How can John Hollinger be wrong?

coolcono · Post #1 » by **coolcono** » Tue Oct 6, 2009 8:59 pm

2 top per players? Playoffs?

Esohny · Post #2 » by **Esohny** » Tue Oct 6, 2009 9:06 pm

I'll assume you're joking and we won't get into the many many ways Hollinger could be wrong.

coolcono · Post #3 » by **coolcono** » Tue Oct 6, 2009 9:13 pm

Yes, I am joking. I guess sarcasm is lost in drinking too early :roll:

Wolves2011 · Post #4 » by **Wolves2011** » Tue Oct 6, 2009 9:13 pm

Hollinger didn't say we would make the playoffs....

He said we will probably win 30 to 35 games.

He said if things go well could could be in the playoff hunt.

Flynn plays like a rookie of the year candidate... would help.

If Sessions and/or Love have "break out seasons... would help

Jefferson and Brewer are healthy etc...

He said we are "likely" to win 30 to 35 games.

But more is possible....

Esohny · Post #5 » by **Esohny** » Tue Oct 6, 2009 9:22 pm

coolcono wrote:Yes, I am joking. I guess sarcasm is lost in drinking too early

I don't just mean his predictions about playoffs and so forth. I mean his methods in general. Ah well.

Wolves2011 · Post #6 » by **Wolves2011** » Tue Oct 6, 2009 9:35 pm

One way he can be wrong... his PER stat doesn't account for defense, which is obviously an important part of the game.

PER is purely a stat that measures contributions on offense.

But when you look at the wages of win site.... win scores for Sessions and Love... they are even higher ranked there and BETTER thought of than by Hollinger

[Note: Big Al is less well thought of using the Wages of Win methodology. There are other subtle differences in the way stats are calculated, but defense is an important reason why.]

Look at my analysis using Hollingers stats.

I looked at all NBA teams and identified how many players each team had in the TOP 40 in PER and how many above average players [PER of 15] on each team.

The teams with the most players in the top 40 and had the most players with above average PER were in most cases winning teams. ...

An in the cases where they were NOT, it was explainable.....Washington had many players in both categories but last year they were injured.

Minnesota had three players in the top 40, but Jefferson was injured, Love played limited minutes and Sessions didn't play for us and also played limited minutes. ...

The analysis was in the link below.... towards the bottom of that page...

viewtopic.php?f=22&t=948101

Wolves2011 · Post #7 » by **Wolves2011** » Tue Oct 6, 2009 9:46 pm

Esohny wrote:
coolcono wrote:Yes, I am joking. I guess sarcasm is lost in drinking too early

I don't just mean his predictions about playoffs and so forth. I mean his methods in general. Ah well.

If you look at the web site

NBA.com

which is the NBA's web site there is an "efficiency" stat.

That was also designed by Hollinger and is very similar to PER.

The NBA [and Team management] uses these types of stats to evaluate players.

Danny Ainge and several of the most successful GM's [I think the San Antonio guys, Portland etc.] have utilizes these type of stats for years.

Gamblers have also successfully used them to make money.

If you don't like them, don't use them.

I've found them to be useful.

So has the NBA.....

Esohny · Post #8 » by **Esohny** » Tue Oct 6, 2009 9:54 pm

I'm going to make a model with as many variables as I can think of that creates an output that I've made up for picking stocks, while arbitrarily setting the "average" output level. It's possible that my model will be better than flipping a coin.

Post #9 » by **revprodeji** » Tue Oct 6, 2009 10:47 pm

It kind of reminds me of the team I am coaching this year. it does not matter if you have 2 good players if the rest of them should be in the stands.

Zeitgeister · Post #10 » by **Zeitgeister** » Tue Oct 6, 2009 11:59 pm

Wolves2011 wrote:One way he can be wrong... his PER stat doesn't account for defense, which is obviously an important part of the game.

PER is purely a stat that measures contributions on offense.

But when you look at the wages of win site.... win scores for Sessions and Love... they are even higher ranked there and BETTER thought of than by Hollinger

[Note: Big Al is less well thought of using the Wages of Win methodology. There are other subtle differences in the way stats are calculated, but defense is an important reason why.]

