RealGM

I've been a fan of Dave Berri's "Wages of Wins Journal", for a while now. The guy's system isn't perfect, but it is still a pretty good barometer of who has value, and who does not. It tends to undervalue guys who had immense value in one area (D.J. with his man defense, Ray Allen with his shooting), but overall it is fair. Defense is the hardest area to accurately judge off of the statsm, anyways.

He just did a list, according to his system, ranking all of the Celtics players since the 77-78 season up to now. It is purely off of his system. It isn't as if he is saying "This guy sucks", or "This guy is the bee's knees". Compare his list to the list of the other teams that he did, which I will post at the end. It truly attempts to be fair, statistically.

http://www.wagesofwins.com/Boston77-09.html

http://dberri.wordpress.com/2009/05/28/ ... ince-1977/

The second one is the article, itself with his reasoning and explanations. Remember, time of service and when these guys played counts. Saying "Cowens was awesome!!!!" doesn't really cut it when you are talking about the aging Cowens. Bill Walton is low down, but his WP48 is excellent - he simply didn't play that many games for us. Context, context , context.

I imagine that you can see why I titled this "Be careful what you wish for." The kid is crazy good. Simply awesome, yet people can't seem to wrap their minds around it. He just goes against the grain. If his first three seasons are like this, just imagine what the next 3 are going to be like. Do you actually want to trade this guy? Really? Berri's numbers actually had Rondo neck and neck with Brandon Roy iin their rookie season. He figured that Roy was better, but the numbers clearly pointed to Rondo as being a bigtime player long before that became proven fact.

The system isn't perfect, but you don't end up in that company by accident. Anyways, here are the other two teams. Ask yourself if he's really that far off.

http://www.wagesofwins.com/Lakers77-09.html

http://www.wagesofwins.com/Jazz77-09.html

Any top 10 list that includes Dee Brown on it is clearly less than fool-proof.

floyd wrote:Any top 10 list that includes Dee Brown on it is clearly less than fool-proof.

My Dee Brown jersey says otherwise.

Scalabrine is dead last, seems like a pretty accurate measure

It has to do with his 8 years of service, and 13, 700 minutes of playing time. He was only averaging 3.5 wins produced per season, which is fairly pedestrian. Eddie House, for example, has produced 4.1 according to this measurement.

Once again - pay attention. Do not take things out of context to "prove" a flimsy point. according to this system, by wins produced per season, the top 10 is

1. Bird
2. Garnett
3. Parish
4. Pierce
5. Rondo
6. Ray Allen
7. Maxwell
8. McHale... gotta remember that he was a 6th man for a long time (keeping his minutes down), had a great prime, and then got hurt and was never the same
9. Gary Payton
10. Dave Cowens

Not a bad top 10 if you ask me. I left Washington out as he didn't play an entire season.

Looking beyond the flawed raw data concept.

Cedric Maxwell was a hell of a player and Kevin Garnett is absolutely awesome.

Danny Ainge loves this stat.

A lot of really crappy basketball players have been on this team in the last 21 years.

sully00 wrote:Looking beyond the flawed raw data concept.

Cedric Maxwell was a hell of a player and Kevin Garnett is absolutely awesome.

Danny Ainge loves this stat.

A lot of really crappy basketball players have been on this team in the last 21 years.

I think that another way of looking at it is

1. HOF
2. HOF
3. HOF
4. HOF
5. Rondo
6. HOF
7. Finals MVP
8. HOF
9. HOF
10. HOF

Starting to see my point? Is this guy "expendable"? Especially since Rondo is on his way up, not down.

andy582 wrote:Scalabrine is dead last, seems like a pretty accurate measure

BRIAN SCALABRINE

THE WORST CELTIC.

Looking at his MVP list for 2009 he has rondo as 6th in the league.

There seems to be some selection effects with this statistic though. The 5th most valuable player is Jason Kidd, so according to Dberri, 25 teams should throw whatever they have at Kidd and Dallas to work a deal out. Is Kidd really that good? Number 7 on the list is Troy Murphy, who like Kidd and Rondo has nice stats, but I don't think Larry would turn down say, a Carmelo for Murphy swap. In general point guards seem to be favored. Denver's best player is Billups, in Milwaukee it's Sessions, in Boston Rondo (Should be Pierce or Ray), in Memphis Conley (should be Mayo or Gay), in Toronto Calderon (should be Bosh). It also favors big time rebounders over scorers, which is why Murphy is listed instead of Granger, and Gasol instead of Kobe.

It could be that conventional wisdom is wrong, and the players that most people think are good are really not that great, but that doesn't seem likely.

