Chijioke
June 29, 2026
WORLD CUP 2026: GOAL SCORING EFFICIENCY AND DEFENSIVE CONTROL (Part 2)
Introduction
Part 1 of this series established the group-stage foundation of the World Cup 2026: Goal Scoring Efficiency and Defensive Efficiency model. It profiled all 48 teams by looking at how they created, converted, allowed, and prevented Big Chances.
Part 2 now shifts the focus to the 32 teams that progressed from the group stage. At this stage of tournament football, the margins become smaller, the opposition becomes stronger, and every decisive chance carries greater weight.
This part aims to profile the remaining teams using the same two connected models: Goal Scoring Efficiency and Defensive Efficiency. The attacking side examines Big Chances Created, Big Chances Scored, BCS%, and BCS Goal, while the defensive side examines Big Chances Allowed, Big Chances Conceded, Big Chances Prevented, and BC Con/GC%.
This part also introduces Expected Goals (xG) and Goals minus Expected Goals (G-xG) as an additional attacking layer. While Big Chances show how teams create and convert clear scoring opportunities, xG helps measure the overall quality of chances created. G-xG then shows whether a team scored more or fewer goals than expected from those chances.
A positive G-xG suggests that a team outperformed its expected goal value, often through strong finishing or efficient execution. A negative G-xG suggests underperformance, where a team created enough chance quality but failed to convert it into goals.
By combining Big Chances with xG and G-xG, Part 2 gives a sharper view of the teams left in the competition. It asks which teams are not only creating decisive chances, but also finishing above expectation, defending efficiently, and showing the kind of profile needed to survive knockout football.
NOTE !!!!
At the time of analysing and composing this report, Canada had successfully defeated South Africa by 1-0, on 28th June, 2026. That result had no impact on this analysis and report.
Method
Data Source and Analysis Tool
The data used for this analysis was sourced from FotMob, with team-level statistics collected for the 2026 FIFA World Cup.
The analysis was carried out using Microsoft Excel. The data was extracted, organised, cleaned, calculated, and analysed in Excel, where attacking and defensive tables were created. Pivot tables were also used to summarise the data, calculate key performance metrics, apply average and median benchmarks, and produce the charts used in the analysis.
For Part 2, the focus was placed on the 32 teams that progressed from the group stage. These teams were identified from the BCC and BCA Pivot Tables, which formed the basis of the attacking and defensive models.
The Expected Goals Table was also included to add another attacking layer through Expected Goals (xG) and Expected Goal Difference (G-xG). The xG data was filtered to include only the 32 teams present in both pivot tables.
Variables Used and Operational Definitions
The attacking model used the following variables:
- Goals Scored (GS): The total number of goals scored by each team.
- Big Chances Created (BCC): The total number of clear goal-scoring opportunities created by each team.
- Big Chances Missed (BCM): The number of big chances a team failed to convert.
- Big Chances Scored (BCS): The number of big chances converted into goals. This was calculated as: BCS = BCC − BCM
- Big Chances Scored Percentage (BCS%): The percentage of big chances created that were converted into goals. This was calculated as: BCS% = BCS ÷ BCC
- BCS Goal: The proportion of a team’s total goals that came from big chances. This was calculated as: BCS Goal = BCS ÷ GS
The Expected Goals layer used the following variables:
- Expected Goals (xG): The estimated goal value of the chances created by each team.
- Expected Goal Difference (G-xG): The difference between actual goals scored and expected goals. This was calculated as: G-xG = GS − xG
A positive Expected Goal Difference means that a team scored more goals than expected from the quality of chances created, while a negative Expected Goal Difference means that a team scored fewer goals than expected.
The defensive model used the following variables:
- Goals Conceded (GC): The total number of goals conceded by each team.
- Big Chances Allowed (BCA): The total number of big chances a team allowed the opposition to create.
- Big Chances Conceded (BC Con): The number of opposition big chances that were converted into goals.
- Big Chances Prevented (BC Prev): The number of opposition big chances that did not become goals. This was calculated as: BC Prev = BCA − BC Con
- BC Con/GC%: The proportion of a team’s total goals conceded that came from opposition big chances. This was calculated as: BC Con/GC% = BC Con ÷ GC
For teams that conceded zero goals, BC Con/GC% was treated as N/A, because the denominator was zero.
Benchmark Approach
Average and median values were used as benchmarks to provide context for the attacking, defensive, and Expected Goals metrics.
The average showed the general performance level across the 32 knockout-stage teams, while the median provided a more stable benchmark because tournament data can be influenced by very high or very low team values.
For the attacking model, benchmark lines were applied to BCC, BCS%, and BCS Goal. For the Expected Goals layer, benchmark lines were applied to Expected Goal Difference (G-xG). These benchmarks helped identify teams performing above or below the knockout-stage group.
Correlation Testing
Correlation testing was used to examine how strongly key performance variables related to goal outcomes among the 32 teams.
The first test examined the relationship between Goals Scored and Big Chances Created. The result produced an R² value of 0.45, meaning BCC explained about 45% of the variation in goals scored among the teams.
The second test examined the relationship between Goals Conceded and Big Chances Allowed. The result produced an R² value of 0.30, meaning BCA explained about 30% of the variation in goals conceded among the teams.
The third test examined the relationship between Goals Scored and Expected Goals. The result produced an R² value of 0.47, showing a moderate relationship between chance quality and actual goals scored.
Model Justification
The Part 2 model builds on the group-stage framework from Part 1 but narrows the focus to the 32 teams that reached the knockout stage.
The attacking model does not only ask which teams created chances. It also asks which teams converted them, how much of their goal output came from big chances, and whether their goals exceeded their expected goal value through Expected Goal Difference (G-xG).
