there aren't other useful stats. Everything has its flaws. As you know WP's has an argument against Berri for over weighted rebounding. etc ...

WS48 is good too but what if a player is able to log more minutes? Actual WS's contributed seem to be more closely related to wins.

And if you look at the list, they seem to get it right - but that does not mean its correct.

Certainly a more realistic wins gauge than RPMS?

I disagree. I've never seen any evidence that there's a consistent, big drop off in production as minutes increase. At most, you could look at WS48 and maybe do a tiny adjustment in your mind. Total WS is too confounded with minutes played though.

Minutes played is the actual minutes played. If a player plays 40 per and contributes more than a player who plays 35 minutes then he contributed more.

If want to make an argument that the player who played 35 minutes with the player who played 5 minutes as his backup then perhaps I can see some truth to that but there were need to be a stat that compiles this combined production. Backups very often dont perform at the level of the starter.

How about RPMS? What stat(s) do you go by as your barometer? You are a big fan of David Berri but now have a problem with WP's. Now I know stats are forever evolving but at any given time, you have to choose something. It can't keep changing based on the required outcome (not saying that you do this, but many do).

I give some weight to most of the advanced stats including the nba tracking data. I definitely give win shares a good amount of weight.
So who would you give a higher contract to? A guy who plays great in 20 mpg or an ordinary player with a higher win share total because he's playing 32 mpg?

A player that can average a large number of minutes and still remain efficient. There certainly is going to be diminished returns if you max out a guys minutes. You don't believe that rest improves efficiency?

This is why Durant will probably get the MVP. He has logged 2889 (2nd to Melo) and is at a freaky .306 (wp48). James is at .265.

Take Kawhi Leonard. At 1800 minutes, he is ranked number 12 in the league in WS48 at .192 - Do you max him out before Pual George, Noah or Howard?

dk7th wrote:here is list of the league. note that tyson chandler is ranked 28th and carmelo anthony is ranked 52nd.

http://espn.go.com/nba/statistics/rpm/_/sort/RPM

Note that Carmelo Anthony is 52 while the vaunted Paul George is 58.

mreinman wrote:

Bonn1997 wrote:

mreinman wrote:

Bonn1997 wrote:

mreinman wrote:

Bonn1997 wrote:

mreinman wrote:Go to the below page and sort by Win Shares and you would probably get the closest advanced stat to reality (there will always be anomalies with every stat)

http://www.basketball-reference.com/leagues/NBA_2014_advanced.html

Top 20:

Rk Player WS ▾
1 Kevin Durant 18.4
2 LeBron James 15.2
3 Kevin Love 13.8
4 Stephen Curry 12.3
5 Blake Griffin 11.8
6 James Harden 11.7
7 Chris Paul 11.7
8 DeAndre Jordan 10.8
9 Kyle Lowry 10.8
10 Paul George 10.6
11 Anthony Davis 10.5
12 Joakim Noah 10.4
13 Carmelo Anthony 10.3
14 Goran Dragic 10.2
15 Dirk Nowitzki 10.2
16 Andre Drummond 9.2
17 Damian Lillard 9.2
18 Robin Lopez 9
19 Serge Ibaka 8.8
20 DeMar DeRozan 8.5

What makes it the closest to reality? At the very least, you can make a good argument for using WS48 because you have to adjust for minutes played but I think there are other useful stats too.

I did not say that there aren't other useful stats. Everything has its flaws. As you know WP's has an argument against Berri for over weighted rebounding. etc ...

WS48 is good too but what if a player is able to log more minutes? Actual WS's contributed seem to be more closely related to wins.

And if you look at the list, they seem to get it right - but that does not mean its correct.

Certainly a more realistic wins gauge than RPMS?

I disagree. I've never seen any evidence that there's a consistent, big drop off in production as minutes increase. At most, you could look at WS48 and maybe do a tiny adjustment in your mind. Total WS is too confounded with minutes played though.

Minutes played is the actual minutes played. If a player plays 40 per and contributes more than a player who plays 35 minutes then he contributed more.

If want to make an argument that the player who played 35 minutes with the player who played 5 minutes as his backup then perhaps I can see some truth to that but there were need to be a stat that compiles this combined production. Backups very often dont perform at the level of the starter.

How about RPMS? What stat(s) do you go by as your barometer? You are a big fan of David Berri but now have a problem with WP's. Now I know stats are forever evolving but at any given time, you have to choose something. It can't keep changing based on the required outcome (not saying that you do this, but many do).

I give some weight to most of the advanced stats including the nba tracking data. I definitely give win shares a good amount of weight.
So who would you give a higher contract to? A guy who plays great in 20 mpg or an ordinary player with a higher win share total because he's playing 32 mpg?

A player that can average a large number of minutes and still remain efficient. There certainly is going to be diminished returns if you max out a guys minutes. You don't believe that rest improves efficiency?

This is why Durant will probably get the MVP. He has logged 2889 (2nd to Melo) and is at a freaky .306 (wp48). James is at .265.

