Odds in TRPGs

I find there is a need for randomization in a TRPG; not just for the thrill of the players, but also as a crutch for the game master when their mental stack is too demanding for them to offer an interesting deal or dilemma, and when just letting it happen feels unsatisfying. However, there is definitely an open question of how to determine the odds of the randomizer (as well as the severity of the results, but that's saved for another blog post). 

 

I believe I've finalized a method to determine the odds that I find very satisfying to rely on. I usually approach things by using some easy to reference mnemonic, and so after some fiddling I arrived at using the empirical rule as my rule of thumb:

Since ~68% of the outcomes of an experiment should be within one standard deviation of the average result (if following a normal distribution, which I chose loosely on the idea that the average influence of all random variables that may affect the outcome would tend to approximate a normal distribution), I decided that I could say that something was "likely" if it followed those odds. Naturally, "unlikely" would be the rest from the law of total probability, i.e. ~ 32% (as you can see I am only using this standard deviation as a rule of thumb for what odds feel "right").

 

I came up with the following set of "difficulties" for rolls based on the previous idea that ~68% is a "likely" outcome to bet on:

 

• ~83% (5 in 6) chance is "very likely/easy"
• ~66% (4 in 6) chance is "likely"
•   50% (3 in 6) chance is "even/moderate"
• ~33% (2 in 6) chance is "unlikely"
• ~16% (1 in 6) chance is "very unlikely/hard"
  

 

I find using the previous reference has given me easy odds on the fly which result in satisfying stakes at the table.