Home

Post-Match Dashboard

Visualizes key events from the match, allowing us to see how and where teams attacked, any patterns in their attacking play, and stretches of the game where they were in control.

Game Control Index

This visualizes both the stretches of time a team was in control of a match as well as the magnitude of control they exerted. Unlike many “momentum” or “flow” charts you normally see, this takes into account both on-ball/attacking data & off-ball/defensive data.

You can use this index to see which team was in control of the match overall, how a team responded after conceding or scoring, and more.

Pass Map

Pass maps show the passing network of teams in a match. We can see where players usually were when they passed as well as their main passing partners.

You can use pass maps to analyze a team’s structure & patterns in possession. Were they held to deep passes between center backs? What was their in-possession shape? Were their fullbacks high? Did their striker typically drop deep?

Passing Sonars

Passing sonars offer a lot of information about a player’s passes at one time. We can see the average directions players passed in (the orientation of the bars), the relative number of passes in each direction (bar color) and the average length of passes in each direction (bar length).

We can use these visuals to look deeper into an individual player’s passing characteristics in a match (or a full season).

Top X Passing Chains

Passing chains show the longest stretches of sustained possession a team had in a match. I usually visualize the top 15 chains, as that also shows us if teams were able to retain possession for long stretches at a time when they were in possession.

A key use for this visual is to see if there were any regular passing patterns a team used when they had possession. We can also see if perhaps they found it difficult to break down their opponent, since they would have plenty of chains that include deep passes or passes on the flanks. We can also see if they were able to turn any longer stretches of possession into shots, which this particular example shows Vegalta Sendai doing just that.

Expected Threat (xT) Heatmap

Expected Threat essentially shows the change in probability of scoring a goal in the next few actions by moving the ball from one place to another. My charts show passing xT, so it’s the increase or decrease in probability of scoring a goal by passing from the passer’s location to the receiver’s location.

We can see lots of good info in these charts, such as where the team mainly created threat (in this example, mainly the right flank), which player was the most threatening from each zone, as well as each player’s total xT.

xT Zone Control

Expanding on the xT heatmap for each team, we can use this chart to see which team was the most threatening from different zones on the pitch, as well as how much more threatening the main team was.

Passes & Shots in the Final Third

This illustrates all passes a team made while they were in the final third, as well as their shots. A key way to use this is to see how a team used their passes in the most advanced area of the pitch, however it also gives insight into their opponent’s defensive structure.

Typically, teams who face well-drilled blocks will find it hard to use the space in front of the box and will find it hard to get the ball into the box. You tend to see a lot of lateral passes against these defenses.

“Key percent” is the percentage of completed passes that lead to shots, and “final third retention rate” is the percent of all passes in the final third that keep possession inside the final third. This is an important statistic to look at when analyzing a team’s final third play.

Defensive Actions

Seeing where teams made on-ball defensive events can paint a picture of how they wanted (or had) to defend.

This image also shows the average height of the team’s defensive line (excluding clearances), as well as their PPDA. PPDA is a team’s passes allowed per defensive action, which attempts to show pressing intensity. A lower number indicates a more intense press.

Pre-Match Final Third Analysis

This large image compares a lot of key event data for the last 4 matches: the focal team’s actions in the final third, and what their opponent has faced in their defensive third.

We can use this to see what could be areas of strength or weakness for a team. This can then be used to preview games and see either where a team should attack or look to shore up their defense.

For example, if the focal team is better at crossing from their right flank then left, and their opponent is weaker defending crosses from this side as well, then the focal team should look to exploit their opponent’s weakness which is also their strength.

This is not comprehensive, but looks to offer some insight into recent strengths or weaknesses of a team’s attack or defense.

Single-Match Actions

These charts are great for seeing what a player did during a match.

From an analysis standpoint, we can use these to look at players beyond just numbers and see if they played their role how they were supposed to play their role, or perhaps if there was an area they could have performed better in to have more/better impact on the game.

Key Attacking Acts

These charts illustrate a player’s key attacking contributions, in the form of dribbles, shot & goal assists, shots, and goals. They allow us to see some of the main goal-creation event data. These can be particularly useful for forwards & attack-minded players.

Beyond attacking acts, I also visualize every element of a player’s game individually, from passes in a specific third to crosses and even free kicks & corners.

Passes After Ball Recovery

An important analysis for midfielders, this chart shows us all passes a player makes after picking up a loose ball and gaining possession.

The interpretation of this chart is up for debate, since some people may prefer players who play safe passes after gaining possession, and others may want players to look for possible counter-attacking options and won’t mind forward passes that immediately turn the ball over. There is no right or wrong: a player’s tendencies may be the outcome of managerial instructions so interpret with a grain of salt.

Defensive Actions

It can be useful to see where a player tends to make defensive actions, as well as how many of each action they perform. Normally, players have many more ball recoveries than any other metric. The best use of these charts is typically to see where a player’s defensive responsibilities lie.

Passing Solar

Passing solars show a player’s main passing targets as well as their average relative location to the player and their average distance as well. They can be very useful for seeing who a player normally passes to and where that receiver normally is in relation to the focal player.

Solars are sort of on the border between analysis and a pure visualization. I code mine to look like a solar system, with each teammate being a “planet” (sized according to number of passes) and each planet having “moons”, which are the standard deviation of that player’s receiving location relative to the focal player. The “orbit” lines of each player are their average distance.

Pass Clusters

Pass clusters are important pieces of analysis, as they allow us to see the most exemplary passes a player makes. These help us get an overview of the player’s passing tendencies: are they usually playing short or long? Do they normally progress the ball forward, move it laterally, or recycle backwards? There are many ways we can interpret these for all different role-positions.

