A sim­ple ques­tion, really.

We must all have heard of an aca­d­e­mic pub­li­ca­tion cita­tion impact mea­sures: pure counts, cen­tral­ity, H-​​index. What­ever. Peo­ple who write jour­nal arti­cles want to know how many times their papers are cited. It affects their aca­d­e­mic jobs, their rep­u­ta­tions, their pay. And to some extent, inven­tors accrue some rep­u­ta­tional rewards from the cita­tion impact of their patents.

But let’s step out­side The Tower for a minute, OK? Sup­pose we had some magic way of see­ing how often some­body uses a tech­nique or a datum or a model or a the­o­rem or an algo­rithm from an aca­d­e­mic paper. By “use” I am not refer­ring to how often some­body writes another paper.

And by “use” I mean use. Peo­ple out here, in the Actual World. Mak­ing actual stuff. And then not writ­ing papers about it. Because we’re not mak­ing the stuff just so we can write another paper.

I ask, because I find I am about to use some­thing Nic McPhee just pub­lished with Ric­cardo Poli. And I real­ize I won’t cite them, except maybe via a blog entry or two. And surely the friends and col­leagues and com­peti­tors I have who pre­dict mar­ket dynam­ics pro­fes­sion­ally using genetic pro­gram­ming, they use GP tech­niques and tricks they’ve read in papers. But the authors, they never hear about it. Or only infre­quently, I’m sure.

Any­way, say we had a mag­i­cal way to mea­sure those actual uses. A way to count imple­men­ta­tions where some aca­d­e­mic work had a real func­tional role.

Would the use­ful­ness be cor­re­lated in any way with the cita­tion rank?