Rocky5000 wrote:Looking at his MVP list for 2009 he has rondo as 6th in the league.

There seems to be some selection effects with this statistic though. The 5th most valuable player is Jason Kidd, so according to Dberri, 25 teams should throw whatever they have at Kidd and Dallas to work a deal out. Is Kidd really that good? Number 7 on the list is Troy Murphy, who like Kidd and Rondo has nice stats, but I don't think Larry would turn down say, a Carmelo for Murphy swap. In general point guards seem to be favored. Denver's best player is Billups, in Milwaukee it's Sessions, in Boston Rondo (Should be Pierce or Ray), in Memphis Conley (should be Mayo or Gay), in Toronto Calderon (should be Bosh). It also favors big time rebounders over scorers, which is why Murphy is listed instead of Granger, and Gasol instead of Kobe.

It could be that conventional wisdom is wrong, and the players that most people think are good are really not that great, but that doesn't seem likely.

I stated already that I think that the system has its flaws. Then again, Kidd is still a really good player. The Mavs were a hell of a lot better this seasonwhen he was out there then when he wasn't.

http://www.82games.com/0809/0809DAL.HTM

A team leading +12.4 oncourt/offcourt. The guy made a big difference for them on teh scoreboard. I also think that defense isn't well accounted for in this system, but that doesn't effect either Rondo or Kidd. Murphy is a hell of a rebounder and a very efficient offensive player, which is why the system favors him. He is also a poor defender, which the system doesn't account for.

Berri is a propenent of possessions gained for your team being a major determinant of your true value as a player. A guy like Rondo who gets a lot of rebounds and steal,s, but who doesn't turn teh ball eover, is favored by the system. I think that there is a lot of truth to his reasoning, because every possession you gain for your team is another chance to score for you at the expense of that chance for the opponent. If your offense scores 1.1 points per possession, then getting two steals and 7 rebounds means that you, on teh avreage, got your team 9.9 points right there. Turn teh ball over twice, and you hand over the opponent the average points you give up per defensive possession.

While I think that he underrates scoring, I do agree with his assessment that scoring is over valued by "conventional wisdom". I find it interesting that when you look at the Plus/minus stats and teh oncourt/offcourt stats, that you will see that most of teh guys that his stat likes also have a big impact on the scoreboard, when it comes to really helping their teams out and improving them on teh court. I think there is a fire where this smoke is coming from.

Rocky5000 wrote:Looking at his MVP list for 2009 he has rondo as 6th in the league.

There seems to be some selection effects with this statistic though. The 5th most valuable player is Jason Kidd, so according to Dberri, 25 teams should throw whatever they have at Kidd and Dallas to work a deal out. Is Kidd really that good? Number 7 on the list is Troy Murphy, who like Kidd and Rondo has nice stats, but I don't think Larry would turn down say, a Carmelo for Murphy swap. In general point guards seem to be favored. Denver's best player is Billups, in Milwaukee it's Sessions, in Boston Rondo (Should be Pierce or Ray), in Memphis Conley (should be Mayo or Gay), in Toronto Calderon (should be Bosh). It also favors big time rebounders over scorers, which is why Murphy is listed instead of Granger, and Gasol instead of Kobe.

It could be that conventional wisdom is wrong, and the players that most people think are good are really not that great, but that doesn't seem likely.

The conventional wisdom you seem to be referencing is that scoring is of supreme importance in basketball. First off, a disclaimer: I think Wages of Wins is incomplete due to it's lack of comprehensive defensive measurements and oversight of distinct roles and match-ups. Win score might love Rondo, Jamario Moon, Shane Battier, Troy Murphy and Joel Pryzbilla would be beloved by WoW, but I wouldn't count on that starting line-up making it to the finals. Still, WoW has a nice track record for predicting the success of teams, FA signing and trades., and I think they have exposed some of the chinks in the "scoring is most important" item of conventional wisdom.

Guys like Mike James, Jamal Crawford, Eddy Curry, Stephon Marbury, Jerry Stackhouse and Allen Iverson are now widely devalued, but WoW was usually ahead of the game in pronouncing most of them relatively unproductive players. Statistically, such players do very little well except accumulate points, in many cases not very efficiently. But that didn't stop fans from drooling over acquiring them or GMs from shelling out huge gobs of money for their scoring ability.

Slartibartfast wrote:

Rocky5000 wrote:Looking at his MVP list for 2009 he has rondo as 6th in the league.

