The individual on the team: some statistical notes

[mrhonline]

82games.com keeps a game-by-game record of whether a player's team outscores their opponents while he is on the floor. This ranking, in my opinion, is more important than the overall +/-, which can be affected by great swings in momentum (e.g., Salim's great comeback vs. the Hornets, Marvin's mop-up time heroics, Al's 1st quarter outburst vs. the Spurs). Instead, this ranking measures how consistently a player's overall contribution is.

I used their records to compile a ranking of players based on the percentage of games that the player had a +/- of 0 or greater. The rankings broke down into several groupings:

A. Marvin, Salim 46%

B. Smith, Al 42%

C. JJ, Lue 39%; Zaza 38%

D. Childress 30%

A couple of observations:

1. No one on the Hawks has a 50%+ winning percentage. Sad.

2. Childress, despite playing with the same players as Marvin Williams and Josh Smith, is significantly less successful. This should put to rest comments that Childress helps the Hawks win by doing the "little things." He clearly is not doing that on a consistent basis, even if he has improved his jumpshot. One has no choice but to question his defense...

3. Marvin's effect on the game, for a rookie, appears to be generally positive despite his lack of great stats. As he becomes a more polished player, I still believe he can be a star (like Rashard Lewis or Jamal Mashburn).

4. Josh Smith does make a big difference on the outcome of the game, and shouldn't be sitting on the bench in the 4th quarter.

5. Salim deserves more minutes, especially on the team with the worst record in the NBA.

(BTW, Ivey would fall somewhere between group C and D, based on his limited minutes).

[Diesel]

Where exactly is that stat found?.. What is it called?

[mrhonline]

Here are the win % rankings for the 04-05, for reference:

A. Drobnjak 41%; Diaw 39%

B. Smith 36%; Delk, Chill 34%

C. Googs 22%; Lue, Al 18%; Walker 17%

Observations:

1. It appears that adding JJ, Marvin, Zaza, and Salim has made Al Harrington much more effective (and his newfound range doesn't hurt, either). The same goes for Tyronn Lue.

2. Diaw and Drobnjak were decent role players for the Hawks last year. It's ironic that both players are now gone...

3. Smith has improved his consistency, while Childress appears to have not. Hmm...

4. Walker sucked eggs.

Quote:

Where exactly is that stat found?.. What is it called?

NBA Player won-lost records

Each player has a won-lost record profile under their "player stats." Remember that I've counted a "tie" as a win, whereas they ignore "ties." (Not sure why they consider a "tie" a bad thing, but on these young Hawks, I think it's a great sign):

Salim's W-L record

[Diesel]

First off, it's not surprising that no Hawk is over 50% because we're a losing team.

2nd... If a player is inserted into a game when we are winning, then their percentage should be better. I think it's kinda why Donta has a good number. Didn't Donta play when we were winning against Indy?

So if a coach feels more comfortable playing a player when we're up and on our way to a win, that number increases. However, if a coach feels like he has to keep certain guys on the floor to mount a comeback in a losing situation, his number decreases...

[exodus]

I don't know if I buy that stat but I do like their new stat which measures a players PER as well as that of their opponent. There are fewer variables in that stat, it just involves the player and the guy they are guarding.

[mrhonline]

Quote:

---

First off, it's not surprising that no Hawk is over 50% because we're a losing team.

---

No kidding.

Quote:

---

2nd... If a player is inserted into a game when we are winning, then their percentage should be better. I think it's kinda why Donta has a good number. Didn't Donta play when we were winning against Indy?

So if a coach feels more comfortable playing a player when we're up and on our way to a win, that number increases. However, if a coach feels like he has to keep certain guys on the floor to mount a comeback in a losing situation, his number decreases...

---

???

It's a measure of the points scored and the points give up while the player is on the floor, not the "score" when the player enters the game.

And I intentionally excluded players who are not in the main rotation (i.e., Ivey, Donta, Batista, Thomas, Edwards).

Again, it's a measure of a player's impact on the score of the game on a game-by-game basis, instead of throwing all those numbers into a big heap and averaging them.

[Diesel]

For instance...

If Al and JJ are just going off. I mean Al has 30 and JJ has 35.. Then they take Zaza out and Put me in with Al and JJ still doing what they do. And we go on to win that game.

Your stat will make it seem like I contributed heavy to the win. However, I had nothing to do with the win. Very little.

By the same token.

If We are losing. And Woody says, I want to play the lineup of Al, JJ, and Diesel for a little longer to see if we can mount a comeback... Then your stat would say that I had something to do with the loss. When really you can't quantify how much I contributed to the loss.

All you can say is tht I was on the floor during a losing game...

