Components Methodology: NBA Lineup Evaluator

This article details the methodology and calculations of the components found on our NBA Lineup Evaluator. Each component represents a different skill or ability an NBA lineup could have. We can use these to asses strengths and weaknesses of NBA lineups that have yet to play together, or that haven’t played enough minutes to accurately evaluate their performance. Data is trained from NBA Lineups from 2015-2017 that played at least 50 minutes together. All data comes from either NBA.com or Basketball-Reference.com.

Introduction

As fans of the NBA, and especially the Minnesota Timberwolves, we often find ourselves sitting and wondering what a lineup with Kristaps Porzingis instead of Andrew Wiggins might look like (especially when Phil Jackson was still the coach). How much better would the rim protection be? Would Towns and Porzingis together space the floor well? Obviously, yes. What about the lineup’s ability to move the ball? As you can imagine, we could go on and on about qualities on which to evaluate a lineup. Our NBA Lineup Evaluator seeks to provide insights into how good a lineup will be overall, offensively and defensively, while also providing insights into specific traits in which we’d expect that lineup to overperform or underperform—such as the ability to protect the rim, space the floor or move the ball, to name a few.

Here’s a list of the components we evaluate, along with a short description of each.



Components

Diversity can also be thought of as unpredictability. We measure this by the diversity of play types a lineup would run; theoretically, the more equally you shoot out of certain play types, the harder it is to gameplan against you. The full calculation and breakdown of this measure, coming soon.

Playmaking (AST%): Rather than simply looking at a cumulative sum of total assists or per-possession assists, we weight each player’s assist percentage in a lineup by their percentage of the total assists contributed to the lineup.

Spacing  is a trendy word in today’s NBA. This component seeks to capture a lineup’s ability to produce and score on the most efficient shots in basketball: shots at the rim and 3 pointers. A full breakdown coming soon.

Scoring Efficiency (TS%): While a highly spaced lineup should in theory score efficiently, that is not always the case. Additionally, our spacing calculation doesn’t account for all types of shots. Some players score efficiently on shots besides threes and layups (e.g. Dirk from mid-range). Our lineup scoring efficiency measure is the sum of the True Shooting Percentage for each player, weighted by their usage.

Ability to Force Turnovers (STL%): This a tough skill to capture, especially at a lineup level. Due to limitations in the available data, we are only accounting for turnovers generated by steals; thus, our lineup steal percentage estimate is the sum of the steal percentages of the individual players in the lineup.

Rim Protection (BLK%)  This is calculated by adding the block percentages of the individual players in the lineup. We will note that blocks aren’t the only result of good rim protection; FiveThirtyEight put together a good piece outlining how the best rim protectors often prevent players from even attempting certain shots (i.e. Rudy Gobert preventing shots at the rim). Nevertheless, a BLK% estimate should give us an adequate measure of rim protection.

Overall Defense (DBPM): In order to capture defensive skills not found within steals and blocks (which is a lot), we have incorporated a “overall” defensive metric. This is calculated as the sum of Defensive Box Plus Minus of each player.

Best Player (Maximum VORP): Who is the player taking your last shot? Who is going to take over the game and control play? Basketball is a team game where the best player in the game can take control. In our investigation, including a measure of a team’s best-player talent helped better explain offensive and defensive rating. (Whether that is more of stabilization than actual inference is another question). Anyway, we decided to use an all-encompassing statistic, of which there are many. For this exercise, we used value over replacement-level player (definition here).

Offensive Rating: Offensive Rating is a measurement created by Dean Oliver to capture “the number of points produced by a player per hundred total individual possessions.” We can apply this to a lineup level by simply using the specific lineup’s points produced.

The calculation can be found below:

ORTG = 100 x Pts / (0.5 * ((Tm FGA + 0.4 * Tm FTA – 1.07 * (Tm ORB / (Tm ORB + Opp DRB)) * (Tm FGA – Tm FG) + Tm TOV) + (Opp FGA + 0.4 * Opp FTA – 1.07 * (Opp ORB / (Opp ORB + Tm DRB)) * (Opp FGA – Opp FG) + Opp TOV)))

which is essentially:

ORTG = (Points Produced / Individual Possessions) x 100

We use a Gradient Boosting Method to impute the ORTG’s of Lineups. For more information and detail on the methodology behind these model coming soon.

Defensive Rating: is a slightly simpler calculation than Offensive Rating.

DRTG = (Opponent’s Points Allowed/ Opponent’s Possessions) x 100

Similarily to ORTG, we also use a Gradient Boosting Method to impute the DRTG of Lineups. While we use similar inputs, the variables obviously have different levels of importance. More information on this model coming soon.

Net Rating: 

This one is pretty simple!

NETRTG = ORTG – DRTG

 

The NBA Lineup Evaluator tool can be found, here.

Written by  Marc Richards and Jack Werner.

96 thoughts on “Components Methodology: NBA Lineup Evaluator

  1. Pingback: ivermectin 20 mg
  2. Pingback: ivermectin ireland
  3. Pingback: cialis generic
  4. Pingback: ivermectin 8000
  5. Pingback: generic for cialis
  6. Pingback: viagra for women
  7. Pingback: ivermectin 9mg
  8. Pingback: order cialis
  9. Pingback: cialis best price
  10. Pingback: tadalafil india
  11. Pingback: cialis dosage
  12. Pingback: order viagra
  13. Pingback: cialis drug
  14. Pingback: tamoxifen oral
  15. Pingback: aralen 50mg
  16. Pingback: nolvadex cost
  17. Pingback: tizanidine 6 mg
  18. Pingback: use of cadista
  19. Pingback: gay sm dating
  20. Pingback: gay dating page
  21. Pingback: dapoxetine priligy
  22. Pingback: gay dating clubs
  23. Pingback: keto margarita

Comments are closed.