Take Kawhi Leonard. At 1800 minutes, he is ranked number 12 in the league in WS48 at .192 - Do you max him out before Pual George, Noah or Howard?

mreinman wrote:

Bonn1997 wrote:

mreinman wrote:

Bonn1997 wrote:

mreinman wrote:

Bonn1997 wrote:

mreinman wrote:

Bonn1997 wrote:

mreinman wrote:Go to the below page and sort by Win Shares and you would probably get the closest advanced stat to reality (there will always be anomalies with every stat)

http://www.basketball-reference.com/leagues/NBA_2014_advanced.html

Top 20:

Rk Player WS ▾
1 Kevin Durant 18.4
2 LeBron James 15.2
3 Kevin Love 13.8
4 Stephen Curry 12.3
5 Blake Griffin 11.8
6 James Harden 11.7
7 Chris Paul 11.7
8 DeAndre Jordan 10.8
9 Kyle Lowry 10.8
10 Paul George 10.6
11 Anthony Davis 10.5
12 Joakim Noah 10.4
13 Carmelo Anthony 10.3
14 Goran Dragic 10.2
15 Dirk Nowitzki 10.2
16 Andre Drummond 9.2
17 Damian Lillard 9.2
18 Robin Lopez 9
19 Serge Ibaka 8.8
20 DeMar DeRozan 8.5

What makes it the closest to reality? At the very least, you can make a good argument for using WS48 because you have to adjust for minutes played but I think there are other useful stats too.

I did not say that there aren't other useful stats. Everything has its flaws. As you know WP's has an argument against Berri for over weighted rebounding. etc ...

WS48 is good too but what if a player is able to log more minutes? Actual WS's contributed seem to be more closely related to wins.

And if you look at the list, they seem to get it right - but that does not mean its correct.

Certainly a more realistic wins gauge than RPMS?

I disagree. I've never seen any evidence that there's a consistent, big drop off in production as minutes increase. At most, you could look at WS48 and maybe do a tiny adjustment in your mind. Total WS is too confounded with minutes played though.

Minutes played is the actual minutes played. If a player plays 40 per and contributes more than a player who plays 35 minutes then he contributed more.

If want to make an argument that the player who played 35 minutes with the player who played 5 minutes as his backup then perhaps I can see some truth to that but there were need to be a stat that compiles this combined production. Backups very often dont perform at the level of the starter.

How about RPMS? What stat(s) do you go by as your barometer? You are a big fan of David Berri but now have a problem with WP's. Now I know stats are forever evolving but at any given time, you have to choose something. It can't keep changing based on the required outcome (not saying that you do this, but many do).

I give some weight to most of the advanced stats including the nba tracking data. I definitely give win shares a good amount of weight.
So who would you give a higher contract to? A guy who plays great in 20 mpg or an ordinary player with a higher win share total because he's playing 32 mpg?

A player that can average a large number of minutes and still remain efficient. There certainly is going to be diminished returns if you max out a guys minutes. You don't believe that rest improves efficiency?

This is why Durant will probably get the MVP. He has logged 2889 (2nd to Melo) and is at a freaky .306 (wp48). James is at .265.

Take Kawhi Leonard. At 1800 minutes, he is ranked number 12 in the league in WS48 at .192 - Do you max him out before Pual George, Noah or Howard?

Actually, it would be pretty regarding which of those I'd give large contracts to. The diminishing returns argument is hypothesis that I haven't seen support for but I was saying I'd probably do a tiny adjustment in my mind to the WS48 based on minutes played. I've done some searching but haven't looked thoroughly for research on this diminishing returns hypothesis though. If anyone has data, I'd be interested in it.

I created a list of the top WS48's with a min of 1800 minutes (KD and Melo are around 2900 mins):

If you don't buy into the Dim Ret's Hypothesis than we really should have signed DJ Augustin! Also, shocked to see Robin Lobez so high.

RK Player min WS48
1 Kevin Durant 2889 0.306
2 Chris Paul 2067 0.271
3 LeBron James 2751 0.265
4 Kevin Love 2610 0.254
5 Stephen Curry 2690 0.219
6 James Harden 2595 0.216
7 Anthony Davis 2358 0.213
8 Blake Griffin 2758 0.205
9 Dirk Nowitzki 2476 0.197
10 Kyle Lowry 2688 0.193
11 Goran Dragic 2551 0.192
12 Kawhi Leonard 1801 0.192
13 Joakim Noah 2620 0.191
14 DeAndre Jordan 2766 0.188
15 Paul George 2823 0.181
16 Andre Drummond 2458 0.18
17 Robin Lopez 2471 0.175
18 DCousins 2146 0.171
19 Tim Duncan 2069 0.17
20 Serge Ibaka 2475 0.17
21 Carmelo Anthony 2915 0.169
22 Dwight Howard 2310 0.165
23 Chris Bosh 2395 0.163
24 David Lee 2201 0.163
25 Terrence Jones 1908 0.162
26 D.J. Augustin 1815 0.161
27 Damian Lillard 2807 0.158
28 George Hill 2363 0.157
29 David West 2413 0.156
30 Mike Conley 2268 0.154

Bonn1997 wrote:

mreinman wrote:

Bonn1997 wrote:

mreinman wrote:

Bonn1997 wrote:

mreinman wrote:

