OT: The more things change ... (Sergio in Sacto)

Yawn....

All that and he can't get his ass off the bench and in the game for a team with a 0.366 winning percentage.

BNM

 

I have no idea why that comment was directed at me, becuase that's EXACTLY the point I've been making regarding the +/- stat for the last 2 - 3 weeks.

Basketball is a team sport. +/- is a team stat. You simply cannot remove the impact of the team on the stat.

If you put a good, or even great, player on a really bad team, they will have a negative +/-.

Basketball = team sport
+/- = team stat
bad team = bad +/-

Conversely, you could have put an 8-year old child on the 1992 Dream Team and they would have had a highly positive +/-

Basketball = team sport
+/- = team stat
great team = great +/-

Anyone misusing +/- as an individual stat is ignoring the team aspect of the game and the team's influence on a player's +/- rating. You simply cannot separate the two. A good player on a bad team will have a worse +/- than a bad player on a good team. Because, any way you slice it, basketball is a team sport and +/- is a team stat.

BNM

 

Well fuck! Why don't his coaches recognize his brilliance? I tell you, it's a fucking conspiracy meant to keep him down. Don't they know they have a future hall of famer on their roster?

 

Of course not, but the same is true for scoring, assists, blocks and the like. They, too, are impacted by the team you are on and the opponents you play.

+/- is an individual stat in the context of the team, it is not a team stat - otherwise every player on the team will have the same +/-

+/- IS an individual stat. Same as PER, same as scoring, same as rebounds. All of those stats are influenced by the team and the opponents as well...

Just like scoring, rebounds, assists... You put Adam Morrison on Gonzaga and he is a great scorer, you put him on the Lakers and he is a ticking expiring contract...

+/- is not a team stat. It is influenced by team members, just like all other individual stats. Pretending it is not is absurd. What can be said is that +/- has more volatility to team influence in small samples, but it is an individual stat in the context we discuss it.

 

Exactly.

Repped.

 

The difference between +/- and PER or scoring average is that +/- is measuring team performance and hoping to glean something from the correlation of a specific player being on the court during that team performance being measured.

Measures like PER and scoring averages, while affected by players around them, measure the individual's productivity.

So while there is teammate/opponent influence on all stats, there's also a significant qualitative difference between indirect stats like +/- (which measure team performance while a player is on the court) and direct statistics like PER, PPG, etc (which measure an individual's production).

 

why are we still talking about sergio?

he will be out of the nba by 2012 playing in europe as a backup

 

Of course, it is more a measure of fit, rather than individual talent. But, it is just as important in the grand scheme of winning a basketball game as it is not. It should be noted, however, that it provides a glimpse into the world of defense which most other "mostly individual" stats do not measure.

Within the system. Again, if a player is relegated to a support role because the team has a better offensive player (or one that the coaches think of as better) - the PER formula will not really provide a very good measuring stick of the player's real talent - which is exactly what BNM was arguing against +/-

I am not buying the "qualitative" argument one bit.

In a very simple, obvious way, you are right. +/- is not an easy to decipher and "verify by observation" statistic, where scoring, rebounding or a compounded version of these (PER) is easier to "see". But, just because you do not see it, does not render it less important. The obvious examples of "good stats on a bad team" are a shining beacon on the obvious issues with these very same "easy to decipher" individual stats.

I go back to the claim that basketball is a team sport, and the value of a player to the team goes beyond his easy to measure simple stats. If you only look for the answers under the street light because it's easy to see there - you should not be surprised to find out that the answers you seek are not found...

You put a great engine on a chassis that is not built to handle it - and the result is a crap automobile. +/-, in large sample sizes, and win% even more so (because it reduces the volatility of the outliers) is a way that gives you a long-term view on the fit and value of the individual parts in the team. Ignoring them, or waving your hand at them with the claim that they are "team stats" is just absurd.

 

That wasn't what I was saying at all. I had a very simple, obvious, correct point that you missed: +/- is measuring team performance (score differential achieved by the team) while the player is on the floor.

Measures like PER or PPG or RPG measure individual performance.

These are clearly qualitatively different things. +/- has an added layer of abstraction, trying to see how individual performance is reflected in team performance. That doesn't make it useless, but it does make it useless in small sample sizes, since an individual's performance is not likely to be well-reflected by team performance in short stretches, where luck or differing caliber of teammates can easily overwhelm the effect of the player you're interested in.

I don't think +/- is a team stat. It's an indirect individual stat and, as such, requires much, much more sample size to generate any meaningful conclusions than do direct individual stats. And I question the value of raw +/- over any sample size. A player can easily be on teams with bad teammates around him over multiple seasons. Adjusted +/- metrics, that attempt to adjust for teammates, I think can be quite useful with multiple seasons of data.

 

The individual stats you mentioned are much more dependent on individual talent, size, athleticism, etc. than +/- is. +/- is MUCH more dependent on the the quality of a player's teammates. It's not even close to the same thing. An individual's scoring is more impacted by his own shooting ability, athleticism, ability to get create high percentage scoring opporttunities than it is by the play of a his teammates.

