Here are some pearls of wisdom I've picked up over the years.

  1. "Start with a simple model you can solve" (Lones Smith). It's easier to solve a simple model. If you have a completed proof for a base case, it's easier to see which bells and whistles can be accommodated without sacrificing tractability.
  2. "Here's an example of a bad paper: 'So-and-so showed that property A holds if the world is shaped like an apple. I will show that property A still holds if the world is shaped like a pear'" (Peter Diamond). It's better to show that property A does NOT hold if the world is shaped like a pear (i.e., show the prior result is not robust). Alternatively show that property A holds for a planet of any shape (i.e., generalize the prior result).
  3. "I'd rather make the assumption and prove the result than not make the assumption and not prove the result" (Stephen Morris). Self-explanatory.
  4. "It's better to show that property A holds as long as x > 0 than to show that the property holds as long as x > -1.2" (Itzhak Gilboa by way of Ady Pauzner). Complicated assumptions can be distracting and controversial.
  5. "Less is more" (origin unknown). Points of vulnerability make it easier for referees to recommend rejection. Don't make their job easy! Purge your paper of dubious claims. If you don't think there are any, present your paper a few times. Audience members will help you identify them.
  6. "Don't be about the literature; be about the economy" (Larry Summers). Self-explanatory.
  7. "You can't make hits if you don't make movies" (origin unknown). Work on the best idea you have at the moment.

- David Frankel

Site Navigation