Bonn1997 wrote:

mreinman wrote:Go to the below page and sort by Win Shares and you would probably get the closest advanced stat to reality (there will always be anomalies with every stat)

http://www.basketball-reference.com/leagues/NBA_2014_advanced.html

Top 20:

Rk Player WS ▾
1 Kevin Durant 18.4
2 LeBron James 15.2
3 Kevin Love 13.8
4 Stephen Curry 12.3
5 Blake Griffin 11.8
6 James Harden 11.7
7 Chris Paul 11.7
8 DeAndre Jordan 10.8
9 Kyle Lowry 10.8
10 Paul George 10.6
11 Anthony Davis 10.5
12 Joakim Noah 10.4
13 Carmelo Anthony 10.3
14 Goran Dragic 10.2
15 Dirk Nowitzki 10.2
16 Andre Drummond 9.2
17 Damian Lillard 9.2
18 Robin Lopez 9
19 Serge Ibaka 8.8
20 DeMar DeRozan 8.5

What makes it the closest to reality? At the very least, you can make a good argument for using WS48 because you have to adjust for minutes played but I think there are other useful stats too.

I did not say that there aren't other useful stats. Everything has its flaws. As you know WP's has an argument against Berri for over weighted rebounding. etc ...

WS48 is good too but what if a player is able to log more minutes? Actual WS's contributed seem to be more closely related to wins.

And if you look at the list, they seem to get it right - but that does not mean its correct.

Certainly a more realistic wins gauge than RPMS?

I disagree. I've never seen any evidence that there's a consistent, big drop off in production as minutes increase. At most, you could look at WS48 and maybe do a tiny adjustment in your mind. Total WS is too confounded with minutes played though.

Minutes played is the actual minutes played. If a player plays 40 per and contributes more than a player who plays 35 minutes then he contributed more.

If want to make an argument that the player who played 35 minutes with the player who played 5 minutes as his backup then perhaps I can see some truth to that but there were need to be a stat that compiles this combined production. Backups very often dont perform at the level of the starter.

How about RPMS? What stat(s) do you go by as your barometer? You are a big fan of David Berri but now have a problem with WP's. Now I know stats are forever evolving but at any given time, you have to choose something. It can't keep changing based on the required outcome (not saying that you do this, but many do).

I give some weight to most of the advanced stats including the nba tracking data. I definitely give win shares a good amount of weight.
So who would you give a higher contract to? A guy who plays great in 20 mpg or an ordinary player with a higher win share total because he's playing 32 mpg?

A player that can average a large number of minutes and still remain efficient. There certainly is going to be diminished returns if you max out a guys minutes. You don't believe that rest improves efficiency?

This is why Durant will probably get the MVP. He has logged 2889 (2nd to Melo) and is at a freaky .306 (wp48). James is at .265.

Take Kawhi Leonard. At 1800 minutes, he is ranked number 12 in the league in WS48 at .192 - Do you max him out before Pual George, Noah or Howard?

Actually, it would be pretty regarding which of those I'd give large contracts to. The diminishing returns argument is hypothesis that I haven't seen support for but I was saying I'd probably do a tiny adjustment in my mind to the WS48 based on minutes played. I've done some searching but haven't looked thoroughly for research on this diminishing returns hypothesis though. If anyone has data, I'd be interested in it.

There are many articles and papers on diminishing returns for rebounding and for scoring. Just google in and you will see.

Among the sabermetrics world they have hotly debated this issue.

mreinman wrote:

Bonn1997 wrote:I just sorted the PFs and saw who was ranked 89th out of the league's 89 PFs!

Haha! That is bloody hilarious!

But we did not trade any players for him so who cares if we paid him 100 million.

I guess that Dk will regret this wonderful new and upcoming advanced metric

since when is having the truth revealed about nba basketball players regrettable? you would have resigned david lee as the gm, right?

mreinman wrote:

dk7th wrote:

mreinman wrote:

Bonn1997 wrote:I just sorted the PFs and saw who was ranked 89th out of the league's 89 PFs!

Haha! That is bloody hilarious!

But we did not trade any players for him so who cares if we paid him 100 million.

I guess that Dk will regret this wonderful new and upcoming advanced metric

since when is having the truth revealed about nba basketball players regrettable? you would have resigned david lee as the gm, right?

The truth? That Amare is ranked 89th out of 89 power forwards? If you are okay with that then its ok.

I am not a GM and would not be good at it. Why does everybody think they'd be or are a great GM? Its hard to be a GM. You need to be far more talented and form than knowledgeable fans.

I was really torn on Lee and was happy it was not me making that decision. Looks like as usual, the knicks made the wrong one.

Amare, I was NOT torn on. I hated that move from the moment I heard about it. How stupid can we be.

Maybe if we gave him 60/3 (his next best and only other offer) I could have swallowed it. I really can't believe people who claim they understand the sport and defend this.

you would have a point if we lived in a vacuum or fantasyland.

but the purchase of amare's services was done in the real world and with very specific extenuating circumstances.

where it went wrong, since time flows forward and causes precede effects, is when carmelo anthony was brought in instead of waiting for an elite point guard to become available.