For kicks, I wrote a quick Monte Carlo simulation of the bridge game. I assumed each player had a 50% chance on each tile, and each player would perfectly remember the safe tiles previous players had uncovered. The results over a million trials were:
0 players remaining: 547 trials, 0.05%
1 players remaining: 3115 trials, 0.31%
2 players remaining: 11572 trials, 1.16%
3 players remaining: 32837 trials, 3.28%
4 players remaining: 70612 trials, 7.06%
5 players remaining: 121468 trials, 12.15%
6 players remaining: 166828 trials, 16.68%
7 players remaining: 185008 trials, 18.50%
8 players remaining: 167495 trials, 16.75%
9 players remaining: 121343 trials, 12.13%
10 players remaining: 70978 trials, 7.10%
11 players remaining: 32481 trials, 3.25%
12 players remaining: 11831 trials, 1.18%
13 players remaining: 3125 trials, 0.31%
14 players remaining: 563 trials, 0.06%
15 players remaining: 68 trials, 0.01%
16 players remaining: 3 trials, 0.00%
So they had a pretty good shot of having at least 2 players remaining, but the most likely result was 7 players remaining.
I know that isn’t quite what happened - at least one time someone forgot a previous tile, and there was the case of 2 players dying on the same tile. And the glass guy “cheated” on 2 tiles, so I’ll run the numbers again trying to account for those.