The defensive model does not only ask which teams conceded goals. It also asks how many big chances they allowed, how many were converted, and how many were prevented.
This approach is important for knockout football because margins are smaller. Teams may create fewer chances and face stronger opposition, so efficiency becomes more important. A team’s ability to convert big chances, outperform expected goals, limit opposition danger, and prevent big chances from becoming goals becomes central to understanding its knockout-stage profile.
Big Chances Created, Big Chances Allowed and Expected Goal Models
Expected Goals and Expected Goal Difference Threshold
For the Expected Goals layer, Expected Goals (xG) and Expected Goal Difference (G-xG) were used to add more context to the attacking model.
While xG measures the quality of chances created, Expected Goal Difference shows whether a team scored more or fewer goals than expected from those chances. This was calculated as:
G-xG = Goals Scored − Expected Goals
To strengthen the analysis, both total values and per-90 values were considered. xG p90 was used to show the rate of expected goal creation, while G-xG p90 was used to show the rate at which teams outperformed or underperformed their expected goals.
For the 32 knockout-stage teams, the average Expected Goal Difference was +1.24, while the median was +1.05. The average Expected Goal Difference per 90 was +0.41, while the median was +0.35.
Teams were classified as follows:
Strong Positive Expected Goal Difference: Teams with a G-xG of +1.24 or higher, or a G-xG p90 of +0.41 or higher. These teams scored clearly above expectation and showed strong finishing efficiency.
Moderate Positive Expected Goal Difference: Teams with a G-xG between +1.05 and +1.23, or a G-xG p90 between +0.35 and +0.40. These teams scored above expectation, but not strongly above the knockout-stage average.
Slight Positive Expected Goal Difference: Teams with a G-xG between 0.00 and +1.04, or a G-xG p90 between 0.00 and +0.34. These teams scored more than expected, but their overperformance was below the knockout-stage median.
Negative Expected Goal Difference: Teams with a G-xG below 0.00, or a G-xG p90 below 0.00. These teams scored fewer goals than expected from their chance quality, suggesting underperformance in finishing.
Including the p90 values helped standardise the Expected Goals layer. It allowed the model to assess not only which teams outperformed expected goals overall, but also which teams were doing so at the strongest rate.
Tiering Framework
The 32 teams were grouped into five tiers based on their combined attacking, defensive, and Expected Goals profile.
The attacking model considered Big Chances Created, Big Chances Scored, BCS%, BCS Goal, Expected Goals (xG), Expected Goal Difference (G-xG), and their per-90 values. The defensive model considered Big Chances Allowed, Big Chances Conceded, Big Chances Prevented, and BC Con/GC%.
The tiering framework aimed to identify which teams had the strongest overall profile for knockout football, rather than ranking teams only by goals scored or goals conceded.
Tier 1: Potential World Cup Contenders
Tier 1 teams are the strongest overall profiles in the model. These teams combine strong attacking output with efficient finishing, positive Expected Goal Difference, and reliable defensive performance.
They are not only creating chances, but also converting them at a strong rate. Defensively, they show the ability to limit opposition danger, prevent big chances from becoming goals, or manage defensive pressure effectively.
These teams have the most complete profiles and are considered potential World Cup contenders.
Tier 2: Dark Horses
Tier 2 teams have strong enough profiles to trouble elite opposition and make a deep knockout-stage run, even if they are not the most complete teams in the model.
They may have one outstanding strength, such as strong finishing, high Expected Goal Difference, excellent defensive prevention, or a strong Big Chance profile. However, they may also carry one or two weaknesses that make them slightly less complete than the Tier 1 teams.
These teams are dangerous knockout opponents because they have enough efficiency or resilience to outperform expectations.
Tier 3: Competitive Knockout Teams
Tier 3 teams have balanced or respectable profiles, but without the same level of dominance or completeness shown by the top two tiers.
They may perform well in some areas, such as chance creation or defensive prevention, but lack consistency across the full attacking and defensive model. These teams can win knockout matches, but their profile suggests they may need strong execution, favourable match conditions, or improved efficiency to progress deeper.
Tier 4: High-Risk Qualifiers
Tier 4 teams progressed from the group stage but carry clear performance risks in the model.
These risks may include low chance creation, weak Big Chance conversion, negative Expected Goal Difference, high defensive exposure, or heavy reliance on preventing opposition chances from becoming goals. Their profiles suggest that they may struggle if they face stronger or more efficient opposition.
These teams can still compete, but their margin for error is smaller.
Tier 5: Fragile Knockout Profiles
Tier 5 teams have the weakest overall profiles among the 32 remaining teams.
They may have reached the knockout stage despite poor attacking efficiency, limited chance creation, weak Expected Goal Difference, or concerning defensive numbers. These teams are the most vulnerable in the model because their performances appear less sustainable when measured through Big Chances, defensive efficiency, and xG-based indicators.
For these teams, knockout progression would likely depend on major improvement, exceptional finishing, defensive resilience, or favourable match scenarios.
Team Profiling
1. Algeria
Attacking Profile: Algeria scored 5 goals, created 6 Big Chances at 2.00 BCC p90, and converted 3 Big Chances. Their BCS% was 50.0%, while 60.0% of their goals came from Big Chances. They recorded 3.8 xG, with an Expected Goal Difference (G-xG) of +1.2 and +0.40 G-xG p90.
Defending Profile: Algeria conceded 7 goals and allowed 4 Big Chances at 1.33 BCA p90. They conceded from 3 Big Chances and prevented only 1, giving them a BC Prev% of 25.0%. Their BC Con/GC% was 42.9%.
Tier: Tier 4 — High-Risk Qualifiers.