Being that we have loss so many games, every player BUT the least played will be effected.

Moreover, I can't understand why Salim would have a high %. A few weeks ago, we noticed that we won games and Salim played LESS.

[mrhonline]

Most predictive sports statistics rely on the law of averages, Diesel.

Law of Averages

Quote:

The law of averages is a lay term used to express the view that eventually, everything "evens out." For example: Two very similar people who drive similar cars in similar circumstances over a long period of time will have roughly the same number of accidents. The more children you have, the more likely you will have an equal division of boys and girls. The longer you flip a coin, the more likely the number of heads and tails will equalize.

The scenarios you supplied are also less relevant to Woodson's rigid substitution patterns, which would tend to quicken the "equalization" process (based on their apparent randomness).

Regardless, the differences in winning % between Josh Smith and Josh Childress cannot simply be shrugged aside. It's a good indication that one player is contributing successfully to the team's success, while the other is not so successful.

[mrhonline]

Quote:

I don't know if I buy that stat but I do like their new stat which measures a players PER as well as that of their opponent. There are fewer variables in that stat, it just involves the player and the guy they are guarding.

I like that one as well, exodus. It does seem to favor guards, though:

Roland Ratings

I'd love to know the rationale behind the 2.5 ratio (of PER difference to +/- difference). It really helps out Childress, and hurts Marvin...

Have you seen the Hawks' rankings?

http://www.82games.com/0506/0506ATLH.HTM

Childress seems to equal out his man well, but the team plays poorly with him on the court. The reverse is true for Marvin.

It will be interesting to see if Marvin's PER diff. improves over the course of the season. I ultimately (a year or so from now) think Childress will have to be traded to make way for Marvin (with Josh Smith as a 3/4 off the bench).

[exodus]

Yeah i have seen it. I think 82games.com is good as a general reference but it shouldn't be treated as the Bible, since there are too many variables.

For example after the first month Marion and Nash were both worse than -20 even though they were putting up monster numbers.

[TheNorthCydeRises]

Quote:

---

Yeah i have seen it. I think 82games.com is good as a general reference but it shouldn't be treated as the Bible, since there are too many variables.

For example after the first month Marion and Nash were both worse than -20 even though they were putting up monster numbers.

---

Exactly.

And MrH, what you have to remember about Childress, is that for the first month of the season, he played HORRIBLY. Ever since then, he's aruguably been the 5th best player on the team, behind Al, JJ, Lue, and Smoove.

Make no mistake about it. When the Hawks won 5 of 7 games last month, Childress played very well in those games.

Like with all stats, you just can't look at one, and make a sweeping judgement about a player's overall game or a player's impact on a game. You have to look at everything and weigh it like that.

Take Al for instance. If you looked at him STRICTLY by his scoring and shooting percentages, you could make a legit case for him being on the All-Star team as a reserve. When you follow him every game, like we do, you know better to even think like that.

[mrhonline]

Quote:

And MrH, what you have to remember about Childress, is that for the first month of the season, he played HORRIBLY. Ever since then, he's aruguably been the 5th best player on the team, behind Al, JJ, Lue, and Smoove.

Two things:

1. Don't patronize me. I know the limitations of statistics as well as you.

2. Your statement about Childress is purely subjective. I could argue that he's the worst player on the team and we've accomplished nothing.

What cannot be debated is that in 70% of Childress' games this season, the Hawks have given up more points with him on the floor than they've scored with him on the floor. That's the lowest percentage on the team, and significantly lower than the percentages of similar, younger players in Smith and Marvin, both of whom have more upside.

I'm not saying he can't contribute, I'm saying he hasn't contributed as consistently as is expected of him. The fact that you so easily remember his good games reinforces my statement.

[Squawker24]

The world is littered with statistics, and the average person is bombarded with five statistics a day. Statistics can be misleading and sometimes deliberately distorting. There are three kinds of untruth commonly recognised;

"Lies, damn lies and statistics." - Mark Twain

[mrhonline]

Statistics are like women; mirrors of purest virtue and truth, or like whores to use as one pleases. ~Theodor Billroth

Smoking is one of the leading causes of statistics. ~Fletcher Knebel

The statistics on sanity are that one out of every four Americans is suffering from some form of mental illness. Think of your three best friends. If they're okay, then it's you. ~Rita Mae Brown

It is the mark of a truly intelligent person to be moved by statistics. - George Bernard Shaw

There are Five kinds of lies: Lies, Damned Lies, Statistics, Politicians quoting statistics, and Novelists quoting Politicians on Statistics. ~Stephen Tagg

[Lascar78]

I don't have a girlfriend. I just know a girl who would be really mad if she heard me say that ~ Mitch Hedburg