As jams on itch.io become more bigger and more popular, I wonder if more advanced rating systems are becoming necessary for scoring to be fair. With larger jams there is inevitably going to be a huge range of ratings for games and the raw averages are always going to be biased towards games with fewer ratings. I know that there is some penalty for being under the median review count, but that still skews games that are right above the median number higher, and those that become popular to have a hard time competing. 

Luckily there are tools out there that can help! Bayesian Scoring can help account for the large difference in ratings for these games and help prevent the trap of popular titles being punished, especially if the games are popular because they are good!

Simple example of the formula (in JS)

// Bayesian rating formula:
// score = (C*m + R*v) / (m + v)
function bayesianScore(R, v, C, m) {
  // R = item's average rating (e.g., 4.6 out of 5)
  // v = number of votes/ratings the item has
  // C = global average rating across *all* items (e.g., 3.7)
  // m = minimum votes threshold before trusting the item's own average
  if (v <= 0) return C; // no votes → return global mean
  return (C * m + R * v) / (m + v);
}