Comment: Algeria’s attacking profile is decent, especially through Big Chance conversion and positive Expected Goal Difference. However, the defensive profile is a major concern. They did not allow a very high number of Big Chances, but when those chances came, they were often punished.
2. Argentina
Attacking Profile: Argentina scored 8 goals, created 7 Big Chances at 2.33 BCC p90, and converted 2 Big Chances. Their BCS% was 28.6%, while only 25.0% of their goals came from Big Chances. They recorded 5.7 xG, with an Expected Goal Difference (G-xG) of +2.3 and +0.77 G-xG p90.
Defending Profile: Argentina conceded only 1 goal and allowed 2 Big Chances at 0.67 BCA p90. They conceded from 1 Big Chance and prevented 1, giving them a BC Prev% of 50.0%. Their BC Con/GC% was 100.0%.
Tier: Tier 2 — Dark Horses.
Comment: The defending champion’s profile is unusual. Their goal output and Expected Goal Difference are strong, but their Big Chance conversion and BCS Goal values are low. They have been efficient overall, but much of their scoring has come outside Big Chance situations (maybe a certain player is a factor?).
3. Australia
Attacking Profile: Australia scored 2 goals, created only 1 Big Chance at 0.33 BCC p90, and converted that chance. Their BCS% was 100.0%, while 50.0% of their goals came from Big Chances. They recorded 2.1 xG, with an Expected Goal Difference (G-xG) of -0.1 and -0.03 G-xG p90.
Defending Profile: Australia conceded 2 goals and allowed 4 Big Chances at 1.33 BCA p90. They conceded from 1 Big Chance and prevented 3, giving them a BC Prev% of 75.0%. Their BC Con/GC% was 50.0%.
Tier: Tier 5 — Fragile Knockout Profiles.
Comment: Australia’s Big Chance conversion looks perfect, but it comes from a very small sample. Their attacking profile is fragile because they created very few clear chances, even though their defensive prevention was respectable.
4. Austria
Attacking Profile: Austria scored 6 goals, created 7 Big Chances at 2.33 BCC p90, and converted 2 Big Chances. Their BCS% was 28.6%, while 33.3% of their goals came from Big Chances. They recorded 3.7 xG, with an Expected Goal Difference (G-xG) of +2.3 and +0.77 G-xG p90.
Defending Profile: Austria conceded 6 goals and allowed 6 Big Chances at 2.00 BCA p90. They conceded from 2 Big Chances and prevented 4, giving them a BC Prev% of 66.7%. Their BC Con/GC% was 33.3%.
Tier: Tier 4 — High-Risk Qualifiers.
Comment: Austria’s Expected Goal Difference is strong, but their Big Chance conversion is weak. Defensively, conceding 6 goals places them in a risky position, even though their Big Chance prevention was fairly solid.
5. Belgium
Attacking Profile: Belgium scored 6 goals, created 7 Big Chances at 2.33 BCC p90, and converted 2 Big Chances. Their BCS% was 28.6%, while 33.3% of their goals came from Big Chances. They recorded 6.8 xG, with an Expected Goal Difference (G-xG) of -0.8 and -0.27 G-xG p90.
Defending Profile: Belgium conceded 2 goals and allowed 3 Big Chances at 1.00 BCA p90. They conceded from 0 Big Chances and prevented 3, giving them a BC Prev% of 100.0%. Their BC Con/GC% was 0.0%.
Tier: Tier 3 — Competitive Knockout Teams.
Comment: Belgium’s defensive profile is strong, but their attacking efficiency is a concern. Their xG suggests they created good chance quality, but the negative Expected Goal Difference and low BCS% show they did not convert enough of that value into goals.
6. Bosnia and Herzegovina
Attacking Profile: Bosnia and Herzegovina scored 5 goals, created only 3 Big Chances at 1.00 BCC p90, and converted 1 Big Chance. Their BCS% was 33.3%, while only 20.0% of their goals came from Big Chances. They recorded 1.9 xG, with an Expected Goal Difference (G-xG) of +3.1 and +1.03 G-xG p90.
Defending Profile: Bosnia and Herzegovina conceded 6 goals and allowed 9 Big Chances at 3.00 BCA p90. They conceded from 4 Big Chances and prevented 5, giving them a BC Prev% of 55.6%. Their BC Con/GC% was 66.7%.
Tier: Tier 5 — Fragile Knockout Profiles.
Comment: Bosnia and Herzegovina strongly outperformed their xG, but their underlying profile is fragile. They created few Big Chances, depended heavily on finishing above expectation, and were exposed defensively.
7. Brazil
Attacking Profile: Brazil scored 7 goals, created 12 Big Chances at 4.00 BCC p90, and converted 6 Big Chances. Their BCS% was 50.0%, while 85.7% of their goals came from Big Chances. They recorded 7.3 xG, with an Expected Goal Difference (G-xG) of -0.3 and -0.10 G-xG p90.
Defending Profile: Brazil conceded only 1 goal and allowed 3 Big Chances at 1.00 BCA p90. They conceded from 1 Big Chance and prevented 2, giving them a BC Prev% of 66.7%. Their BC Con/GC% was 100.0%.
Tier: Tier 1 — Potential World Cup Contenders.
Comment: Brazil’s profile is built on strong chance creation and heavy Big Chance involvement. Their negative Expected Goal Difference shows they did not fully outperform their chance quality, but their high xG, high BCC, and low goals conceded keep them in the contender category.
8. Canada
Attacking Profile: Canada scored 8 goals, created 11 Big Chances at 3.67 BCC p90, and converted 4 Big Chances. Their BCS% was 36.4%, while 50.0% of their goals came from Big Chances. They recorded 7.5 xG, with an Expected Goal Difference (G-xG) of +0.5 and +0.17 G-xG p90.
