04 Nov Who builds up best in LaLiga? It depends on what you count
Athletic Club ranks 18th in LaLiga for low build-ups that reach the final third: 2.4 per match. Ask the question slightly differently, what share of its low build-ups get there, and it ranks 4th, at 21.8%. Elche does the opposite: 5th by count (3.7 per match) and 14th by share (18.8%).
Neither ranking is wrong. They answer different questions, and across the 20 teams they agree only partly (Spearman correlation 0.50).
This post follows one family of metrics through three choices that reshape a build-up ranking: what you divide by, what you include, and what the definition lets you see. It ends with a five-question checklist and a file for each of the 20 teams, in final-table order.
What counts as a build-up here
In this data, a build-up is a controlled passing sequence in which the opposition’s midfield and defence stay behind the ball for at least eight seconds. Each one is classified as low (a phase taking place mainly in the team’s own defensive third) or mid (mainly in the middle third), and the two always add up to the total.
Two details matter for everything below. «Low» describes where a phase mainly takes place, not where it starts. And a sequence shorter than eight seconds is not a build-up, so fast direct play barely shows up.
A build-up «reaches» a zone when the phase progresses until it. Reaching the box implies reaching the final third, which implies reaching the opposition half. The analysis covers 760 team-match records from the 2025-26 season.
Lever 1: what you divide by
A count and a rate answer different questions, and the gap between them is how many build-ups a team has. Counted per match, low build-ups reaching the final third rank Barcelona first (4.3) and Getafe last (1.4). Divided by each team’s low build-ups, the same numerator ranks Atlético first (24.7%) and Mallorca last (16.3%).
Low build-ups per match range from 7.9 (Getafe) to 20.6 (Girona), a 2.6-fold spread. Conversion rates range from 16.3% to 24.7%, only 1.5-fold. A count is mostly a measure of opportunity.
Athletic shows it best. Only 31.7% of its build-ups take place mainly in its own third, the lowest in LaLiga, so it has 11.0 low build-ups per match against 18.3 for Barcelona. Few opportunities and high conversion: 4th by rate, 18th by count. Elche and Girona sit at the other end, with 19.6 and 20.6 low build-ups per match but average conversion (18.8% and 19.9%).
Dividing by attack phases instead of matches barely matters. Teams have between 132 and 173 attack phases per match, so that ranking stays close to the per-match one (Spearman 0.93). The choice that matters is dividing by build-ups.
The effect is strongest for the final third from low build-ups. For mid build-ups reaching the final third, count and rate agree at 0.77, and for low build-ups reaching the box at 0.71. Report both, or say which one you used.
Lever 2: what you include
Whether short goal kicks count changes who looks most goalkeeper-centric. The metric exists in two versions in the data, with and without them, and the team rankings agree only at 0.81. The version with short goal kicks is higher for every team, and the extra appears only in low build-ups, which is consistent with build-ups that start from a short goal kick.
Teams differ in how much of their goalkeeper involvement comes from restarts: the gap runs from 2.5 percentage points (Alavés) to 13.9 (Levante). Include them and the metric partly describes goal-kick style. Exclude them and it describes involvement in open play.
Athletic moves from 16th to 8th, Getafe from 13th to 7th and Alavés from 20th to 15th. Villarreal and Betis fall five places each (14th to 19th and 7th to 12th). Elche is first either way, at 61.6% with short goal kicks and 51.6% without.
The analysis behind these team files uses the version without short goal kicks. If the question is how a team restarts, use both and report the gap.
Lever 3: what the definition lets you see
Every build-up metric inherits the limits of its definition, and four of them change how a ranking should be read.
- Eight seconds. Direct play that never strings eight seconds of controlled passing does not appear as a build-up. Getafe spends 6.8% of match time in build-ups against a league average of 13.4%, and has 22.5 build-ups per match against Barcelona’s 57.2. That describes how often it circulates the ball, not how much of it it has.
- «Low» is not «starts low». A phase counts as low when it takes place mainly in the team’s own defensive third. Levante has the highest share (50.2% of its build-ups) and Athletic the lowest (31.7%); both describe where phases are played, not where they start.
- Long passes. Own-half losses split exactly into two types: dispossession or a blocked or intercepted pass, and a loss after attempting a long pass. A long pass that carries the ball into the opposition half counts as progress even if the ball is then lost, so long-pass losses are a floor for direct play, not a measure of it.
- Unclassified endings. Add up the build-ups that reach the opposition half, those lost in the team’s own half and those that end with a foul won there, and about three per match and type are left over. The data does not say how they end.
One definition choice does not matter. Measuring time in build-up as a share of the whole match or as a share of the team’s attacking time gives almost the same ranking (Spearman 0.97).
The 20 team files
Each card shows a team’s season profile and, in color, the estimated link between every metric and winning rather than losing. The captions add the lens of this post: where the team stands by volume and by rate, and how its goalkeeper-touch rank changes without short goal kicks. Teams follow the final table.
1 · Barcelona (94 pts)
Count and rate agree: 1st per match (4.3) and 2nd as a share of low build-ups (23.8%). The volume comes from 57.2 build-ups per match, the most in LaLiga. Goalkeeper touch: 18th with and without short goal kicks.
2 · Real Madrid (86 pts)
Rate beats count: 3rd as a share (23.2%), 4th per match (3.9). It has fewer low build-ups than the top teams (16.8 per match, 12th), so conversion carries the ranking. Goalkeeper touch: 15th with short goal kicks, 17th without.
