SPORTS ANALYTICS CONSULTING · PROCESS DEMONSTRATION

Entirely fictional data. This is a reporting example, not a client case study, forecast or evidence of service outcomes. Two errors were deliberately inserted.

Better results. Stronger performance?

A review of 12 fictional matches shows why a report should distinguish scoreboard improvement from evidence about how a team is playing.

Results improved, but these data do not establish better underlying performance. Points rose from 5 to 16 across equal six-match windows. Average shot differential stayed close to zero. One late-period match lacks opponent shots, so the shot comparison uses unequal coverage.

1. Compare the result with the supporting evidence

Measure	Matches 1–6	Matches 7–12
League points (3 win / 1 draw / 0 loss)	5 from 6 matches	16 from 6 matches
Points per match	0.83	2.67
Average shot differential	-0.33 (6 valid matches)	+0.00 (5 valid matches)

Shot differential means shots for minus shots against. It does not measure shot quality, opponent strength or tactical execution. No expected-goals values were invented. These fictional match IDs represent sequence, not actual dates.

2. Flag errors before interpreting the report

Match	Field	Finding and handling
11	on_target_for	Exceeds total shots; field excluded pending correction
12	shots_against	Missing; excluded from dependent metrics

Only the affected fields are excluded: match 11 still contributes valid total shots and goals. Match 12 contributes goals and points, but not shot differential. Nothing is silently replaced with zero. Request corrections from the source owner before issuing a final client report.

3. Give the analyst a defensible conclusion

“The later six matches produced more points. The available shot counts show little descriptive change in shot balance, and one match is incomplete. This short sequence cannot distinguish improvement from finishing variation, opponent differences or other influences.”

Next useful work: correct both source errors; examine authorized shot-quality data if available; review opponent, venue and score-state context with the soccer analyst. More goals alone do not justify a causal or predictive claim.

Match-level audit trail

Match	Shots for	Shots against	Goals for–against
1	10	12	0–1
2	12	10	1–1
3	9	11	1–2
4	11	12	0–1
5	13	11	2–1
6	10	11	1–1
7	11	12	2–1
8	12	10	2–0
9	10	11	1–0
10	11	12	2–1
11	12	11	2–2
12	11	Missing	1–0

What this demonstrates

Field-level data checks, explicit denominators, reproducible calculations and restrained interpretation. It does not demonstrate tactical expertise, predictive model accuracy, measured time savings or willingness to pay.

Source: locally authored fictional_matches.csv. Generated by build_report.py using Python standard library only. No external data, APIs or paid services. Published demonstration using fictional data.