Data for Donuts: Putting the Moe's Doughs vs Peter Pan Debate to Rest

In Brooklyn there is a baker named Moe. Moe worked at Peter Pan Donut & Pastry Shop in Greenpoint making donuts for years and years. One day Moe decided to venture out on his own. Moe set up his own donut shop just around the corner from Peter Pan, using Peter Pan's recipes, and named it Moe's Doughs Doughnuts.

Drama has ensued. Debate has raged among donut connoisseurs ever since. 

The New York Times, in a profile piece on both bakeries, refused to declare which was better. The Village Voice didn't pass judgment either. This blogger and that blogger both ate a variety from each shop (like I did last year), but wrote only about the qualities of the donuts and did so only from their personal perspectives.  

I, like many others, appreciate a good donut. I appreciate the finer things in life. I don't do well with ambiguity. I must know with certainty what the finest things are.

Not knowing which bakery made the finest donuts was, erm, eating at me. Also, I wanted to know if I was better off going to Moe's when the queue snaked out the door at Peter Pan and all available seats were snatched by late-to-the-party-hipsters and tourists extending and Instagramming authentic experiential moments for way longer than the time necessary to have a donut. Moe's never has a line when I go there. But should I go there? What if the donuts at Peter Pan are better? I should wait for those then. But how do I get the truth without real data?

I set Moe's against Peter Pan, donut against donut, Old Fashioned against Old Fashioned, Glazed against Glazed in a blind taste test. 

It would be an "any given day" experiment I'd conduct. No call aheads. No hinting at what I was doing. Just a guy walking in buying a variety of donuts. 

I walked into each shop around 8:00 AM, bought two dozen donuts and a cinnamon bun. I couldn't get a 1:1 match of flavors due to I guess what we'll call market volatility. All these donuts came with me to the office.

Flavors were matched from row to row. I randomized the donuts in each row by shop. Sometimes plate A had a donut from Moe's, sometimes it had one from Peter Pan. Nobody would know which shop's donut they were eating. I carved each donut into quadrants, stuck toothpicks with Sharpie'd IDs in each quadrant, and plated them in parallel-flavored rows. The cinnamon buns were so large that they had to be cut into six parts. Altogether this let me make 46 comparisons. I placed all the donuts in a communal area at my office and invited my coworkers to participate. Word spread quickly.

It wasn't a totally randomized set of participant voters, but it's a large enough sample set that I think it returned true results. I printed out a few copies of this to leave near the donut array:

Here’s the situation:

Some say Moe does donuts better. Others say Peter Pan is the original and best. These are the opinions of individuals, biased and unscientific in their research, eating one donut at a time. But we are employees at a clinical trial software (as a solution) company! Together, we can do better. We can answer the questions and put the debate to rest.

Here's how:

1. Take one donut quadrant from a plate in row A (e.g.. 3A)
2. Take one donut quadrant from a matching plate in B with a number (e.g.. 3B)
   1. Whatever you try from row A, you must also try from row B.
   2. This way we’re comparing the same type donut from one shop against the other.  
3. Eat them both.
4. Savor this moment.
5. Place the toothpick with an ID flag from the better donut quadrant in the Winner! bin near the rows of donuts.
6. Toss the other flagged toothpick in the trash.

Flags in the Winner! bin shall then be tallied. The winning shop and individual winning donuts shall then be declared. Results will be posted and debates settled when all donuts are consumed.

It's not the hardest science, but it aspires. 

People had lots of commentary on the donuts. They talked about texture, amounts of sugar or salt, freshness, etc. All I wanted to know was of the two they just ate, which one was better?

N.B.: My employer did not sponsor this research, has no opinion on these results. Nor did they disapprove of this data collection. And I don't speak on their behalf. 

Overall, after the numbers came in, it was Moe's that won.

25 total votes for Moe's Doughs Donuts to 14 for Peter Pan Donuts & Bakery. One of my coworkers with a Ph.D charted the data:

Some people just grabbed donuts and walked off without voting. Dr. Coworker said there's a margin of error of 7. Even adjusted with that margin of error in favor of Peter Pan, Moe's is still the winner. 

I didn't see that coming, but now I know where I'm going next time I need a fix.