3 · Villarreal (72 pts)
Mid-table on both: 8th per match (3.4) and 7th as a share (20.1%). Its goalkeeper-touch rank falls from 14th to 19th without short goal kicks (41.9% to 28.9%), one of the largest drops.
4 · Atlético Madrid (69 pts)
The best rate in LaLiga: 24.7% of low build-ups reach the final third (1st), and 4.3 per match (2nd). With 17.3 low build-ups per match (8th), the ranking rests on conversion, not volume.
5 · Real Betis (60 pts)
Plenty of opportunities, low conversion: 17.7 low build-ups per match (7th), but only 17.4% reach the final third (19th). Per match it is 13th (3.1). Goalkeeper touch: 7th with short goal kicks, 12th without.
6 · Celta Vigo (54 pts)
Volume, not rate: 7th per match (3.6) and 12th as a share (19.3%), helped by 18.4 low build-ups per match (3rd). Its goalkeeper touch is 19th with short goal kicks (35.4%) and last without them (26.2%).
7 · Getafe (51 pts)
Last in the count (1.4 per match) because it has the fewest low build-ups (7.9 per match). As a share it is 18th (17.7%), much closer to the pack. Direct play that never reaches eight seconds of passing does not show up here.
8 · Rayo Vallecano (50 pts)
A stable middle: 6th per match (3.6) and 9th as a share (19.9%). Its goalkeeper touch is 3rd with short goal kicks (52.5%) and 6th without (39.7%), a 12.8-point gap.
9 · Valencia (49 pts)
11th per match (3.2) and 15th as a share (18.7%). Without short goal kicks its goalkeeper-touch share falls 11.2 points, from 9th to 13th.
10 · Real Sociedad (46 pts)
Below average on both: 14th per match (2.9) and 17th as a share (17.9%). Its goalkeeper-touch rank holds up whichever version you use: 4th with short goal kicks, 3rd without.
11 · Espanyol (46 pts)
A rate story: 16th per match (2.7) but 5th as a share (20.8%). It has few low build-ups (13.1 per match, 17th) and converts them well.
12 · Athletic Club (45 pts)
The widest gap in LaLiga: 18th per match (2.4) but 4th as a share (21.8%). Only 31.7% of its build-ups happen mainly in its own third, so it has 11.0 low build-ups per match to convert. Goalkeeper touch: 16th with short goal kicks, 8th without.
13 · Sevilla (43 pts)
Consistent: 12th per match (3.1) and 10th as a share (19.8%). It loses only 5.5 points of goalkeeper touch when short goal kicks are excluded, so its rank rises from 8th to 4th.
14 · Alavés (43 pts)
19th per match (2.1) but 11th as a share (19.3%), with only 10.9 low build-ups per match (19th). Its goalkeeper involvement is almost all open play: removing short goal kicks costs 2.5 points and lifts its rank from 20th to 15th.
15 · Elche (43 pts)
5th per match (3.7) but 14th as a share (18.8%): the second-most low build-ups in LaLiga (19.6 per match), converted below average. First in goalkeeper touch with (61.6%) and without (51.6%) short goal kicks.
16 · Levante (42 pts)
8th per match (3.4) and 6th as a share (20.3%). Its goalkeeper involvement leans most on short goal kicks: 53.8% with them, 39.9% without, the largest gap in LaLiga (13.9 points), and its rank falls from 2nd to 5th.
17 · Osasuna (42 pts)
Below the middle on both: 15th per match (2.7) and 13th as a share (18.9%), with 14.5 low build-ups per match (16th). Goalkeeper touch: 11th with short goal kicks, 10th without.
18 · Mallorca (42 pts)
The lowest rate in LaLiga: 16.3% of low build-ups reach the final third (20th). Volume partly compensates, at 16th per match (2.7) with 16.6 low build-ups per match (13th). Goalkeeper touch: 6th with short goal kicks, 9th without.
19 · Girona (41 pts)
The most low build-ups in LaLiga (20.6 per match) put it 3rd per match (4.1), but conversion is average: 8th as a share (19.9%). Goalkeeper touch: 12th with short goal kicks, 14th without.
20 · Real Oviedo (29 pts)
10th per match (3.3) and 16th as a share (18.1%). Its 18.2 low build-ups per match (6th) prop up the count. Goalkeeper touch: 5th with short goal kicks, 2nd without (40.9%).
Five questions before trusting a build-up ranking
- What is it divided by? A count measures opportunity and a share measures conversion. For low build-ups reaching the final third, the two rankings agree at 0.50.
- What does the metric include? Short goal kicks, the eight-second minimum and the meaning of «low» all change what is being ranked. Goalkeeper touch moved Athletic eight places depending on one of them.
- What can the definition not see? Direct play that never forms a build-up, long passes that end in the opposition half, and the endings the data does not classify.
- Is the gap bigger than the noise? Neighbouring teams in the per-match ranking above differ by a median of 0.13, while each team’s season average carries a standard error of about 0.3 per match. For 18 of the 19 neighbouring pairs, the gap is smaller than its own standard error. Read ranks as tiers, not as positions.
- Does the claim describe or explain? A ranking describes. Saying why a team wins or loses needs a different design, and it still has to face the scoreline: teams play differently when ahead or behind.
Limits
This is one season (38 matches per team, 760 team-match records) of build-up data, with no goals or match minutes, so it cannot separate style from the scoreline. The post does not say which ranking is the right one. It shows which question each one answers, and how far a ranking moves when the question changes.
A companion analysis looks at how these same metrics relate to winning and losing within each team. It reaches a similar conclusion from a different angle: most apparent team-specific stories do not survive a check for noise.