Defending Profile: Canada conceded 3 goals and allowed 5 Big Chances at 1.67 BCA p90. They conceded from 2 Big Chances and prevented 3, giving them a BC Prev% of 60.0%. Their BC Con/GC% was 66.7%.
Tier: Tier 2 — Dark Horses.
Comment: Canada’s attacking process is strong because their xG and Big Chance creation are both high. Their finishing is not elite, but the chance quality is strong. Defensively, they are not as secure as the top contenders, which keeps them in the dark horse category.
9. Cape Verde
Attacking Profile: Cape Verde scored 2 goals, created 3 Big Chances at 1.00 BCC p90, and converted 1 Big Chance. Their BCS% was 33.3%, while 50.0% of their goals came from Big Chances. They recorded 2.5 xG, with an Expected Goal Difference (G-xG) of -0.5 and -0.17 G-xG p90.
Defending Profile: Cape Verde conceded 2 goals and allowed 5 Big Chances at 1.67 BCA p90. They conceded from 2 Big Chances and prevented 3, giving them a BC Prev% of 60.0%. Their BC Con/GC% was 100.0%.
Tier: Tier 5 — Fragile Knockout Profiles.
Comment: Cape Verde’s profile shows limited attacking output and negative Expected Goal Difference. Their defensive numbers are not disastrous, but they do not have enough attacking strength in the model to look secure.
10. Colombia
Attacking Profile: Colombia scored 4 goals, created 8 Big Chances at 2.67 BCC p90, and converted 3 Big Chances. Their BCS% was 37.5%, while 75.0% of their goals came from Big Chances. They recorded 4.2 xG, with an Expected Goal Difference (G-xG) of -0.2 and -0.07 G-xG p90.
Defending Profile: Colombia conceded only 1 goal and allowed 2 Big Chances at 0.67 BCA p90. They conceded from 0 Big Chances and prevented 2, giving them a BC Prev% of 100.0%. Their BC Con/GC% was 0.0%.
Tier: Tier 2 — Dark Horses.
Comment: Colombia’s attacking output is not elite, but their defensive profile is strong. They limited opposition danger well and did not concede from Big Chances. If their finishing improves, their defensive base gives them a strong platform.
11. Croatia
Attacking Profile: Croatia scored 5 goals, created 5 Big Chances at 1.67 BCC p90, and converted 2 Big Chances. Their BCS% was 40.0%, while 40.0% of their goals came from Big Chances. They recorded 2.8 xG, with an Expected Goal Difference (G-xG) of +2.2 and +0.73 G-xG p90.
Defending Profile: Croatia conceded 5 goals and allowed 8 Big Chances at 2.67 BCA p90. They conceded from 2 Big Chances and prevented 6, giving them a BC Prev% of 75.0%. Their BC Con/GC% was 40.0%.
Tier: Tier 4 — High-Risk Qualifiers.
Comment: Croatia outperformed their xG, but their chance creation was limited and their defensive exposure was high. Their ability to prevent Big Chances helped them survive, but the profile is risky against stronger opponents.
12. DR Congo
Attacking Profile: DR Congo scored 4 goals, created 4 Big Chances at 1.33 BCC p90, and converted 3 Big Chances. Their BCS% was 75.0%, while 75.0% of their goals came from Big Chances. They recorded 3.6 xG, with an Expected Goal Difference (G-xG) of +0.4 and +0.13 G-xG p90.
Defending Profile: DR Congo conceded 3 goals and allowed only 2 Big Chances at 0.67 BCA p90. They conceded from 0 Big Chances and prevented 2, giving them a BC Prev% of 100.0%. Their BC Con/GC% was 0.0%.
Tier: Tier 3 — Competitive Knockout Teams.
Comment: DR Congo have strong efficiency and defensive prevention, but their attacking volume is limited. They can be difficult to beat because they do not allow many Big Chances, but they may need to create more to progress deeper.
13. Ecuador
Attacking Profile: Ecuador scored 2 goals, created 9 Big Chances at 3.00 BCC p90, and converted only 1 Big Chance. Their BCS% was 11.1%, while 50.0% of their goals came from Big Chances. They recorded 5.1 xG, with an Expected Goal Difference (G-xG) of -3.1 and -1.03 G-xG p90.
Defending Profile: Ecuador conceded 2 goals and allowed 4 Big Chances at 1.33 BCA p90. They conceded from 1 Big Chance and prevented 3, giving them a BC Prev% of 75.0%. Their BC Con/GC% was 50.0%.
Tier: Tier 4 — High-Risk Qualifiers.
Comment: Ecuador’s defensive profile is respectable, but their attacking underperformance is the biggest issue. They created enough chance quality, but their finishing was poor. This makes them a high-risk team despite solid defensive numbers.
14. Egypt
Attacking Profile: Egypt scored 5 goals, created 5 Big Chances at 1.67 BCC p90, and converted only 1 Big Chance. Their BCS% was 20.0%, while 20.0% of their goals came from Big Chances. They recorded 3.8 xG, with an Expected Goal Difference (G-xG) of +1.2 and +0.40 G-xG p90.
Defending Profile: Egypt conceded 3 goals and allowed 10 Big Chances at 3.33 BCA p90, the highest BCA total among the 32 teams. They conceded from 2 Big Chances and prevented 8, giving them a BC Prev% of 80.0%. Their BC Con/GC% was 66.7%.
Tier: Tier 4 — High-Risk Qualifiers.
Comment: Egypt’s profile is risky because they allowed too many Big Chances and converted very few of their own. Their defensive prevention was strong, but relying on that level of survival may not be sustainable against stronger knockout opponents.
