Meaning of Statistics

[legacyinthemakin89c]

Hey I'm doing a research paper for my freshman college writing class and its on basketball statistics.

I'm thinking of writing about what common statistical similarities are there between winning teams. Like when you guys look at stats, what jumps out as you as being essential to win a game/championship?

Any stats websites, general opinions or advice you guys could give me on this would be amazing since this is 40% of my overall grade.

[Sroek]

Do your own homework.

[Schad]

Hey I'm doing a research paper for my freshman college writing class and its on basketball statistics.

I'm thinking of writing about what common statistical similarities are there between winning teams. Like when you guys look at stats, what jumps out as you as being essential to win a game/championship?

Any stats websites, general opinions or advice you guys could give me on this would be amazing since this is 40% of my overall grade.

www.basketball-reference.com. Stats for every team in NBA history, and you can judge them against the league averages for the year.

[Mags FTW]

Like when you guys look at stats, what jumps out as you as being essential to win a game/championship?

Boston, Houston, Detroit, Dallas, San Antonio, and LA Lakers are your league leaders in Opponent FG%. That's usually a good place to start.

[legacyinthemakin89c]

Do your own homework.

This is doing my homework, why don't you calm down and if you don't have anything to contribute, how about you just don't post?

Its a legit question even if I didn't have a paper on it. What statistics make a great team?

[BrooklynBulls]

I think looking at defensive and offensive efficiency (points scored per 100 possessions) is the simplest way of cutting through the BS.

It literally tells you how good the defense and offense is, with no confounding factors.

Another statistic to look into is rebounding differential. The amount of rebounds you get vs the other team is an important statistic in determining the total possessions, and therefore shots, available.

The final statistic you should look at is TO's vs TO's forced. This is a completely overlooked factor in determining total possessions, which again, separates the normal teams from the great ones. Teams like San Antonio keep the ball, almost never throwing it away. Specifically Detroit, is a team I know never, ever, loses the TO battle. If they do, they lose. Detroit doesn't have a high-powered offense, instead using smart defense, rebounding, and controlling the tempo, and turnovers in order to maximize possessions while minimizing the opposing team's offense.

I think you can get all of these stats on knickerblogger.net, 82games.com, basketball-reference.com, or just NBA.com.

Hope that helps.

[miller31time]

I think looking at defensive and offensive efficiency (points scored per 100 possessions) is the simplest way of cutting through the BS.

It literally tells you how good the defense and offense is, with no confounding factors.

Another statistic to look into is rebounding differential. The amount of rebounds you get vs the other team is an important statistic in determining the total possessions, and therefore shots, available.

The final statistic you should look at is TO's vs TO's forced. This is a completely overlooked factor in determining total possessions, which again, separates the normal teams from the great ones. Teams like San Antonio keep the ball, almost never throwing it away. Specifically Detroit, is a team I know never, ever, loses the TO battle. If they do, they lose. Detroit doesn't have a high-powered offense, instead using smart defense, rebounding, and controlling the tempo, and turnovers in order to maximize possessions while minimizing the opposing team's offense.

I think you can get all of these stats on knickerblogger.net, 82games.com, basketball-reference.com, or just NBA.com.

Hope that helps.

Took the words right out of my mouth.

[Tim Lehrbach]

what jumps out as you as being essential to win a game/championship?

Google "Dean Oliver".

And follow others' advice and check out basketball-reference.com.

[tkb]

I looked up defensive rebounding in the playoffs a year or two ago. Looked up all games the past 10 years and over 70% of the time the team who had the most defensive rebounds also won the game.

A couple sites I use a lot:
82games.com
basketball-reference.com

Google APBRmetrics too.

[Samuel-J]

http://en.wikipedia.org/wiki/APBRmetrics


---

[Idunkon1stdates]

Hey I'm doing a research paper for my freshman college writing class and its on basketball statistics.

I'm thinking of writing about what common statistical similarities are there between winning teams. Like when you guys look at stats, what jumps out as you as being essential to win a game/championship?

