One question that I have- on the surface it feels to me like there is a slight bias towards success in the calculations, since we have a positive outcome and a moderately positive outcome but only one negative outcome. Is there an explanation for this? Intuitively it seems that if we added more "tiers" we'd get more precision, though they'd probably be harder to define then the 3 you have outlined
Good point. You’re right that increasing tiers could potentially improve the accuracy but will also increase complexity. Your may consider it slightly biased towards the positive outcome, probably because it’s trying to propel action vs decision paralysis. At the end of the day, this is going to provide a EV that you can choose to use or ignore, and the most critical part is the actual likelihood and outcome values.
I love this framework!
One question that I have- on the surface it feels to me like there is a slight bias towards success in the calculations, since we have a positive outcome and a moderately positive outcome but only one negative outcome. Is there an explanation for this? Intuitively it seems that if we added more "tiers" we'd get more precision, though they'd probably be harder to define then the 3 you have outlined
Good point. You’re right that increasing tiers could potentially improve the accuracy but will also increase complexity. Your may consider it slightly biased towards the positive outcome, probably because it’s trying to propel action vs decision paralysis. At the end of the day, this is going to provide a EV that you can choose to use or ignore, and the most critical part is the actual likelihood and outcome values.