Carry Maps

Definition

Carry maps visualize ball carries (also called dribbles or runs with the ball) on the pitch, showing the start and end positions of each carry as lines or arrows. They highlight how players progress the ball through individual movement rather than passing, revealing dribblers, progressive carriers, and players who advance play by running with the ball.

History & Origins

Ball carries were historically undervalued in football analytics because early event data focused heavily on passes and shots. A player who received the ball, ran 30 meters forward into dangerous space, and then passed sideways would get credit for a short sideways pass but none for the carry that created the danger.

This changed as data providers began explicitly logging carries as events. StatsBomb was a pioneer in this area, introducing carry events as a formal part of their event data specification around 2018–2019. Before this, carries had to be inferred: if a player received the ball at point A and their next action (pass, shot, etc.) was at point B, the movement between A and B was implicitly a carry. Opta and Wyscout adopted similar approaches, though the level of detail varies between providers.

The visualization of carries gained traction in the analytics community around 2019–2020, driven by the realization that some of the most impactful players in football (e.g., Adama Traoré, Lionel Messi, Neymar, Alphonso Davies) created enormous value through carrying rather than passing. Analysts like those at StatsBomb, The Athletic, and independent Twitter analysts began publishing carry maps alongside pass maps to give a more complete picture of ball progression.

The concept of Progressive Passes & Carries|progressive carries — carries that move the ball significantly forward — became a standard metric, mirroring the progressive pass framework. FBref now displays progressive carries per 90 as a default statistic.

How It Works

1. Identify all carry events in the data — either explicitly logged or inferred from consecutive events by the same player at different locations
2. For each carry, record the start position and end position
3. Draw lines or arrows from start to end on a pitch representation
4. Optionally filter by type: progressive carries, carries into the final third, carries into the penalty area
5. Optionally color by value: xT - Expected Threat|xT gained, distance covered, or outcome

Carry Detection (when not explicitly in the data)

If working with event data that doesn't log carries:

• Player A receives the ball at position (x₁, y₁)
• Player A's next action (pass, shot, dribble) is at position (x₂, y₂)
• If (x₁, y₁) ≠ (x₂, y₂), the movement between them is a carry
• Distance and direction can be computed from the coordinate difference

Variants

Progressive Carry Maps

Filtered to show only Progressive Passes & Carries|progressive carries — typically carries that move the ball 10+ meters forward or into the final third / penalty area. These are the most analytically useful for identifying players who drive play forward.

xT-Colored Carry Maps

Carries colored by the xT - Expected Threat|xT value gained, similar to xT-colored Pass Maps|pass maps. A carry from the halfway line into the edge of the box lights up; a carry sideways along the touchline stays neutral.

Dribble Maps

A subset focusing specifically on take-ons (1v1 dribbles past an opponent) rather than all ball carries. These show where a player attempts to beat defenders, with outcomes (successful/failed) typically color-coded.

Carry + Pass Combined Maps

Showing both carries and passes for a player in the same visualization, revealing their complete ball progression profile. This distinguishes a player who progresses through passing (e.g., Toni Kroos) from one who carries (e.g., Adama Traoré) or one who does both (e.g., Messi).

What They Reveal

• Carry profiles: identifies players who progress the ball through movement vs. passing
• Positional tendencies: does a fullback carry up the wing or cut inside? Does a midfielder carry centrally or drift wide?
• Pressing escape: carries out of defensive areas indicate a player who can break the press under pressure
• Counter-attacking threat: long carries into transition spaces reveal speed and directness
• Role identification: some center-backs (e.g., Rúben Dias, Virgil van Dijk) carry forward significantly; carry maps reveal this dimension of their game

Limitations & Debates

• Carry definition inconsistency: different providers define and log carries differently. StatsBomb logs them explicitly; others require inference. The inferred method can create false carries (e.g., a player standing still but the ball bouncing nearby)
• No opposition context: a carry in open space and a carry under intense pressure look identical on a carry map
• Speed not captured in event data: a burst of pace and a slow jog with the ball are the same carry in event data. Tracking data would distinguish them but isn't available in most carry visualizations
• Overvaluing volume: players who carry a lot aren't necessarily carrying well — some players carry sideways or backward frequently. Filtering for progressive carries helps
• Complementary, not standalone: carry maps are most useful alongside Pass Maps|pass maps and Heat Maps|heat maps, not in isolation

Related Metrics

• Progressive Passes & Carries — the threshold-based framework for identifying valuable carries
• xT - Expected Threat — used to value carries by the threat gained
• VAEP - Valuing Actions — values carries as part of its comprehensive action valuation
• Packing Rate — alternative: measures defenders bypassed during a carry/dribble

Related Visualizations

• Pass Maps — the passing counterpart to carry maps
• Heat Maps — shows where a player is active but doesn't distinguish carries from other actions

---

Tags: #football #analytics #visualization #carries #dribbling #ball-progression