Sure, a player's individual stats are impacted to a small degree by his teammates, but if you put Michael Jordan on the Clippers during his prime he still wins a ton of scoring titles. Ditto for Dennis Rodman. He would have led the league in rebounding no matter who he played for in his prime (he led the league in rebounding multiple times for three different teams).

Wrong. Every player on the team does get the exact same +/- for the time period they are on the floor together. Every players +/- is highly influenced by the play of his teammates and his opponents. Bad players are rewarded by playing with better teammates and good players are penalized by playing with subpar teammates. For example:

The same five man unit plays a six minute stretch together. Player A plays great basketball. He scores 14 points on 4-4 from the field and 6-6 from the line as he draws three fouls on the opposition, getting their big men in foul trouble and his team shooting the penalty early. He plays lockdown defense and holds his man scoreless. He dishes out three assists, pulls down four rebounds, has a blocked shot, two steals and zero turnovers.

His teammate, Player B stinks up the gym. He shoots 0-6, turns the ball over four times, has zero assists, zero steals, zero blocked shots and zero rebounds. On the other end, he gets lit up by his man for 12 points. So, at the end of that six minute stretch, both players, along with their three other teammates, will have the exact same +/-, because +/- is a team stat. No individual +/- ratings are awarded. The +/- all players get is indentical and not a direct reflection of their individual contributions.

If they are ahead, Player B, in spite of his truly awful play, is rewarded with the exact same positive +/- as Player A. If they are behind, Player A, in spite of doing everything possible to help his team win, gets a negative +/- thanks to teammate who can't shoot for shit, fumbles the ball away and plays non-existent defense.

That, is exactly the problem with the way people in this forum misuse the +/- stat.

No, it's not. If five players play together for six minutes they will ALL have the exact same +/- for that six minute stretch. During those same six minutes, they will not all have exactly the same PER, scoring and rebounding stats. If the team grabs a total of 10 rebounds (6 for Player A, 3 fopr Player B, 1 for Player C and zero for players D and E), they don't all get credit for having 10 rebounds. They each get credit for the rebounds they grabbed. Same for scoring. +/- is TEAM stat and is based on TEAM performance, not individual performance, during those six minutes. PER, scoring and rebounding measure each players contributions in those specific areas. Totally, totally different.

To a MUCH, MUCH, MUCH lesser degree than +/-. It's not even close to the same.

Are you saying Adam Morrison's inability to score in the NBA is the fault of his inferior Laker teammates? I don't think that's what you meant, but that's how it came out. Adam Morrison doesn't suck because he's on a bad team. Morrison's inability to score in the NBA is nobody's fault but his own. He simply is not athletic enough to get his shot off against the athletes he plays against in the NBA. He happens to already play on one of the best teams in the league, but he still sucks. Surround him with even better players, and guess what, he'll still suck.

Nope. Wrong again. It is a team stat. All five players who are on the court at the same time get the exact same +/- reguardless of how much or how little they contribute. If it was an individual stat, they would all get different +/- ratings based on their individual contributions.

Over the course of a game, a season or a career, it is simply the summation of the player's teams' +/- while they were on the court and not a direct reflection of a players individual contributions to his teams' success.

BNM

 

Exactly! Put a good player on a bad team and he'll have a negative +/-. Doesn't mean he's a bad player. It just means he's playing on a bad team.

Put that exact same player on a team where he's surrounded by great players and suddenly his +/- skyrockets. The rapid increase in his +/- is not a function of his individual play. He didn't instantaneously go from being a "bad" player (negative +/-) to a "great" one (highly positive +/-). He simply went from a bad team to a great team - and his +/- rating is a direct relfection of the quality of his teammates, not the quality of his own play. Because... +/- is a team stat, not an individual stat.

BNM

 

So... if a player has the BEST +/- on the team, he miraculously manages to luck into being on the court when the team is playing well better than any other player. Kind of a lucky charm, I guess. You'd think, given how sucky the Kings are, they might want more of the lucky charm rather than less. But then, there are things like "contracts" and "showcasing to offload" and stuff.

 

Indeed he does - currently 41% for the season. Of course, Sergio's no slouch - 36% for the season. Hey, maybe he's not one-dimensional either!

True.

Not as good a clip as Sergio (2.4 to 2.0 steals per 48), but not bad.

This is a classic, even for you, Ed. I love how you come up with ridiculous quantifications. "Nearly half"? Why only "nearly half"? And how much of a game is passing, do tell. 27% perhaps?

Look, to me it boils down to this: Kevin Martin is a shorter version of last year's Kevin Durant. He does the flashy stuff that people notice (putting the ball in the basket). But if you've played basketball much, you know that that doesn't mean all that much. That just means he wants the glamor job. But does he help his teammates get easier scores? Does he work hard on D? Or is he that jackass who cherrypicks and never helps out?

