Advanced Statistics – What They Mean and How to Use Them

As we mentioned in the last post, advanced basketball statistics or “APBRmetrics” attempt to take the stats we already keep and make them MUCH more meaningful.  There is a reason why NBA teams are paying whole teams of statisticians a lot of money to dig into these numbers, they are very insightful.  Here are a few of the things we can gain from looking at these stats at just a basic level:

• Further the ability to determine how good a team or player is on offense or defense.
• They create an even playing field so you can compare teams regardless of their style of play or the pace at which they play
• Allow us to look at players regardless of how many minutes they play
• Help to determine the most effective line-ups for coaches
• Serve as a supplement to player analysis, coach analysis, and more.
• Can help shape how you approach your (or your team’s) basketball training.

You will find a file(s) attached at the bottom.  It is an excel document that examines an actual Division 3 Men’s Basketball team’s statistics at an individual statistic level, as well as looking at a few teams within a division 3 conference.  This is intended for you to use to generate these same stats for your teams.  The formulas are all loaded in already, you just need to plug in team/player stats and you are set to go!

A quick word on individual statistics…

It’s really not fair to judge a player that plays 10 minutes a game against a player that plays 35 minutes, so we put everyone on an equal footing when comparing players.  It is okay to still look at how players perform an absolute level, but the per 40 minute calculation used in our attachment is where you really start to find some good info.  That said…you probably don’t want to extend this to guys who play less than 10 minutes a game.  It is not fair to extrapolate the stats of a guy who only plays at the end of the game in “clean-up” time.  So, in our template you will see both absolute and per 40 minute statistics.

Let’s talk possessions…

One of the first tenants of basketball statistics is to look at how good a team is offensively and defensively.  There is a problem in just looking at regular stats when comparing teams….every team plays a different style.  Some teams run up and down the floor and score a lot of points, some teams slow down the game and work the shot clock, and most teams are somewhere in between the two.  In order to most accurately compare teams, we look at how each team does by possession.

We know that possessions end in just a few different ways: a missed shot, a missed free throw, or a turnover.  However it is not perfectly straight forward because sometimes a player scores and is fouled (so there is a field goal attempt and a free throw attempt)…so we have to apply a factor (approximately 44% of free throw attempts end a possession) to the free-throws to account for this.  The formula for possessions is:

Possessions = .96 * (FGA − ORb + TO + (.44 * FTA))

Not that hard yet, right?  Okay, now we calculate an offensive rating and a defensive rating.

Offensive Rating = (Points Scored * 100)/Possessions

Defensive Rating = (Points allowed * 100)/Possessions

These ratings actually tell you how good or bad you are at offense or defense.  This ignores the pace at which you play and other external factors.  If a team has an O-rating of 112 that means they score 112 points in 100 possessions, even if that team is among the low scoring teams in a league, they have the most efficient offense.    The same principal goes for defense, except a lower D-rating is better than a higher rating.

Use & Details: This is most easily calculated for team statistics, and takes on even more meaning when you look at these ratings for when a particular line-up of players is in the game.  This can also be calculated on an individual player level, but it is a much more complex formula.

Effective FG% = (Field Goals + .5 (3-pt Field Goals)) / Field Goal Attempts

This is just a small twist on a normal field goal % which gives made 3 pointers a higher weighting.  The ability to hit 3 pointers at a higher rate makes a player a more efficient shooter from a statistical standpoint.

Use & Details: This is used for both teams and players.

True Shooting % = Points Scored / (2 * (Field Goal Attempts + 0.44 * Free Throw Attempts))

True shooting % is similar to the effective FG% except it takes into account field goals, 3 point field goals, and free throws.   It looks at the number of points scored on your various shot attempts.

Use & Details: This is used for both teams and players.