There seems to be some selection effects with this statistic though. The 5th most valuable player is Jason Kidd, so according to Dberri, 25 teams should throw whatever they have at Kidd and Dallas to work a deal out. Is Kidd really that good? Number 7 on the list is Troy Murphy, who like Kidd and Rondo has nice stats, but I don't think Larry would turn down say, a Carmelo for Murphy swap. In general point guards seem to be favored. Denver's best player is Billups, in Milwaukee it's Sessions, in Boston Rondo (Should be Pierce or Ray), in Memphis Conley (should be Mayo or Gay), in Toronto Calderon (should be Bosh). It also favors big time rebounders over scorers, which is why Murphy is listed instead of Granger, and Gasol instead of Kobe.

It could be that conventional wisdom is wrong, and the players that most people think are good are really not that great, but that doesn't seem likely.

The conventional wisdom you seem to be referencing is that scoring is of supreme importance in basketball. First off, a disclaimer: I think Wages of Wins is incomplete due to it's lack of comprehensive defensive measurements and oversight of distinct roles and match-ups. Win score might love Rondo, Jamario Moon, Shane Battier, Troy Murphy and Joel Pryzbilla would be beloved by WoW, but I wouldn't count on that starting line-up making it to the finals. Still, WoW has a nice track record for predicting the success of teams, FA signing and trades., and I think they have exposed some of the chinks in the "scoring is most important" item of conventional wisdom.

Guys like Mike James, Jamal Crawford, Eddy Curry, Stephon Marbury, Jerry Stackhouse and Allen Iverson are now widely devalued, but WoW was usually ahead of the game in pronouncing most of them relatively unproductive players. Statistically, such players do very little well except accumulate points, in many cases not very efficiently. But that didn't stop fans from drooling over acquiring them or GMs from shelling out huge gobs of money for their scoring ability.

Very well put, Slartibartfast. I think a good example of how roles that are filled impacts a player's value is Scalabrine this season. He still got hammered in teh ratings, but actually had value for us in his role. His defense wasn't factored into this, nor were his picks, box outs, or his decent three point shooting. He didn't have good stats, but he was effective. I don't think that Berri is really saying that these numbers are solid. He just thinks that they are good indicators of value. Yur lineup is a good example, as would be a lineup consisted of five tewenty point scorers. There are only so many shots to go around, so where is teh value of having 5 of them, and what do those guys do beyond score? You need some dirty work guys to make things work.

Greendreamer, you are embarassing yourself.

Let me get this straight. Rondo, who just completed his third year in the NBA is a more valuable player than Bob Cousy?

Would you trade Rondo for an in prime Cousy? You would get laughed out of the building.

Here is what I learned in stats 101 in college. When evaluating statistics and models, you have to have a measure of skepticism at the results. If the results of the statistical analysis don't past muster (the smell test) then you have to explore how your data was synthesized.

You can't have cognitive dissonance between yourreality and the stats. Either Rondo is indeed superior to Kevin McHale and Bob Cousy, or the model isn't useful

The fact Rondo is in the top 10 is defacto silly. It is not plausible.

That statistical analysis doesn't pass the common sense test. It measures something, but what?

Flawed model.

Rondo more valuable than Cousy....

GreenDreamer wrote:

sully00 wrote:Looking beyond the flawed raw data concept.

Cedric Maxwell was a hell of a player and Kevin Garnett is absolutely awesome.

Danny Ainge loves this stat.

A lot of really crappy basketball players have been on this team in the last 21 years.

I think that another way of looking at it is

1. HOF
2. HOF
3. HOF
4. HOF
5. Rondo
6. HOF
7. Finals MVP
8. HOF
9. HOF
10. HOF

Starting to see my point? Is this guy "expendable"? Especially since Rondo is on his way up, not down.

First off you are preaching to the choir on Rondo I wanted to use the lottery pick to get him. Rondo is not expendable he also is not a max salary player nor is untouchable in a deal. He is a perfect pg for this team I am concerned about how he fits with the next team. I said when we picked him that he will be the next Mo Cheeks and he is well on his way. But Mo Cheeks as the best player on his team, while still a very good player wasn't a very good team.

In the end if I am picking between Rondo and Wade that is no choice at all. I also see no need to sign a guy to an extension a year early so that it can only tie your hands going forward.

Now the Tyreke Evans/Jason Thompson or Mike Conley/Rudy Gay is just doubling your money. I think Evans and Conley are equal prospects to Rondo and Thompson or Gay just put it over the top, there are risks in everything but while Rondo may average a triple double and be a great, great player I think you will always need someone to carry the scoring load.

Also, Rondo's effect is infalted by the fact in two of his three years the Celtics won 60 something games.

After 10 years, where will rondo be. It would seem Rondo hasn't had enough longevity to accurately affect this model.

