4 ms·
The ROI for a given bit of science over a one year span will almost always be a big negative. The ROI for "Science" over the last century is a huge portion of t
by digikata 12y ago
The ROI for a given bit of science over a one year span will almost always be a big negative. The ROI for "Science" over the last century is a huge portion of the GDP of the world.
I'd assert that for optimizing some mostly known process, building an ROI model for investment decisions is valid, but for long-term exploration of the unknown it's meaningless because you can't build a predictable model for ROI.
- swatow 12y agoWhere does "one year" come into it? ROI means the total discounted future returns divided by total present investment. Talking about ROI is nothing to do with short vs long term thinking. ROI should always be taken to mean expected ROI. Now I'm not dogmatic when it comes to probability. If you say that the probability distribution of a given projects future returns is not well defined, that's fine. Nonetheless, we implicitly always use something roughly similar to probabilities when evaluating uncertain outcomes. I'm not even saying that doing this explicitly with numbers is even useful. I'm just saying that there is no fundamental difference between saying "we should(n't) do X" and calculating the ROI of X.