Talk:The Game Master

Drop rate stats are inconsistent

The provided stats for drops don't line up.

Provided details:

    Checkmate (Rare — 5.71%)
    Gamblers Gambit (Common — 45.71%)
    Schemer's Guise (Common — 51.43%)

Drop rate notes: Percentages are based on 35 kills. Both times Checkmate dropped, it dropped alongside one of the two other items.

I assume these % values come from 2/35 (.0571), 16/35 (.4571), and 18/35 (.5143) respectively. However, that means there were only 34 drops with either Gamblers Gambit or Schemer's Guise, so either there was 1 kill with no loot (which I don't think is possible) or 1 kill with only Checkmate (contrary to the original edit).

I have logs from 19 kills: 10 Gamblers Gambit, 9 Schemer's Guise, 0 Checkmate. Combining that with the original counts as given (2, 16, and 18 out of 35) makes 54 kills.

This gives an observed drop rate of:

Checkmate (3.7%) Gamblers Gambit (48%) Schemer's Guise (50%)

I'd like to see clarification on the data initially provided as it isn't consistent with the claims. But at best it gives confidence intervals, with only 54 observations recorded. Each of the ranges below is to 95% confidence:

Checkmate (1-12.5%) Gamblers Gambit (35-61%) Schemer's Guise (37-63%)

These ranges are from the Wilson score interval (lower, upper = (p̂ + z²/2n ± z·√(p̂(1-p̂)/n + z²/4n²)) / (1 + z²/n)), where p̂ is the observed drop rate, n is the number of kills, and z is 1.96 for 95% confidence.

By the drop frequency table on Help:Contents, 45.71% would make Gamblers Gambit Uncommon rather than Common. The pooled rates would label Checkmate Ultra Rare (under 5%), Gamblers Gambit Uncommon (25% to under 50%), and Schemer's Guise Common (50% or more, right at the cutoff), but the ranges above cross those thresholds, so 54 kills can't settle the labels either way. Since the original poster's Gambit rate was below 50% (16 of 35) and my 19 kills alone were above it (10 of 19), I'm going to list both Gamblers Gambit and Schemer's Guise as Common (about 50%) and then list the 95% confidence interval. Additionally, if the drop rate for the two items is 50%, most actual measurements will have one slightly below 50% and one slightly above 50% (since each kill drops exactly one of the two), endlessly flipping between Common and Uncommon.

More data is needed for tighter estimates, and it should come from a variety of sources, not just from runs people chose to post. Ideally the original poster will update their runs, I'll certainly update mine, though my subscription ends soon. Those interested in searching their logs can use this regex to find drops from The Game Master:

^\[(.+?)\] You (?:have )?looted\s+(.+?)\s+from The Game Master's corpse(?: to create (.+))?

Drop rates also need the kill count, which isn't the same as the number of loot lines, since one kill can produce more than one (Checkmate dropped alongside another item). Kills can be counted with the following regex, but you will need to be sure it's your group's kill, and not a kill by someone outside you or your group that your log picked up:

(?:You have slain The Game Master|The Game Master has been slain)

Justatest90 (talk) 14:56, 28 October 2026 (EDT)