Meanwhile, if you've ever played with a guy like Sergio (whose gifts, however "one-dimensional" are a fuck of a lot rarer than Martin's), it's fun for the whole team. Everybody gets to score. Everybody feels like they're actually part of the team and valued, and they work harder.

Now, I'm not saying that Sergio is a great player. And maybe he doesn't deserve big minutes in the NBA. But to say that he deserves DNP-CDs on the Kings is laughable. Yes, his defense is bad. But it's probably better than half their fucking team. Look at Hawes. Look at Udrih. Look at Martin. In fact, Evans, Brockman and possibly Nocioni (although injuries have taken a lot out of him) are the only players obviously better than him. And what message are you sending if you bench the player who, for whatever reason, and for whatever sample-size, has been helping your team more than any other while he's been on the court this season?

 

So maybe, just maybe, the way to evaluate +/- is not to look at absolute values, but relative to a player's teammates? Or would that just make too much sense?

You make it sound like players operate in a vacuum and have nothing to do with other players or are not adaptable. And it doesn't seem to occur to you that the converse is true: that a player's PER is very much dependent on how a coach uses him, and to what extent his teammates help him get easy shots or feed him the ball.

 

Rasta... are you really claiming that Martin isn't obviously better then Sergio but Brockman is???

STOMP

 

So what? Basketball games are not won only on individual talent.

Which is why we need more than a small sample size. If basketball games were played by rigid 5 men groups, the stat would have very little significance. Since it does not...

Again, anecdotal evidence and statistics do not mix well. I understand pretty well the idea of the inter-dependence between players and their +/- with their team-mates in each INDIVIDUAL segment. The idea about statistics is to take a lot of these small instances with many "random" events (with random in this sense being who their team-mates and opponents were at that specific segment) - and come to a quantitative value for their contributions using a small enough margin of error. Of course, if you look at the formulas to calculate the margin of error (or confidence in the mean) you will see that for Gausian formulas (which we can assume it is because of the unpredictable combinations that basketball games are played at) - it is directly related to the sample size.

No, it is not. The problem is your inability to accept that math has a way to deal with confidence levels based on individual "random" events.

Right. Small sample size, but right.

Of course not, but the exact same PER over a stretch of a small segment of basketball game is worthless information. Again with the anecdotal evidence.

No, I am just showing you that scoring is just as volatile to the team-mates and opponents as +/- - I am telling you that all these individual stats you present are just as situation dependent... negating the crux of your argument.

If you have a great scorer that never gets the ball from his team-mates, his scoring will suffer, if you have a great set-up man with players that can not score - his assists will suffer, etc, etc, etc...

 

Huh?

Oh, right! Yes, yes I am. Among other things.

 

By that logic, Anderson Varejao is a better basketball player than Lebron James.

As I have said repeatedly regarding +/-... bad (or less good) players benefit from sharing the court with great players. Good (or great) players see their +/- negatively impacted by sharing the court with lesser players.

A player's +/- is also impacted by the five opponents on the court when he's in the game. That's why it's not unusual for a bench player to have a higher +/- than an obviously superior starter. The starter plays the bulk of his minutes against the other teams' best players. A bench player doesn't.

In Sergio's specific case, the sample size is very small, but he rarely plays against an opponents best players and lately, when he does play it's for six or seven minutes of garbage time. So, it's not suprising at all that he would have a better +/- than the Kings starters who play the bulk of their minutes against their opponents' best players.

This is directly reflected in the Kings +/- stats. Their five starters have the lowest +/- on the team. Above that is their second unit, and above that are their scrubs and garbage time players. No suprise there. This is typical for a team with a losing record. And, once again illustrates the fallacy of the +/- stat. A good player on a bad team will have a bad +/-.

BNM

 

No, that makes no sense at all. It's a function of who you play with, who you play against and how many minutes you play. Is Anderson Varejao REALLY better than Lebron James? Hlee no he's not, but he plays on the same team and has a higher +/-. In fact, Anderson Varejao has the highest +/- in the entire NBA. Does that mean he's the best player in the league? No, it means he plays on the best team in the league and gets to play a lot of minutes next to the best players on that team, including one of the best players in the league. His +/- benefits from playing on good team next to a great player.

No, I have never claimed players operate in the vaccuum. As I said many, many times, basketball is a team sport and +/- is a team stat. How do you get "vacuum" out of that?

Yes, a player's individual stats can be impacted by the quality of their teammates and their coach - to a limited degreee. They are still individual stats and more greatly influenced by the players own individual talents and skills. You can get Dan Gadzuric all the easy shots in the world, but he's still Dan Gadzuric and he'll still brick a large percentage of them.

BNM

 

sergio has played 23% of the time for the kings. he has played 8% of their "clutch minutes"(4th quarter or overtime, less than 5 minutes left, neither team ahead by more than 5 points).

that seems to indicate that sergio doesn't play a whole lot of minutes that matter(which is what you would expect from the 3rd string pg). playing a lot of your minutes during garbage time does weird things to +/-.