<result> <column_name>1st%</column_name> <family>uncertain</family> <subtype>none</subtype> <confidence>0.3</confidence> <requires_user_confirmation>true</requires_user_confirmation> <sum_meaningful>false</sum_meaningful> <signed>false</signed> <reducers> <primary_reducer>mean</primary_reducer> <valid_reducers>mean, min, max, median</valid_reducers> <excluded_reducers>sum (percent shares do not accumulate), product (values are not sequential growth factors)</excluded_reducers> <average_semantics>arithmetic_mean</average_semantics> </reducers> <inner_outer_measure> <inner_measure_reducer>n/a</inner_measure_reducer> <outer_measure_reducer>n/a</outer_measure_reducer> </inner_outer_measure> <multiplicative_partner_derivable>false</multiplicative_partner_derivable> <reasoning>The table is PATH with a Date sequence axis, so the gate allows a full analysis. The "%" in the name and the percent-formatted values are multiplicative naming signals. However, the column is VARCHAR, so no numeric statistics are available and there is no reconstruction evidence. The samples (0.0%, 10.0%, 10.5%, plus "-" and blank placeholders) look like a share or percentage level, for example a share of first-place results. They are not compounding growth factors centered on 1.0, and no center-of-mass or cumprod evidence supports a ratio. Because the naming signal and the apparent semantics conflict and there is no data to settle it, I abstain. The column needs parsing (strip "%", map "-" and blanks to null) and human confirmation of its meaning. If it is a share level, mean-style reducers are the safest.</reasoning> <table_type_conflict>false</table_type_conflict> </result>