Adjusted Value = (PTS * weighting) + (AST * weighting) + (OReb * weighting) + (DReb * weighting) + (STL * weighting) + (BLK * weighting) – (Missed FG * weighting) – (Missed FT * weighting) – (Turnovers * weighting) – (Fouls * weighting)

Adjusted Value attempts to look at a player’s total contributions to a team.  This can also be calculated at a team level to compare team vs. team in terms of statistical efficiency.  In a nutshell good things give you points, bad things take away points.  Each of these stats is multiplied by a weighting of importance.  There are a number of different weightings out there and various pieces of analysis for deciding what that rating may be.  The rating I use is actually an average of 10 different weighting models from Dean Oliver’s “Basketball on Paper” (a truly awesome book).  This is a great way to look at the value a player adds to your team.

*As mentioned, this is “adjusted value” there is also a “value” calculation which would be this formula without the weightings.

You can see the weightings in the spreadsheet attached, but steals and offensive rebounds are viewed as the most favorable (with a weighting of 1.27 and 1.19 respectively) and turnovers are the most costly (with a weighting of 1.28…basically an offset to a steal)…so a player that has a lot of steals may improve some type of adjusted Assist: Turnover ratio.

Use & Details: Most effective to evaluate adjusted value to compare players, but it can also be useful to compare teams, as it is indicator for which team may be expected to win in a match-up or how effective coaching is (i.e. winning more games than your adjusted value might indicate. It is often more useful to look at this from a per-game perspective.

Offensive Rebound % = (Offensive Rebounds * 100) / (Field Goal Attempts – Field Goals)

This statistic actually shows how many times a team rebounds one of their missed shots.  When you rebound your own shot you often have the opportunity for a very easy put-back or close range shot.  These are often very high percentage shots making this a very valuable statistic

Use & Details: This is used for both individuals and teams.  However, when used individually this should be adjusted for the amount of time on the floor (minutes).

Free Throw Rate = Free Throws Made / Field Goals Attempted
This measures how often a team gets to the line and how well they shoot from the line.  Getting to the line helps a team eliminate the effect of defense and turnovers, giving them a free shot(s) at the line.

How do you compare to other teams in your conference?  What are your ratings on a game by game basis?

Use & Details: This is used for both individuals and teams.

Expected Win % = (FG% – Defensive FG%) x 5 + 0.5

This is a very high level estimate of how many wins a team might be expected to win based on several pieces of analysis that show an extremely high correlation of the difference of FG% & Opponent FG% to Wins.  This basically says you will games in proportion to how much better you shoot than your opponent.

Use & Details: Team stat only.  You should compare this to Actual Win %.  There are a few things this means…. 1) In theory a coach with a higher actual win % than expected is doing a good job as a coach (this is admittedly very short sighted, as there are MANY more factors at play, this is just one way of showing how a team is doing…and perhaps if they are outperforming or underperforming).  2) A team has had some tough losses they probably should have won, but lost by a close margin.

Creation to Turnover Ratio = ((Ast * .99)+(Blk * 1.02)+(Stl * 1.27)) / (Turnovers * 1.28)

This shows what an individual player (or team) might contribute to a team regardless of scoring & shooting.  The weighting assigned to each assumes the relative importance of each (i.e. often times a steal leads to easy points (lay-up or open 3)) therefore it is weighted heavier.  On the flip side, a turnover often leads to the same thing for the other team.  An assist and a block are often considered less critical to a team’s success than a steal or turnover, so they are weighted lower. (These weights are also from “Basketball on Paper“.

Use & Details: Mostly used as an individual stat.  This should be looked at in terms of what a player might bring to the table in terms of possession/point creation vs. possession/point giveaways.

Still to come:

• What are the most significant statistics to look at?
• Stats you may want to begin tracking…
• Other miscellaneous statistical items

NOTE: IF YOU HAVE TROUBLE WITH DOWNLOADING THE ATTACHEMENTS SHOOT ME A NOTE AT  JOE@THEWORLDOFHOOPS.COM.

What are your thoughts on advanced statistics?  Do you think they are important?  Do you use them?  Are you going to start?  Be sure to leave your comments below!