Basketball IQ

Team Offensive Decision Making IQ

Abstract

In this study we introduce simplified Team’s Offensive basketball IQ metric. It is known that a player’s basketball IQ constitutes of multiple, complex and effective decisions. Team’s IQ is in correlation with the IQ of its Players. One of the approaches to quantify Basketball IQ is to consider that an intelligent decision means that a Player will take a Field Goal with potential equal or higher Percentage compared to rest of the Team’s potential FG%. Currently it seems too complex to create an algorithm for calculating this IQ automatically, due to numerous possible situations. So, via simplification, we want to quantify the Basketball IQ of a Team in offense with some fundamental approach. In order to do that we are going to characterize each team’s play having an Optimal or Not Optimal Decision, meaning an intelligent or not decision. Ηigh Basketball IQ means that Team is reaching to its potential capabilities. In addition, we introduce the ‘IQ for missed Field Goals’ (IQM). The difference between IQ vs IQM, is that IQ refers for all Plays, but IQM only for Missed FGs. Same metrics can be defined respectively for defense.

1. Introduction

In live basketball games and in post-game reviews, coaches are constantly analyzing each Team’s plays and decisions. Good coaches can evaluate if each Team’s play is a result of a correct decision or not. Great coaches can do that faultless, and also can have an accurate overview of each Player’s Decisions and of all Team’s decisions; and they can adjust accordingly. What is missing is a standardization and quantification of coach’s analysis for this Basketball IQ.

A thorough approach to calculate a Team’s Basketball IQ would dictate to utilize and evaluate stats from Box Scores, Synergy, Tracking NBA SPORTVU system, and on top of them to have a coaching expertise to evaluate the situation according to all possible variables, such as set play (PNR, Isolation , Post Up, Cut, Transition, Handoff etc), area spot, contested shot, time, and yes the much debated hot hand, etc. In this study, we will approach the IQ scenario with less steps and details. Similar to Synergy, manually logging is necessary in order to tag if a Play has an Optimal or Not Optimal Decision, ie if a Play has or not an ‘IQ and Effectiveness’ decision. Inevitably, subjectivity is part of the process. Though, we will try to give the logic as much as possible as an objective algorithm. Eventually the algorithm may be feasible to be translated to a Programming Language having as input necessary available data.

2. Definition of IQ

IQ is a metric of Optimal Decisions and effectiveness for all Plays in a Game. By definition Optimal Decisions are including all Made FGs and 0.44 * FTs. Furthermore, Turnovers are not Optimal Decisions. Value of IQ ranges approximately between 40-90%.

Outcomes of each Play can be:

• An Optimal Decision
• A Not optimal Decision

IQ and Effectiveness (symbol IQ) is calculated by formula:

, where

• Optimal Decisions = Made FGs + 0.44*Made FTs + (Optimal Plays in Missed FGs) - Turnovers
• Plays = FGA + FTA * 0,44 + Turnovers

Max and Min values of IQ are:

IQ_MAX means that all Missed FGs where Optimal Decisions. Value is approximately 90% [1], and less than 100% [2].

,where

• Scoring Plays = Made FGs + 0.44 * FTs

IQ_MIN means that none of Missed FGs where Optimal Decisions. Value is approximately 45% [3] .

3. Definition of IQ for Missed FGs (IQM)

We can further introduce IQ only for the Missed FGs. In essence it illustrates the ratio of missed Shots that were a good option. The difference between IQ vs IQM (IQ for Missed FGs), is that IQ refers for all Plays, but IQM only for Missed FGs. Value of IQM ranges theoretically between 0-100%.

IQ for Missed FGs (IQM) is calculated by formula:

If all of Team’s Missed shots are Optimal Decisions, then the Team is playing according to their actual potential. That is, Team’s FG% is illustrating maximum FG% that team can reach. It’s obvious that having always the right decision, doesn’t mean that you will have 100% FGs. On the contrary, if Team’s Not Optimal Decisions are increased, equally the Team is declined from its Potential, and actually is playing worse than its ability. So:

IQM Max = 100%, ie Team is playing to its actual Potential (all Missed Shots are Optimal Decisions)

IQM Min = 0%, ie Team cannot play worse (none of Missed Shots are Optimal Decisions)

4. Definition of Defensive IQ and IQM (IQ for Missed FGs)

Similar to Offense, we can define IQ and IQM for defense (DIQ and DIQM). We apply same logic, but now we tag Opponent’s Field Goals. Higher DIQ or DIQM means a bad defense that allows Optimal Decisions on Offense.