That said: You don't have to go to the spreadsheet to tell me how good rondo is.I watched him, and I see his effect on the game. Rondo isn't a primary option, is a liability at times on offense and is inconsistent.

On the plus side, he is devastatingly quick, a natural feel for the game and an excellent rebounder for his size (which this model extremely over represents the value of). Rondo is a terrific player, but an all-time great? Not a chance. He doesn't have that talent.

IF Rondo was so great, how come the Celtics were the worst team in the league his rookie year? What changed inbetween? Hmm. This says more about Paul Pierce and KGs ability than Rondo's.

KG makes other players better.

Slartibartfast wrote:

Rocky5000 wrote:Looking at his MVP list for 2009 he has rondo as 6th in the league.

There seems to be some selection effects with this statistic though. The 5th most valuable player is Jason Kidd, so according to Dberri, 25 teams should throw whatever they have at Kidd and Dallas to work a deal out. Is Kidd really that good? Number 7 on the list is Troy Murphy, who like Kidd and Rondo has nice stats, but I don't think Larry would turn down say, a Carmelo for Murphy swap. In general point guards seem to be favored. Denver's best player is Billups, in Milwaukee it's Sessions, in Boston Rondo (Should be Pierce or Ray), in Memphis Conley (should be Mayo or Gay), in Toronto Calderon (should be Bosh). It also favors big time rebounders over scorers, which is why Murphy is listed instead of Granger, and Gasol instead of Kobe.

It could be that conventional wisdom is wrong, and the players that most people think are good are really not that great, but that doesn't seem likely.

The conventional wisdom you seem to be referencing is that scoring is of supreme importance in basketball. First off, a disclaimer: I think Wages of Wins is incomplete due to it's lack of comprehensive defensive measurements and oversight of distinct roles and match-ups. Win score might love Rondo, Jamario Moon, Shane Battier, Troy Murphy and Joel Pryzbilla would be beloved by WoW, but I wouldn't count on that starting line-up making it to the finals. Still, WoW has a nice track record for predicting the success of teams, FA signing and trades., and I think they have exposed some of the chinks in the "scoring is most important" item of conventional wisdom.

Guys like Mike James, Jamal Crawford, Eddy Curry, Stephon Marbury, Jerry Stackhouse and Allen Iverson are now widely devalued, but WoW was usually ahead of the game in pronouncing most of them relatively unproductive players. Statistically, such players do very little well except accumulate points, in many cases not very efficiently. But that didn't stop fans from drooling over acquiring them or GMs from shelling out huge gobs of money for their scoring ability.

Excuse me sir? Scoring is of supreme importance. It is the object of the game. Every other statistic in basketball is related to scoring.

Scoring efficiency is the holy grail. You want a guy who can score at a high clip and efficiently (the only good high volume shooter in the NBA is Allen Iverson, who is unique in his own right).

What world are these people in? This isn't baseball where HRs, AVG. and RBIs aren't great measure of a player. In baseball, these advanced sabermatics are useful because the conditions that they are created under are relatively stable.

In hoops, you are so dependent on who is on the floor with you, these advanced models are all flawed because they don't take that into account. There is no remedy for putting up stats on a bad team.

In baseball, you add A-Rod to your team, even as the best player in the game, the effect isn't as pronounced if you added Michael Jordan to your team. IN the NBA, one player is the difference between 20 wins and 60 wins.

JFS. First of all, the list in question only rated Celtics from 1977 onwards, so Cooz wasn't a part of the discussion. Maybe Tiny Archibald, Kenny Anderson and Gary Payton could make better reference "points" in that list.

Sully. You really think Conley and Gay are equal prospects to Rondo. What makes Gay better than Al Harrington or Cuttino Mobley? He's a scorer of average efficiency who doesn't really do anything else particularly well.

I would choose Gay over Rondo. If you put Gay in the current Boston environment, he would thrive.

Gay is a superior athlete to Al Harrington, and more of a true wing player while Harrington is a hybrid. Cuttino Mobley is a poor comp also because Mobley is a combo guard.

Your also devaluing scoring ability way too much. Most rondo fanatics disregard this ability. They do so at their peril. It's the No.1 ability needed in an NBA player.

The ineffciencies in Gay's game can be alleviated by improved players surrounding him. In addition, when you combine his skill set with his athleticism and finishing ability,he has the makings of an elite scoring wing in excess of 20 ppg. Right now,straight up talent-wise, I choose Gay.

Rondo is a product of the big 3. Yes, he is a good player, but because of the ability of the three HOFs around him, he is able to produce at a higher level. He wouldn't nearly be the player he is now on the Grizzlies.

If we traded Rondo foe Gay srraight up,and then signed Andre Miller to a deal...it would be better now and in the future.