15. England
Attacking Profile: England scored 6 goals, created 13 Big Chances at 4.33 BCC p90, and converted 4 Big Chances. Their BCS% was 30.8%, while 66.7% of their goals came from Big Chances. They recorded 6.1 xG, with an Expected Goal Difference (G-xG) of -0.1 and -0.03 G-xG p90.
Defending Profile: England conceded 2 goals and allowed 3 Big Chances at 1.00 BCA p90. They conceded from 1 Big Chance and prevented 2, giving them a BC Prev% of 66.7%. Their BC Con/GC% was 50.0%.
Tier: Tier 2 — Dark Horses.
Comment: England created chances at a contender level, but their Big Chance conversion was below the strongest teams. Their defensive profile is solid, and their high BCC gives them upside. The main question is whether they can become more clinical.
16. France
Attacking Profile: France scored 10 goals, created 12 Big Chances at 4.00 BCC p90, and converted 5 Big Chances. Their BCS% was 41.7%, while 50.0% of their goals came from Big Chances. They recorded 6.0 xG, with an Expected Goal Difference (G-xG) of +4.0 and +1.33 G-xG p90.
Defending Profile: France conceded 2 goals and allowed 7 Big Chances at 2.33 BCA p90. They conceded from only 1 Big Chance and prevented 6, giving them a BC Prev% of 85.7%. Their BC Con/GC% was 50.0%.
Tier: Tier 1 — Potential World Cup Contenders.
Comment: France combine elite goal output with strong finishing overperformance. Their defensive exposure was higher than some other contenders, but their ability to prevent Big Chances from becoming goals was excellent.
17. Germany
Attacking Profile: Germany scored 10 goals, created 13 Big Chances at 4.33 BCC p90, and converted 7 Big Chances. Their BCS% was 53.8%, while 70.0% of their goals came from Big Chances. They recorded 6.8 xG, with an Expected Goal Difference (G-xG) of +3.2 and +1.07 G-xG p90.
Defending Profile: Germany conceded 4 goals and allowed 4 Big Chances at 1.33 BCA p90. They conceded from 2 Big Chances and prevented 2, giving them a BC Prev% of 50.0%. Their BC Con/GC% was 50.0%.
Tier: Tier 1 — Potential World Cup Contenders.
Comment: Germany’s profile is strong because their attacking numbers are supported by both volume and efficiency. They created at a high level, converted well, and significantly outperformed their xG. Their defensive numbers are not perfect, but their attacking power gives them a strong contender profile.
18. Ghana
Attacking Profile: Ghana scored 2 goals, created 4 Big Chances at 1.33 BCC p90, and converted 2 Big Chances. Their BCS% was 50.0%, while 100.0% of their goals came from Big Chances. They recorded 2.1 xG, with an Expected Goal Difference (G-xG) of -0.1 and -0.03 G-xG p90.
Defending Profile: Ghana conceded 2 goals and allowed 6 Big Chances at 2.00 BCA p90. They conceded from 1 Big Chance and prevented 5, giving them a BC Prev% of 83.3%. Their BC Con/GC% was 50.0%.
Tier: Tier 4 — High-Risk Qualifiers.
Comment: Ghana prevented Big Chances well, but their attacking volume was limited. Their goals came entirely from Big Chances, which shows dependency, but they did not create enough to look secure going into knockout football.
19. Ivory Coast
Attacking Profile: Ivory Coast scored 4 goals, created 7 Big Chances at 2.33 BCC p90, and converted 3 Big Chances. Their BCS% was 42.9%, while 75.0% of their goals came from Big Chances. They recorded 4.0 xG, with an Expected Goal Difference (G-xG) of 0.0 and 0.00 G-xG p90.
Defending Profile: Ivory Coast conceded 2 goals and allowed 3 Big Chances at 1.00 BCA p90. They conceded from 1 Big Chance and prevented 2, giving them a BC Prev% of 66.7%. Their BC Con/GC% was 50.0%.
Tier: Tier 3 — Competitive Knockout Teams.
Comment: Ivory Coast have a balanced profile without being dominant. Their Big Chance dependency is high, and their defensive numbers are respectable. However, the neutral Expected Goal Difference suggests they are performing close to expectation rather than above it.
20. Japan
Attacking Profile: Japan scored 7 goals, created 6 Big Chances at 2.00 BCC p90, and converted 3 Big Chances. Their BCS% was 50.0%, while 42.9% of their goals came from Big Chances. They recorded 3.9 xG, with an Expected Goal Difference (G-xG) of +3.1 and +1.03 G-xG p90.
Defending Profile: Japan conceded 3 goals but allowed only 1 Big Chance at 0.33 BCA p90. They conceded from 0 Big Chances and prevented 1, giving them a BC Prev% of 100.0%. Their BC Con/GC% was 0.0%.
Tier: Tier 2 — Dark Horses.
Comment: Japan’s profile is very efficient. They did not create a large number of Big Chances, but they strongly outperformed xG and allowed the fewest Big Chances among the 32 teams. Their profile makes them a dangerous knockout opponent.
21. Mexico
Attacking Profile: Mexico scored 6 goals, created 9 Big Chances at 3.00 BCC p90, and converted 6 Big Chances. Their BCS% was 66.7%, while 100.0% of their goals came from Big Chances. They recorded 3.7 xG, with an Expected Goal Difference (G-xG) of +2.3 and +0.77 G-xG p90.
Defending Profile: Mexico conceded 0 goals and allowed only 3 Big Chances at 1.00 BCA p90. They conceded from 0 Big Chances and prevented 3, giving them a BC Prev% of 100.0%. Their BC Con/GC% was N/A because they conceded no goals.
Tier: Tier 1 — Potential World Cup Contenders.