Key Distributions

Key distributions include attacking switches and progressive passes.

Attacking switches are passes that travel at least 50% of the pitch from side-to-side and at least 10% forward at the same time. Typically, these are switches with an intent to start or continue an attack.

Progressive passes have many definitions, but I use passes that move the ball at least 25% closer to the goal (based on remaining area of the pitch with the ball as the focal point), or any pass into the box.

Both of these types of passes, when taken together, can give us a decent look into the player’s tendency/ability to start or move attacks forward.

Goal Kick Analysis

These visuals show all of a goalkeeper’s goal kicks, and then breaks the successful kicks down into ones that were initially successful but didn’t retain possession (particularly important for long kicks), and ones that did.

The statistics on the right side allow for a deeper dive into a goalkeeper’s tendency or ability with goal kicks: do they normally take them short, are they good at completing long goal kicks, etc.

Typical Buildup Passes

As goalkeepers become more and more a part of teams’ buildups (see Magdeburg, HSV, Mamelodi Sundowns, and Fujieda MYFC for examples), it’s increasingly important to look at the patterns of a keeper’s most-played passes in buildup.

This specific graphic shows the most typical buildup passes, but I have several variations of this and deeper dives into this concept in my arsenal, such as the images below.

Position-Role Clusters

Using an Affinity Propagation algorithm trained on a diverse range of leagues from around the world, I can cluster players into one of 54 different roles based on their events heatmap, similar to this paper doing something similar in the NHL.

The algorithm spit out 54 different clusters. I then went through each cluster & the players the algorithm placed in each, to subjectively name them and put them into positon groups of GK, CB, FB, DM, CM, AM, Winger, and ST. These roles are supposed to be general, like “half-space CAM” and “box-entering winger”, as other data will tell us if a player may be a “playmaker” or “shooter”, etc. The heatmaps can’t tell us that.

The idea is to better understand where a player is actually playing during games (not on the team sheet), and perhaps get general info on their style at a glance (progressing RW vs. box-entering RW, for example).

VAEP Profiles

VAEP stands for “valuing actions by estimating probabilities”, and is used to value a player’s actions in terms of the added probability of it creating a goal (or conceding one) within the next X number of actions, given the previous Y number of actions and other context. Here is a great interactive explanation.

My profiles break down VAEP into 7 different categories: Dribbles, Aerials, Tackles & Interceptions, Recoveries, and open play Receiving, Passing, and Shooting. This enables us to get a good picture of not just how valuable a player might be, but dive into their strengths and weaknesses.

I use the role cluster groups (FB, ST, DM, etc) explained above to create the comparison samples for these profiles.

VAEP Event Overview

These visuals further break down a player’s VAEP profile and allow us to see a random sample of specific actions and their values.

With this view, we can get a much better picture into a player’s style & value of play.

Cutbacks by Team

Cutbacks are one of the most dangerous types of passes a team can play. Seeing how teams use cutbacks, both how often and their patterns and effectiveness, is important. These visuals allow us to see teams who might be trying lots of cutbacks, or maybe even teams who should try more cutbacks because they’re effective with the relatively few they make.

Pass Clusters

We can analyze teams’ pass clusters in the same way we can analyze players’. It’s even more powerful to break passes down into specific type of passes, like key passes (shown), crosses, passes into the final third, passes when in the final third (or middle or defensive), and much more. These are often very good for exploratory team analysis.

Defending the Box: Opponents’ Passes Heatmap

Analyzing a team’s defensive play with event data is difficult, but one of my favorite areas to explore. Typically, we need to flip event data around and look at the patterns of a team’s opponents, like in this specific graph. It shows where a team’s opponents typically complete passes to when they’re in the final third. The teams with the best (and worst) box defenses/defensive blocks will stand out. Good defenses should be able to force opponents onto the flanks and away from the box.

Pass Target Areas Between Boxes

How a team uses the space between boxes is important to know. Do they use the flanks to move the ball forward or maybe try to move it inside? How do they use half-spaces? What patterns do they exhibit? All of these questions help us understand a team’s in-possession play.

In the Specific example shown here, we see Buriram United use midfield and attacking half-spaces mainly to move the ball wider, but defensive half-space to move the ball both central and wide.

Player Average Location & Territory

Particularly interesting when analyzing defensive lines, seeing the average location and territory of several players at once adds a rich layer of detail when analyzing a team. We can see which players may be tasked with covering certain areas, either in possession (such as the image shown) or out of possession with defensive actions.

Vertical Passes Faced

Another piece of defensive analysis revolves around seeing how well a team defends against their opponents’ vertical passes when they’re in front of the box. These are naturally very dangerous passes, so need to see how well a team deals with them.

Using the completion rate and key percentage in tandem with any patterns of completions or key passes (are they normally on the right, left, center, or do the key passes originate from a particular spot?) adds even more richness to a team analysis.

PDI

PDI aims to rank players by how dangerous their passing style might be, based on the level and magnitude of things like shot assists, deep completions, key passes, and passes aiming to create attacking chances.

ARR

ARR attempts to determine the overall receiving rate added (or subtracted) from their teammates. A higher number means players are consistently able to play passes to players that are easier to receive than passes from an average passer, given the same pass: perhaps they’re really good at playing passes to their receiver’s strong foot, or good at threading the needle. A negative value means their passes may be harder to receive than the average player’s pass: maybe they are poor at the exact placement of their passes to their receiver.