Practical guide · Data literacy
Two teams can have different-looking statistics because they played different opponents, maps or numbers of matches. A percentage without its denominator can exaggerate the evidence. Start by asking what was counted and over which period.
Align the sample
Compare the same game, stage, map scope and time window wherever possible. Online and LAN results may have different contexts. A team with three recent maps should not be presented as equally well measured as one with thirty.
Read the denominator
A 100% record can mean one win from one attempt. Display the number of attempts beside the rate. Missing data is not zero, and a provider's absent value should not become a low performance score.
Separate description from prediction
A historical win rate describes the selected sample. It does not automatically estimate the next match's chance. Roster changes and opponent selection can weaken that interpretation further. Explain what the statistic measures without calling it a forecast.
Use a comparison note
Record source, retrieval time, date range, map count and exclusions. If two providers define a metric differently, keep their figures separate. A transparent small sample can still be useful for a question; it is the unsupported claim of precision that makes the comparison misleading.
Compare two fictional samples
Imagine Team A won four of five maps on a particular map, while Team B won twelve of twenty. The displayed rates are eighty and sixty percent, but the samples differ greatly in size and may involve different opponents. That observation alone does not establish that Team A is stronger. Write the counts beside the percentages and inspect the date range. If Team A played only weaker qualifying opponents while Team B faced a later playoff field, the comparison needs that context before it can support even a descriptive conclusion.
Decide the question first
A statistic should answer a defined question. Are you describing recent results, map selection frequency or performance with a current roster? Each needs a different sample. Choosing the most impressive number after looking through many metrics can produce a misleading story. State the question and inclusion rules before collecting the figures. If the available data cannot answer it, narrow the question. A table of transparent observations is more useful than a model-like score whose inputs and exclusions cannot be explained.
Watch for selection effects
Public match records may omit lower-tier events, qualifiers or certain regions. A provider may cover one competition more completely than another. Check the coverage notes before calling a dataset comprehensive. Missing matches can change rates and rankings, especially in small samples. Do not fill the gaps with guessed results. If the sample is limited to the matches available through one provider, say exactly that. This limitation does not make the data useless; it defines the boundary of the claim you can responsibly make.
Separate zero from unavailable
A zero value means a measured quantity was zero under the metric definition. An unavailable value means it was not supplied or could not be established. Replacing unavailable values with zero can unfairly lower a team average and create false precision. Keep them explicitly missing and state how the summary handles them. If you exclude incomplete matches, disclose that exclusion. A comparison should not reward the better-covered team merely because the other team has missing fields that were silently treated as poor performance.
Write a careful conclusion
A useful conclusion might say that one team had a higher observed rate in a specified sample, while noting the denominator and context. It should not jump from that statement to a guaranteed result. If you discuss a possible explanation, label it as analysis and cite the evidence behind it. Recheck source timestamps before publishing a live preview. A small sample can motivate a question for viewers to watch, but it cannot carry the certainty of a large, well-defined measurement.
Separate descriptive notes from forecasts
A statistics worksheet can describe what happened without supporting a prediction about the next match. Give every table a short purpose statement, such as comparing map appearances within one completed stage. That prevents a reader from treating the largest number as a universal ranking. Include the number of observations beside an average and mark missing data rather than turning it into zero. If two players have different roles or faced different opponents, say so before drawing a comparison. A chart can look precise even when its categories are poorly matched. Check whether the source counts overtime, substitutions or incomplete maps, and retain its definition when presenting your own summary. If you cannot establish a definition, limit the article to the raw observation and explain the uncertainty. Avoid adding a percentage improvement merely because two reported numbers differ; first establish that both measure the same thing over comparable periods. The final note should tell the reader what the data can answer, what it cannot answer and where to find the original record. This is useful analysis even when the honest conclusion is that the available sample is too small.
Related reading and sources
Continue with Tournament directory and Match centre. Reference: Official LoL Esports schedule, checked 7 September 2026. The workflow above is editorial guidance; examples are hypothetical and are not reported test results.