This is not to mention Gay's potential has only scratched the surface. That 18.9 ppg. on 45 percent shooting was done easy. His talent, at the same age as rondo, is off the charts. He averaged 20.1 a year ago. Kids an elite scorer playing with trash in Memphis.

Gay >Rondo.

Like both, but Gay is being severely underrated here.

If Gay is more athletic than Harrington, what does it matter if he doesn't do anything better than him? He's too small and too poor of a rebounder to play effectively at the 4, and he's too poor a playmaker to be creating for himself off the wing. Last season he averaged almost a full turnover more than assists. Essentially all he's good at is scoring, and he's not that great at that. He shoots around 45% from the field, 35% from 3 and is around average at drawing fouls. While those numbers aren't bad, they're not exactly singing his praises, especially when the rest of his statline is playing a baseline of mediocrity. He could get better, he's still young, but Rondo is just as young and has accomplished far more. Rondo is a stellar rebounder, a great passer, and a seriously flawed but still downright efficient scorer. And he steals the ball more and turns the ball over less. Those aren't just areas of potential. Rondo is demonstrably better than Gay in almost every field but distance shooting and FT shooting.

And since when has playing with "trash" effected a player's scoring average negatively. Elite scorers tend to do even more scoring on bad teams. In fact, bad teams allow for gunners like Mike James, Ricky Davis, Cat Mobley, Zach Randolph, Jamal Crawford, Corey Maggette, Big Dog Robinson, Marbury and countless others to throw up ridiculous statlines without competition from better scorers for shots.

Meanwhile elite scorers like Pierce, Allen and Garnett see precipitous drops in their scoring averages as soon as they join a good team.

jfs1000d wrote:Also, Rondo's effect is infalted by the fact in two of his three years the Celtics won 60 something games.

After 10 years, where will rondo be. It would seem Rondo hasn't had enough longevity to accurately affect this model.

That said: You don't have to go to the spreadsheet to tell me how good rondo is.I watched him, and I see his effect on the game. Rondo isn't a primary option, is a liability at times on offense and is inconsistent.

On the plus side, he is devastatingly quick, a natural feel for the game and an excellent rebounder for his size (which this model extremely over represents the value of). Rondo is a terrific player, but an all-time great? Not a chance. He doesn't have that talent.

IF Rondo was so great, how come the Celtics were the worst team in the league his rookie year? What changed inbetween? Hmm. This says more about Paul Pierce and KGs ability than Rondo's.

KG makes other players better.

How the Celtics were so bad? A massive wave of injuries, intentional losing at the end of the season, and a poor identification by Doc of who was good and what they were good at.

Here's the 82games.com stats for that season

http://www.82games.com/0607/0607BOS.HTM

Do you see that +8.3 next to Rondo there? That was larger than Paul's. Here are the raw plus/minus stats

http://www.82games.com/0607/0607BOS1.HTM

Notice the 35 and 34 W/L next to Rondo's. That means that in 35 games that he played they outscored the opposition with him out there, and 34 they were outscored. So the stats were spread out and not merely him having 1 good game and then 2 bad ones. That they were aterrible team wasn't his fault. He didn't control who he played with or when he played. He just made them a lot better when he was out there. The problems were when he wasn't out there. Thattakes care of that part.

There is NO inflation because of 60 win records. These winsaxores do not "divide up the wins" between the players. They don't take a team's record into account at all, until the end. They seek to indentify what makes a team win, and who contributes to those wins. A players INDIVIDUAL stats are analysed, and a win score rating is assessed to him based on those stats. They then add up all of the winscores from the different players on the team to see how wel that it worked. It isn't perfect, but it usually comes pretty close.

http://www.wagesofwins.com/Boston0507.html

I'm figuring that the rash of injuries on the team messed up his figures a little ehere, because the error usually isn't so high. Afterall, he's trying to find a system which can be applied to any team as a baseline, and most teams don't have the number of injuries that we had, especially to the core. Note that Rondo is ranked just after Al and Paul.

Now eher are the paged for the enxt two season.s

http://www.wagesofwins.com/Boston0708.html

he hasn't realeased one for this past season which is full, so I'll put his one through 59 games in

http://www.wagesofwins.com/Boston590809.html

Ofcouse this misses the bulk of Garnett's injury woes, but as you can see that his method was working pretty well overall. He does each guy individually, based on his numbers, and then adds up the winscores and the result is very close to what happened in reality.

The man simply evaluates players in a fundamentally different way than has been the norm up to this point.

By the way, Rudy Gay is a complete slug according to Berri.

http://www.wagesofwins.com/Memphis620809.html

but at keast Conley's a nice player.