Decision Analytics

The Brainball Approach

Transform subjective decisions made using subjective factors
into expert decisions made using objective factors

Game of Opinions

In football, many decisions are made subjectively to a greater or lesser degree, namely transfers, squad selection, team selections, position placements, and substitutions. In each of these cases the decision maker has a series of internal and external subjective factors that they use to make a decision, and these subjective factors have been fed by a series of observations. The subjective factors can differ according to experience, emotions, temperament, amount of time witnessed, confirmation bias and external influences.

Expert Decisions
Our definition of an expert decision is a decision that uses the objective factors that are known predictors / measurements of output performance for a particualr role. When looking at a player, BRAINBALL have currently researched these factors according to the following process:

Objective Factors

The key objective factors that we have defined that are known predictors / measurement of player performance are:

Decision Levels
Using a series of specific criteria, we can provide objective technical player performance factors to help make an expert decision. These factors are more advanced than the commonly used factors like pass success rate, goals scored, and minutes played. They provide a complete technical and player performance input for decision-making, eliminating the need for subjective opinions about player performance.
Expert Layer
Comparing our OBJECTIVE FACTORS assessing the number of data entry points used to inform the decision against a normal decision in a similar subject field EG: Item 8: Goals against XG

Item 8
NORMAL DECISION:
Goals scored, has only 1 data point, the goal.

EXPERT DECISION: XG, has 5 additional data points. Position on field, right or left or headed attempt, on or off target, rating according to chance of scoring. These factors combined produce a percentage chance of scoring with multiple layers of objective data.

Defining Our Objective Factors

Bring the game into focus

It is important to note that these are not obviously the complete set of objective factors used, but do cover our view of the required technical player performance factors. Other groups of factors required are temperament/emotional, physical/fitness, and team make up. We will not about these factors in this paper but they should be objectively constructed

Previous slide
Next slide
Objective Data
Now once we have decided what factors we need - by eliminating the need for subjective data - we can ensure that the factors are defined by objective statistically significant data. Brainball anaylysesa series of matches, ensuring that each player has played at least 270 minutes of football (equivalent of 3 football games). The analysis involves 3 key steps.

1

2

3

Filming and recording 270 minutes of matches that involve each player in a competitive situation.

Analysts reviewing each recording and transcribing the information into a series of descriptive statistics that do not require any subjectivity. Each action in the football match is recorded together with the success and outcome of each action.

Summarising the tagging of matches for each player, normalising by minutes played and calculating the factors/measured required for an expert decision

Data Tagging SAMPLE

Operations
BRAINBALL utilises the HBUFC employment hub to:

- Organise and film matches to enable 270 minutes of recorded competitive playtime for each player

- Tagging the matches to create an objective transcribed data view of a football match, action by action

The data is then transformed by the BRAINBALL team to produce useable statistical data and then present it in a way to be useable and referenced in expert decisions. This is presented in a series of player cards by position to be used in expert decisions.

Player Cards. Midfield Focus
The objective factors that we have defined differ in importance to the position / decision being made. At a high level the following guide is used.
Sample Player Cards
Attacking Midfield
Previous slide
Next slide
Player Comparison
Attacking Midfield
Expert
Decision:
SHAHEED: THE BEST ATTACKING MIDFIELD

Shaheed is by far the strongest attacking midfield player, he creates at least 3 times as many shot creating actions as his team mates, twice the shots, his XG is more than 2 goals every 3 games and the number of people past with actions are phenomenal, he drives the team forward, creates momentum and he is the most effective attacking player in terms of creating and taking chances. However he does make mistakes...so cover is required for these with a more risk adverse player. Luka and Putty are similar with shot rate and XG and people passed etc. Luke slightly more attacking, Putty more involved defensively. He should be partnered with Shaheed.

Play Smarter

Verified by MonsterInsights