Google vs Librarians: When It Comes to Recipes, Old Fashioned Research Trumps Technology

SF logo_2013

I spent Sunday at the UCLA Science and Pie extravaganza.  It was pretty much food geek heaven.  The UCLA Science and Food class had a “science pie” competition for their last public lecture.  Students were asked to choose a part of pie making for scientific inquiry and present the results of their research for the public and a group of judges that included two scientists Andrea Kasko and Sally Krasne, the day’s lecturers Christina Tosi and Zoe Nathan, me and Jonathan Gold.  It was a combination of old fashioned science project poster displays of inquiry and results along with pie tastings and best of all,  two perfectly complementary lectures by pastry chef Tosi and baker Nathan.

Here is the main takeaway I have from the day.  If you want to learn how to make a recipe that has very few ingredients don’t turn to a search engine to find the best recipe to do it, ask a librarian to help you instead. Remember that there is a difference between mass-market taste and flavor-textural excellence as defined by masters of the craft.  Pie dough recipes are the perfect example of why cookbooks and, in particular, highly reviewed and acclaimed cookbook authors still have such high value for those on a quest for culinary mastery that will result in deliciousness.  For that’s the goal, right?  Cookbooks played an essential role in teaching me how to cook.  I didn’t go to culinary school.  Instead I learned from real people whose food I loved and real people whose books I loved.  You can’t substitute a “ranking” for a master to transmit cooking knowledge.  It’s a physical task and like many physical tasks (think ballet or ceramics), the subtleties are often lost in a boilerplate recipe.

During the judging I asked the teams where they got their original recipe on which to base their inquiry and experimentation.  Over and over again I heard “google”.  They figured that the “best” recipes would be those with the highest rankings.  That is an interesting but, I think, flawed assumption that we explored in this Good Food interview with computer programmer Lada Adamic who created an algorithm to determine the success of online recipes.

One team even printed on their poster that the “perfect pie ratio is 5-2-1 (flour-fat-liquid)”, which stopped me in my tracks.  The real perfect pie ratio is 3-2-1.  I learned this from one of my mom’s friends, a talented home baker when I was 12.  Ask any professional pie baker and they’ll tell you it’s 3-2-1.  In Michael Ruhlman’s book/app Ratio it’s 3-2-1.  So where did this young woman get her information?  She figured it out from a cursory internet search of a few top ranked recipes.  But she was wrong.  And she was at UCLA, a repository of an extensive cookbook collection and a gaggle of smart, passionate librarians who would have loved, I’m sure to provide her with the beginning tools she needed in the form of a recipe developed by a “master”.