Hollinger PER - interesting names on the top

[jerrywest]

#3 Power Forward: Josh Smith.

#8 Power Forward: Shelden Williams.

http://insider.espn.go.com/nba/hollinger/rankings?pos=6&action=login&appRedirect=http%3a%2f%2finsider.espn.go.com%2fnba%2fhollinger%2frankings%3fpos%3d6

#14 Center: Horford.

http://insider.espn.go.com/nba/hollinger/rankings?pos=9&action=login&appRedirect=http%3a%2f%2finsider.espn.go.com%2fnba%2fhollinger%2frankings%3fpos%3d9

#2 Point Guard: Acie Law

#7 Point Guard: Jason Terry

http://insider.espn.go.com/nba/hollinger/rankings?pos=1&action=login&appRedirect=http%3a%2f%2finsider.espn.go.com%2fnba%2fhollinger%2frankings%3fpos%3d1

#14 Shooting guard: JJ

http://insider.espn.go.com/nba/hollinger/rankings?pos=2&action=login&appRedirect=http%3a%2f%2finsider.espn.go.com%2fnba%2fhollinger%2frankings%3fpos%3d2

#1 player in the NBA: Kosta Koufas

#10: Acie Law.

http://insider.espn.go.com/nba/hollinger/rankings?&action=login&appRedirect=http%3a%2f%2finsider.espn.go.com%2fnba%2fhollinger%2frankings

[exodus]

Acie has played only 42 minutes on the season so it is a bit early to make any judgements about him.

[macdaddy]

Shelden has been very productive. its good to see. Including Acie is a joke. He's played like 30 total minutes on the season.

[DrReality]

Very happy for Horf. Glad to see him emerge. Wonder how much of Smoove's game playing inside has helped him out? Has to be a correlation!

[Packfill]

Very happy for Horf. Glad to see him emerge. Wonder how much of Smoove's game playing inside has helped him out? Has to be a correlation!

I think it was two years ago where Smoove had the highest per on the team, so I think it has more to do with the fact that he accumulates stats across the board (i.e., blocked shots in addition to points, etc.).

[Packfill]

How good is Maresse Speights? He is currently fifth overall in PER and his numbers look pretty good given his age, experience and playing time.

[Guest]

PER is a flawed stat and its even more flawed when you take a sample size of less than 10 games.

Don't drink the Hollinger kool-aid, there are much better statistics out there than PER. And you shouldn't be looking at those right now, its way too early.

[capstone21]

#3 Power Forward: Josh Smith.

#8 Power Forward: Shelden Williams.

http://insider.espn....kings%3fpos%3d6

#14 Center: Horford.

http://insider.espn....kings%3fpos%3d9

#2 Point Guard: Acie Law

#7 Point Guard: Jason Terry

http://insider.espn....kings%3fpos%3d1

#14 Shooting guard: JJ

http://insider.espn....kings%3fpos%3d2

#1 player in the NBA: Kosta Koufas

#10: Acie Law.

http://insider.espn....nger%2frankings

I am not a fan of the 'per'. I just don't think it is an accurate representation of the impact someone will bring when he is on the floor.

[AHF]

PER is a flawed stat and its even more flawed when you take a sample size of less than 10 games.

Don't drink the Hollinger kool-aid, there are much better statistics out there than PER. And you shouldn't be looking at those right now, its way too early.

What are your preferred stats for measuring efficiency or overall effectiveness as PER is intended to measure?

[Guest]

What are your preferred stats for measuring efficiency or overall effectiveness as PER is intended to measure?

I find win shares to be a better measure. Its from Dave Berri's book Wages of Wins. Hollinger's PER is just a monster of him tinkering his formula enough times to make it look the way he wanted it. Win shares puts a solid foundation behind the measure (regressing statistics on wins, then using the coefficients as weights...) that isn't concerned with who is on top as much as how you measure it.

But even then, its been a while since I've read WOW so I can't recall everything. My mind might have changed but I know for sure I will never endorse PER. Hollinger is a quack.

[AHF]

I find win shares to be a better measure. Its from Dave Berri's book Wages of Wins. Hollinger's PER is just a monster of him tinkering his formula enough times to make it look the way he wanted it. Win shares puts a solid foundation behind the measure (regressing statistics on wins, then using the coefficients as weights...) that isn't concerned with who is on top as much as how you measure it.

But even then, its been a while since I've read WOW so I can't recall everything. My mind might have changed but I know for sure I will never endorse PER. Hollinger is a quack.

Thanks!

For last season, the constrast between win shares and PER looks most intersting for Josh Smith and Marvin Williams/Mike Bibby.

WS

Bibby 6.9

Marvin 6.1

Josh 4.9

PER

Josh 17.2

Bibby 16.3

Marvin 16.0

[Guest]

Thanks!

For last season, the constrast between win shares and PER looks most intersting for Josh Smith and Marvin Williams/Mike Bibby.

WS

Bibby 6.9

Marvin 6.1

Josh 4.9

PER

Josh 17.2

Bibby 16.3

Marvin 16.0

You should always remember what they are trying to describe. PER is trying to fit whatever Hollinger thinks looks the coolest. Win Shares is trying to project how you contribute to wins on your team. Some people's definition of a "good player" may not conform to the definition of a player that helps you win games. They are usually close, but sometimes they are far apart.

My guess for Josh's low WS would be he hurts the team a lot. Just look at turnovers and 3 pt shooting. All win shares are adjusted for position and for his position, Josh is one of the worst in the league for those stats. So all the good he helps out with gets balanced out by the bad. Again, its been a while I have been preoccupied with other things and haven't really gone back and looked into WOW for probably 2 years now so I may actually be butchering the stat (or giving it too much credit).