Look at my analysis using Hollingers stats.

I looked at all NBA teams and identified how many players each team had in the TOP 40 in PER and how many above average players [PER of 15] on each team.

The teams with the most players in the top 40 and had the most players with above average PER were in most cases winning teams. ...

An in the cases where they were NOT, it was explainable.....Washington had many players in both categories but last year they were injured.

Minnesota had three players in the top 40, but Jefferson was injured, Love played limited minutes and Sessions didn't play for us and also played limited minutes. ...

The analysis was in the link below.... towards the bottom of that page...

viewtopic.php?f=22&t=948101

Not entirely. PER recognizes blocked shots, steals, and defensive rebounds which are all defensive stats. Yes, I know that there is a lot more to defense than that but PER isn't purely an offensive stat.

Wolves2011 · Post #11 » by **Wolves2011** » Wed Oct 7, 2009 12:24 am

Zeitgeister wrote:
Wolves2011 wrote:One way he can be wrong... his PER stat doesn't account for defense, which is obviously an important part of the game.

PER is purely a stat that measures contributions on offense.

But when you look at the wages of win site.... win scores for Sessions and Love... they are even higher ranked there and BETTER thought of than by Hollinger

[Note: Big Al is less well thought of using the Wages of Win methodology. There are other subtle differences in the way stats are calculated, but defense is an important reason why.]

Look at my analysis using Hollingers stats.

I looked at all NBA teams and identified how many players each team had in the TOP 40 in PER and how many above average players [PER of 15] on each team.

The teams with the most players in the top 40 and had the most players with above average PER were in most cases winning teams. ...

An in the cases where they were NOT, it was explainable.....Washington had many players in both categories but last year they were injured.

Minnesota had three players in the top 40, but Jefferson was injured, Love played limited minutes and Sessions didn't play for us and also played limited minutes. ...

The analysis was in the link below.... towards the bottom of that page...

viewtopic.php?f=22&t=948101

Not entirely. PER recognizes blocked shots, steals, and defensive rebounds which are all defensive stats. Yes, I know that there is a lot more to defense than that but PER isn't purely an offensive stat.

lol... I know, I've calculated PER, Win score etc....but in general its an offensive stat. They use anything in the box score....lol...except plus minus

John Doe [MIN] · Post #12 » by **John Doe [MIN]** » Wed Oct 7, 2009 6:38 am

It's not as if Hollinger sat down and hand picked the order of the players' projected PERs. He designed a formula, applied it to everyone, and reported the results.

What I've noticed is his projections seemed to predict declines in the vast majority of players. They also seemed to lump most of the elite players together right in the 22 range. Jefferson's #2 is a legitimate prediction. He was third last year, and Tim Duncan is very passable.

Love's #4 ranking is a lot sketchier, but he got there for two reasons. He was one of the few players projected to improve upon their previous year's PER, and the PF position is stacked with guys in the 18s, most of which happened to come just a bit short of Love's mark.

Wolves2011 · Post #13 » by **Wolves2011** » Wed Oct 7, 2009 7:25 am

John Doe [MN] wrote:It's not as if Hollinger sat down and hand picked the order of the players' projected PERs. He designed a formula, applied it to everyone, and reported the results.

What I've noticed is his projections seemed to predict declines in the vast majority of players. They also seemed to lump most of the elite players together right in the 22 range. Jefferson's #2 is a legitimate prediction. He was third last year, and Tim Duncan is very passable.

Love's #4 ranking is a lot sketchier, but he got there for two reasons. He was one of the few players projected to improve upon their previous year's PER, and the PF position is stacked with guys in the 18s, most of which happened to come just a bit short of Love's mark.

Most veterans are shown to be going down slightly because historically players "of a certain age" have on average shown declining performance. Thats not surprising, is it?

Love's performance was shown improving because of his performance last season. Virtually constant improvement.

In his "CHAT" today, Hollinger was asked, which player he thought most likely to have a "breakout season" like Devin Harris did last season, Hollinger said Sessions.

In a follow up, he was asked who was the next most likely to have a break out season? He said Love.

This was not in the "forecast". This is "UPSIDE" from what was shown.

As I've said, it will be an interesting season for Wolves fans.