Comment: Mexico have one of the cleanest efficiency profiles in the model. Their attacking output came entirely from Big Chances, while defensively they did not concede. Their profile shows strong execution at both ends of the pitch.
22. Morocco
Attacking Profile: Morocco scored 6 goals, created 10 Big Chances at 3.33 BCC p90, and converted 5 Big Chances. Their BCS% was 50.0%, while 83.3% of their goals came from Big Chances. They recorded 6.1 xG, with an Expected Goal Difference (G-xG) of -0.1 and -0.03 G-xG p90.
Defending Profile: Morocco conceded 3 goals and allowed only 3 Big Chances at 1.00 BCA p90. They conceded from 0 Big Chances and prevented 3, giving them a BC Prev% of 100.0%. Their BC Con/GC% was 0.0%.
Tier: Tier 1 — Potential World Cup Contenders.
Comment: Morocco’s profile is strong because they combine good chance creation with excellent defensive prevention. Their Expected Goal Difference is slightly negative, but their Big Chance conversion and defensive control make them a serious contender profile.
23. Netherlands
Attacking Profile: The Netherlands scored 10 goals, created 5 Big Chances at 1.67 BCC p90, and converted 4 Big Chances. Their BCS% was 80.0%, while 40.0% of their goals came from Big Chances. They recorded 5.2 xG, with an Expected Goal Difference (G-xG) of +4.8 and +1.60 G-xG p90.
Defending Profile: The Netherlands conceded 4 goals and allowed 4 Big Chances at 1.33 BCA p90. They conceded from 2 Big Chances and prevented 2, giving them a BC Prev% of 50.0%. Their BC Con/GC% was 50.0%.
Tier: Tier 2 — Dark Horses.
Comment: The Netherlands are dangerous because their finishing has been exceptional. Their Big Chance volume is not high, but their Expected Goal Difference is one of the strongest in the model. The concern is whether this level of finishing overperformance can continue.
24. Norway
Attacking Profile: Norway scored 8 goals, created 14 Big Chances at 4.67 BCC p90, the highest BCC total among the 32 teams, and converted 6 Big Chances. Their BCS% was 42.9%, while 75.0% of their goals came from Big Chances. They recorded 6.4 xG, with an Expected Goal Difference (G-xG) of +1.6 and +0.53 G-xG p90.
Defending Profile: Norway conceded 7 goals and allowed 7 Big Chances at 2.33 BCA p90. They conceded from 3 Big Chances and prevented 4, giving them a BC Prev% of 57.1%. Their BC Con/GC% was 42.9%.
Tier: Tier 2 — Dark Horses.
Comment: Norway are one of the most dangerous attacking teams in the model, but their defensive numbers are a concern. Their high BCC, strong BCS Goal, and positive Expected Goal Difference make them dangerous, but conceding 7 goals makes them a high-variance dark horse.
25. Paraguay
Attacking Profile: Paraguay scored 2 goals, created only 1 Big Chance at 0.33 BCC p90, and converted that chance. Their BCS% was 100.0%, while 50.0% of their goals came from Big Chances. They recorded 1.1 xG, with an Expected Goal Difference (G-xG) of +0.9 and +0.30 G-xG p90.
Defending Profile: Paraguay conceded 4 goals and allowed 9 Big Chances at 3.00 BCA p90. They conceded from 2 Big Chances and prevented 7, giving them a BC Prev% of 77.8%. Their BC Con/GC% was 50.0%.
Tier: Tier 5 — Fragile Knockout Profiles.
Comment: Paraguay’s attacking profile is extremely limited. Their defensive prevention helped them survive, but allowing 9 Big Chances shows a high level of exposure. They are likely to struggle if forced to create more chances.
26. Portugal
Attacking Profile: Portugal scored 6 goals, created 8 Big Chances at 2.67 BCC p90, and converted 2 Big Chances. Their BCS% was 25.0%, while 33.3% of their goals came from Big Chances. They recorded 4.0 xG, with an Expected Goal Difference (G-xG) of +2.0 and +0.67 G-xG p90.
Defending Profile: Portugal conceded only 1 goal and allowed 3 Big Chances at 1.00 BCA p90. They conceded from 1 Big Chance and prevented 2, giving them a BC Prev% of 66.7%. Their BC Con/GC% was 100.0%.
Tier: Tier 3 — Competitive Knockout Teams.
Comment: Portugal have a strong defensive base and a positive Expected Goal Difference, but their Big Chance conversion is weak. Their profile suggests they can compete, but their attacking efficiency from clear chances needs improvement.
27. Senegal
Attacking Profile: Senegal scored 8 goals, created 10 Big Chances at 3.33 BCC p90, and converted 4 Big Chances. Their BCS% was 40.0%, while 50.0% of their goals came from Big Chances. They recorded 5.3 xG, with an Expected Goal Difference (G-xG) of +2.7 and +0.90 G-xG p90.
Defending Profile: Senegal conceded 6 goals and allowed 9 Big Chances at 3.00 BCA p90. They conceded from 4 Big Chances and prevented 5, giving them a BC Prev% of 55.6%. Their BC Con/GC% was 66.7%.
Tier: Tier 3 — Competitive Knockout Teams.
Comment: Senegal have enough attacking power to trouble opponents, but their defensive exposure is high. Their positive Expected Goal Difference is a strength, but allowing 9 Big Chances and conceding 6 goals makes their knockout profile less secure.
28. South Africa
Attacking Profile: South Africa scored 2 goals, created 2 Big Chances at 0.67 BCC p90, and converted 1 Big Chance. Their BCS% was 50.0%, while 50.0% of their goals came from Big Chances. They recorded 2.6 xG, with an Expected Goal Difference (G-xG) of -0.6 and -0.20 G-xG p90.
