diminishing returns from good ideas

70% you'll have a good idea

20% you'll know what to do next

10% you'll know how to course correct

5% you'll know when to cut your losses

1% you'll know how to actually do the actual work!

= some vanishly small probability of "completion"