5. Axioms

In order to define Basketball IQ and also to simplify its calculation, we are going to set below Axioms:

• IQ is defined as the number of Optimal Decisions per number of Plays
• IQM (IQ for Missed FGs) is defined as the number of Optimal Decision per Missed FGs
• The difference between IQ vs IQM, is that IQ refers for all Plays, but IQM only for Missed FGs
• All Made Field Goals are considered as Optimal Decisions. Even if a shot selection seems not optimal, we are obliged to count it as Optimal due to the effectiveness and value of the Made Shot. Eventually, the winner is always the more Effective Team
• All Plays that result to FTs are considered as an Optimal Decision. FTs Percentage dictates this rule
• Turnovers are Not optimal Plays
• A non scoring Play ie a missed FG, can be either an Optimal or Not Optimal Decision. It is possible to have an Optimal Decision that leads to a Missed Play (it seems rational, that the Team with the most Optimal Decisions will be more likely to win the game)
• A contested or not shot, can be either Optimal or Not Optimal Decision. Some players have the ability to make Contested shots. So, a contested shot doesn’t mean a ‘Not Optimal’ Decision
• A Player missed shot that has Lower Percentage compared to Team’s Percentage, is defined as Not Optimal. For example, if a Player is shooting with 30%, while Team’s percentage is more than 30%, then this shot is considered as Not Optimal
• For this task Synergy, or Tracking NBA, or spatial statistics could provide support, since a Player will have different percentages depending on various spots, or on how contested is his shot
• Furthermore, the type of shot (spot, off balance, layup, etc) can vary between players. Hence, it seems difficult to objectively calculate with precision the percentage of a shot
• A manual subjective check is needed per missed field Goals, for all Missed FGs via a Decision IQ Flow, which tries to reduce the subjectivity

6. Play Categories

In order to define if a Missed FG is an Optimal Decision, we will apply previously axioms and following rules per Play category.

Transition

Transition is defined as Play ending in a 1vs1 situation, or if Offensive team is having more Numbers (2vs1, 3vs2, 4vs3, 5vs4), or when Defense is not yet set. We categorize result as Optimal or not Optimal Shot option:

• IQ Optimal Decision: Missed Layup
• IQ Optimal Decision: Missed Open FG, when you don’t have a layup
• IQ Not Optimal Decision: Missed Contested Shot

Set Play (PNR, Isolation, Post Up, Cut, Off Screen)

We categorize result as Optimal or not Optimal Shot option:

• Objectively: Via Synergy, Tracking Stats, or Spatial stats comparing attempted Player’s FG% versus Team’s Average FG%. If attempted Player’s FG% is less than Team’s Average FG%, then this is a Not Optimal Decision
• Subjectively (if above is not feasible): Via judgement if this is a Shot with higher or less percentage than Team’s Average Percentage
• Furthermore, in some cases it can be easy to evaluate that a pass instead of a Shot, could lead to better eFG%. This is the moment that again coaching/logging subjectivity will judge the case
• Example: In some PNRs it’s easier to see some better options. For example, if PNR Ball Handler had an option to pass to a wide-open Teammate, that is having eFG%, than the hard defended Player that actually took the shot, then this is a Not Optimal Decision. Hypothetical Scenario in a Fenerbahce PNR: If Sloukas is passing to PNRoll Man who misses a hard defended shot with potential 55%, and in parallel Datome is Wide Open on the 3pt Wing (who has a 60% in Open 3pts shots), then this is a Not Optimal Pass Decision for Sloukas
• Due to its dominance in the frequency usage, PNR can be distinguished in more different options that rest of Set Plays

Special Situation

There can be a lot of Special Situations in a game. In our model, we will limit Special Situations only to last 24’’ of first 3 quarters. Also, at the last 24’’ seconds of 4^th Quarter but only considering situations where there is a game draw (eg 75-75 and a Team has possession in last 24’’); for 4^th quarter intention fouling will not be handled.

A Team that has a possession in last 24’’ of a Quarter can utilize all 24’’ without the Opponent to make any other offense. So, at last 24’’ the Team with the possession should have equal or more points produced than its opponent. This should be a derivative of an Optimal Decision. On the contrary, it will be a Not Optimal Decision if the opponent will eventually produce more points.

7. Offensive Decision Making Basketball IQ Flow

Base to previous paragraphs, below follows a diagram flow in order to illustrate if a Missed FG, is an Optimal (OK) or not Optimal (NOK) Basketball IQ option.

8. Case Studies

1. 2018-19 Regular Season, Euroleague, Bayern Munich – Fenerbahce Beko 90-86

The game was one of the most intriguing games of Euroleague, not only due to the 2 Overtimes, but also due to numerous difficult Shots made by both teams. Analyzing all Plays of the game, and applying above formulas for IQ and IQM (IQ for missed shots), the final results are:

Bayern’s IQ 69%

Fenerbahce’s IQ 66%

, and

Bayern’s IQM 54%

Fenerbahce’s IQM 62%

Complete analysis can be found here. Also, 4 Video Clips for an Optimal or not Optimal decision for each Team can be found to this Youtube playlist.

2. 2019 FIBA World Cup, Spain – Serbia 81-69

Reaching an almost perfect game, Spain had an IQ of 78% and an IQM of 87%. This actually means that they made almost in any Play the Optimal Decision.

Both of these numbers can be used as a benchmark. This means that Team’s could set as a goal to have 80% of IQ and 90% of IQM.

9. Further Study – More related metrics

Apart of IQ and Effectiveness which considers as Optimal Decision a Made FG, we can even consider to count purely as IQ the higher percentage shots compared to Team’s Average, irrespectively if they are Missed or Made.

So, we can also generate these metrics:

* Refer to metric EFG+ in ‘Quantifying Shot Quality in the NBA’:

http://www.sloansportsconference.com/wp-content/uploads/2014/02/2014-SSAC-Quantifying-Shot-Quality-in-the-NBA.pdf