Defending Profile: South Africa conceded 3 goals and allowed 6 Big Chances at 2.00 BCA p90. They conceded from 3 Big Chances and prevented 3, giving them a BC Prev% of 50.0%. Their BC Con/GC% was 100.0%.
Tier: Tier 5 — Fragile Knockout Profiles.
Comment: South Africa’s profile is fragile because they combined low chance creation with negative Expected Goal Difference and moderate defensive exposure. Their margin for error is small.
29. Spain
Attacking Profile: Spain scored 5 goals, created 8 Big Chances at 2.67 BCC p90, and converted 3 Big Chances. Their BCS% was 37.5%, while 60.0% of their goals came from Big Chances. They recorded 5.3 xG, with an Expected Goal Difference (G-xG) of -0.3 and -0.10 G-xG p90.
Defending Profile: Spain conceded 0 goals and allowed only 2 Big Chances at 0.67 BCA p90. They conceded from 0 Big Chances and prevented 2, giving them a BC Prev% of 100.0%. Their BC Con/GC% was N/A because they conceded no goals.
Tier: Tier 1 — Potential World Cup Contenders.
Comment: Spain’s attacking efficiency is not as strong as the other Tier 1 teams, but their defensive profile is elite. Conceding no goals while allowing very few Big Chances gives them a strong knockout foundation. Their main concern is finishing efficiency.
30. Sweden
Attacking Profile: Sweden scored 7 goals, created 7 Big Chances at 2.33 BCC p90, and converted 3 Big Chances. Their BCS% was 42.9%, while 42.9% of their goals came from Big Chances. They recorded 3.0 xG, with an Expected Goal Difference (G-xG) of +4.0 and +1.33 G-xG p90.
Defending Profile: Sweden conceded 7 goals and allowed 5 Big Chances at 1.67 BCA p90. They conceded from 4 Big Chances and prevented only 1, giving them a BC Prev% of 20.0%. Their BC Con/GC% was 57.1%.
Tier: Tier 4 — High-Risk Qualifiers.
Comment: Sweden’s attacking overperformance is excellent, but the defensive numbers are fragile. Their low Big Chance prevention rate means they are vulnerable when opponents create clear chances.
31. Switzerland
Attacking Profile: Switzerland scored 7 goals, created 13 Big Chances at 4.33 BCC p90, and converted 5 Big Chances. Their BCS% was 38.5%, while 71.4% of their goals came from Big Chances. They recorded 6.4 xG, with an Expected Goal Difference (G-xG) of +0.6 and +0.20 G-xG p90.
Defending Profile: Switzerland conceded 3 goals and allowed 4 Big Chances at 1.33 BCA p90. They conceded from 1 Big Chance and prevented 3, giving them a BC Prev% of 75.0%. Their BC Con/GC% was 33.3%.
Tier: Tier 2 — Dark Horses.
Comment: Switzerland have a strong creation profile and a solid defensive base. Their finishing is not as clinical as the top contenders, but their high BCC and strong defensive prevention make them capable of troubling stronger opposition.
32. USA
Attacking Profile: USA scored 8 goals, created 10 Big Chances at 3.33 BCC p90, and converted 4 Big Chances. Their BCS% was 40.0%, while 50.0% of their goals came from Big Chances. They recorded 4.6 xG, with an Expected Goal Difference (G-xG) of +3.4 and +1.13 G-xG p90.
Defending Profile: USA conceded 4 goals and allowed 6 Big Chances at 2.00 BCA p90. They conceded from 4 Big Chances and prevented only 2, giving them a BC Prev% of 33.3%. Their BC Con/GC% was 100.0%.
Tier: Tier 3 — Competitive Knockout Teams.
Comment: USA’s attacking profile is strong, especially through Expected Goal Difference. However, their defensive numbers are risky. Conceding from 4 Big Chances shows that opposition clear chances are becoming goals too often.
Discussion
Part 2 shows that knockout-stage profiling requires more than simply looking at goals scored or goals conceded. Across the 32 teams analysed, there were 183 goals scored, 240 Big Chances Created, and 100 Big Chances Scored. This gave the knockout-stage group an average of 5.72 goals, 7.50 Big Chances Created, and 3.13 Big Chances Scored per team.
The attacking benchmarks helped separate the stronger profiles from the weaker ones. The average BCS% was 45.5%, while the median was 40.8%. This means teams converting above this range were performing above the general knockout-stage level. The average BCS Goal value was 55.6%, with a median of 50.0%, showing that, for most teams, at least half of their goals were connected to Big Chance conversion.
Chance creation was also important. Norway created the most Big Chances with 14, followed by Germany, England, and Switzerland with 13 each. France and Brazil followed with 12 each. However, creation alone was not enough. Ecuador created 9 Big Chances but scored only 2 goals, with an Expected Goal Difference of -3.1, showing that poor finishing can weaken an otherwise promising attacking process.
The Expected Goals layer added more context to attacking performance. Across the 32 teams, the average xG was 4.48, while the median was 4.10. The average Expected Goal Difference (G-xG) was +1.24, with a median of +1.05. On a per-90 basis, the average G-xG p90 was +0.41, while the median was +0.35. These figures showed which teams were not only creating chance quality, but also finishing above expectation.
The strongest Expected Goal Difference profiles came from the Netherlands (+4.8), France (+4.0), Sweden (+4.0), USA (+3.4), Germany (+3.2), and Bosnia and Herzegovina (+3.1). However, high G-xG also needed context. For example, the Netherlands showed exceptional finishing efficiency, but created only 5 Big Chances, while Germany combined a strong G-xG with 13 Big Chances Created, making their attacking profile more sustainable.
