GEPSI

Methodology

Live Survey Data

GEPSI does not calculate global popularity by simply adding all participants together. Preference is first calculated within each country. Continental and global scores then use equal-country weighting to reduce distortions caused by differences in national sample sizes.

  1. 1Individual ResponsesEach valid submitted respondent selects titles (multi-select).
  2. 2Country Selection %Respondents in a country who selected the game ÷ valid respondents in that country.
  3. 3Minimum Sample ValidationCountries below the minimum number of valid respondents are excluded as Insufficient Data.
  4. 4Country NormalizationEach country becomes a percentage, removing sample-size differences.
  5. 5Equal Country WeightingEvery eligible country has weight 1, regardless of respondent count.
  6. 6Continental IndexMean of eligible country scores within the continent.
  7. 7Global IndexMean of all eligible country scores worldwide (not a mean of continents).

Worked example

South Korea: 600 of 1,000 respondents select Game A = 60%.

China: 50,000 of 100,000 respondents select Game A = 50%.

For continental aggregation, South Korea weight = 1 country and China weight = 1 country. The index averages across eligible countries, not respondents — a larger participant count never increases a country's weight.

A game receiving zero selections in an eligible country contributes 0% to the average — it is not treated as missing data. Fan Preference and Event Selection are calculated completely separately.

Demographic & Time Filters

When demographic or time filters are applied, preference is recalculated within each country's filtered respondents before equal-country weighting is applied. Eligibility is re-checked for every filter and every time period.

Age 18–24 example: Korea 60 / 100 = 60%, Japan 40 / 100 = 40%. Regional index = (60 + 40) / 2 = 50% — the average of eligible country scores, not 100 / 200 pooled respondents. Groups smaller than the minimum public cell size are shown as Insufficient Data. Methodology version: GEPSI-M1.

GEPSI-S1 Structural Suitability Model

Research-derived Delphi model with five criteria. Experts rate each criterion on a 1–5 Likert scale; the official score uses the research weights below (not equal weighting).

1Esports Ecosystem Activation1.1420.88%
2Copyright / Licensing Ease1.1420.88%
3Dedicated Observer Mode1.1420.88%
4Global Popularity1.0419.05%
5Online Competitive Environment Stability1.0018.32%
Σ5.46100%

SSI (1–5) = Σ(mean rating × weight) ÷ Σ weight, Σ weight = 5.46. Display score (0–100) = (SSI − 1) ÷ 4 × 100.

The Delphi study weights the five SSI criteria only. It does not establish a PESI/SSI weighting, so any PESI + SSI composite is shown only as a scenario / sensitivity analysis.

Host Market / Event Fit is not part of GEPSI-S1.

Data Limitations

  • GEPSI is a voluntary participant survey.
  • Results represent GEPSI respondents, not national population estimates.
  • Country normalization reduces distortion from unequal national respondent counts.
  • Equal-country weighting gives every eligible country the same weight.
  • Small filtered groups may be suppressed.