Any stats websites, general opinions or advice you guys could give me on this would be amazing since this is 40% of my overall grade.

dberri.wordpress.com

He wrote "Wages of Wins," which uses regression analysis to determine each stat's significance in determine how many games a team will win. For example, he found scoring alone has a low correlation to wins while FG% (adjusted FG% to be more precise) has a high correlation to wins. Agree or disagree with his conclusions, he arrives at them (and his model) using statistics.

I looked up defensive rebounding in the playoffs a year or two ago. Looked up all games the past 10 years and over 70% of the time the team who had the most defensive rebounds also won the game.

What surprised me is that offensive rebounds do not positively correlate to wins. The problem, of course, is that a team that shoots poorly will have more chances to grab offensive rebounds -- this is where the data can fool people.

Oh, and this is a forum that Dean Oliver (Basketball on Paper), John Hollinger (PER), and others post: http://sonicscentral.com/apbrmetrics/viewforum.php?f=1.

[Schad]

I looked up defensive rebounding in the playoffs a year or two ago. Looked up all games the past 10 years and over 70% of the time the team who had the most defensive rebounds also won the game.

What surprised me is that offensive rebounds do not positively correlate to wins. The problem, of course, is that a team that shoots poorly will have more chances to grab offensive rebounds -- this is where the data can fool people.

Yep, there's a definite connection between these two ideas; the team that grabs the most defensive rebounds tends to be the team forcing the opponent into the most misses, and thus the team likely to win on any given night.

The Lakers are a good example this season: they grab 2.3 more defensive rebounds than opponents per night this year. Even adjusting for pace, that's a significant gap. However, that doesn't make them a great rebounding team...they surrender offensive boards on 26.8% of missed shots, which ranks right in the middle of the league, and grab a lower percentage of offensive rebounds than they surrender (26.2%).

But because they hold opponents to 44.6% shooting, there are plenty of defensive rebounds to be had, while their 47.6% shooting presents fewer opportunities for the other team.

[Duiz]

Utah Jazz rebounds are great examples. they have extremely efficient shooting, and their defense isn't too great either, so they don't have many rebounds. However, they are a great offensive rebounding team, and they have the second highest point differential in the NBA.

Look at the Celtics, with great rebounding they do hold the highest point differential.

[TheSecretWeapon]

I looked up defensive rebounding in the playoffs a year or two ago. Looked up all games the past 10 years and over 70% of the time the team who had the most defensive rebounds also won the game.

A couple sites I use a lot:
82games.com
basketball-reference.com

Google APBRmetrics too.

As Schadenfreude points out, if you ran through those same games you'd probably find that the team with more defensive rebounds also held the opponent to a lower shooting percentage. The single most important stat is efg differential -- the team that shoots best from the floor wins about 80% of the time in the NBA.

For the original poster -- there are 4 key factors that decide winning and losing. Shooting, ball handling, rebounding and fouls. The factors are not equally important. Losing one (or more) does not mean losing the game -- a team can make up for poor shooting, for example, by getting to the ft line, not committing turnovers and by getting offensive rebounds.

These categories are measured as follows by most of us basketball stat goobers:

- shooting -- effective field goal percentage -- (fg + .5 x 3fg) / fga

- turnovers -- turnover percentage -- tov / possessions

- rebounding -- rebounding percentage -- for this we usually just use offensive rebounding percentages calculated by oreb / (oreb + opponent dreb). Do the reverse to get opponent offensive rebounding percentage.

- fouling -- free throws made / fga -- For whatever reason, getting to the FT line actually shows up as more important than making the free throws, but we still use FTs made as a measure of efficiency from the line.

In terms of general importance to winning, efg differential is a 10, turnovers about a 5-6, rebounding about a 3-4, and FTs about a 1-2.

Possessions are calculated by .976 x (FGA + .44 x FTA + TOV - OREB). Do that for both teams, then average the two numbers.

All of this combines to make offensive and defensive ratings. Offensive rating is 100 x (pts / possessions). Defensive rating is 100 x (pts allowed / possessions).