The correlation tests also supported the model. The relationship between Goals Scored and Big Chances Created produced an R² value of 0.45, meaning Big Chances Created explained about 45% of the variation in goals scored. The relationship between Goals Scored and Expected Goals produced an R² value of 0.47, showing a similar moderate relationship between chance quality and goal output. These results suggest that creation and chance quality matter, but finishing efficiency still plays a major role.
Defensively, the 32 teams conceded 101 goals and allowed 152 Big Chances. The average team conceded 3.16 goals and allowed 4.75 Big Chances, with a median of 3 goals conceded and 4 Big Chances Allowed. The average BCA p90 was 1.58, while the median was 1.33.
The defensive model showed that limiting Big Chances was a key separator. Japan allowed only 1 Big Chance, while Argentina, Colombia, DR Congo, and Spain allowed only 2 each. At the other end, Egypt allowed 10 Big Chances, while Bosnia and Herzegovina, Paraguay, and Senegal allowed 9 each. This created clear differences in defensive risk.
Big Chance prevention also mattered. The average BC Prev% was 69.6%, while the median was 66.7%. Belgium, Colombia, DR Congo, Japan, Mexico, Morocco, and Spain all recorded 100.0% BC Prev%, meaning they prevented every Big Chance they allowed from becoming a goal. France also performed strongly with 85.7%, while Ghana recorded 83.3% and Egypt 80.0%.
However, the defensive correlation was weaker than the attacking model. The relationship between Goals Conceded and Big Chances Allowed produced an R² value of 0.30, meaning Big Chances Allowed explained about 30% of the variation in goals conceded. This showed that defensive outcomes were not only about how many Big Chances teams allowed, but also about whether those chances were converted, prevented, or saved.
A key numerical point from the knockout-stage group was the African representation. Nine of the 32 teams were African representatives: Algeria, Cape Verde, DR Congo, Egypt, Ghana, Ivory Coast, Morocco, Senegal, and South Africa. This represented 28.1% of the knockout-stage teams. In past World Cups, African progress was often remembered through individual landmark teams, such as Cameroon in 1990, Senegal in 2002, Ghana in 2010, and Morocco in 2022. In this analysis, the difference was numerical depth: African representation was spread across 9 teams, rather than being carried by one standout team alone.
Within that African group, the profiles varied. Morocco had the strongest African profile and were placed in Tier 1, supported by 10 Big Chances Created, 50.0% BCS%, 83.3% BCS Goal, only 3 Big Chances Allowed, and 100.0% BC Prev%. Senegal were strong offensively, scoring 8 goals with 10 Big Chances Created and a +2.7 G-xG, but they also allowed 9 Big Chances and conceded 6 goals (They need to work on that to survive). DR Congo and Ivory Coast had competitive profiles, especially through defensive prevention, while Algeria, Egypt, Ghana, Cape Verde, and South Africa carried clearer attacking or defensive risks.
Overall, the Part 2 model shows that the strongest knockout-stage teams were not always the teams with the most goals alone. The more complete profiles were built on a combination of Big Chance creation, finishing efficiency, Expected Goal Difference, defensive control, and Big Chance prevention. Teams such as Germany, France, Brazil, Mexico, Morocco, and Spain stood out because they showed strength in several of these areas, while other teams had one strong area but carried clear risks elsewhere.
Conclusion
Part 2 of the World Cup 2026: Goal Scoring Efficiency and Defensive Efficiency series narrowed the analysis to the 32 teams that progressed from the group stage. The aim was to understand which teams had the strongest knockout-stage profiles when attacking efficiency, defensive efficiency, Big Chances, Expected Goals, and Expected Goal Difference were considered together.
Across the 32 teams, there were 183 goals scored, 240 Big Chances Created, and 100 Big Chances Scored. This produced an average of 7.50 Big Chances Created per team, with an average BCS% of 45.5% and an average BCS Goal value of 55.6%. These figures showed that Big Chances remained a major part of attacking output, but they did not explain everything on their own.
The Expected Goals layer added another important view. The average Expected Goal Difference (G-xG) was +1.24, while the average G-xG p90 was +0.41. This helped separate teams that were simply creating chances from teams that were finishing above expectation.
Defensively, the 32 teams conceded 101 goals and allowed 152 Big Chances. The average team allowed 4.75 Big Chances, while the average BC Prev% was 69.6%. Teams that limited Big Chances and prevented them from becoming goals had stronger knockout profiles, especially Spain, Mexico, Morocco, Belgium, Colombia, DR Congo, and Japan.
The correlation results also supported the model. Goals Scored vs Big Chances Created produced an R² of 0.45, while Goals Scored vs Expected Goals produced an R² of 0.47. Defensively, Goals Conceded vs Big Chances Allowed produced an R² of 0.30. These results showed that chance creation and chance quality mattered, but finishing efficiency, defensive prevention, and match context still shaped outcomes.
African representation was also significant. Nine of the 32 knockout-stage teams were African representatives, accounting for 28.1% of the teams analysed. This showed strong numerical presence in the knockout stage, although the profiles varied across the teams.
Based on the bracket route, one potential semi-final from the left side of the bracket is Germany vs Spain, a matchup between Germany’s attacking volume and Spain’s defensive control. From the right side of the bracket, one potential semi-final is Brazil vs Argentina, a matchup that would bring together Brazil’s strong Big Chance profile and Argentina’s efficient goal output.
Overall, Part 2 shows that the strongest knockout-stage teams were not simply the teams that scored the most goals. The most complete profiles belonged to teams that could create and convert Big Chances, outperform or match their Expected Goals, limit opposition Big Chances, and prevent clear chances from becoming goals. Based on the model, Germany, France, Brazil, Mexico, Morocco, and Spain stood out as the strongest contender-level profiles heading deeper into the tournament.