Beechum Drew - Experimental Homebrewing: Mad Science in the Pursuit of Great Beer

When we first sat down and debated what this book was going to look like, we spent a lot of time talking about our brew day processes. In other words, how should we look at Dennys techniques for brewing a world-famous bourbon vanilla porter or Drews constant twists and turns on a familiar style, like saison?

As we sat and pondered our successful beers, we realized the weird and wacky side of things only worked due to some reasoned experimentation. We can talk endlessly about the strange ideas we have, but the success with a recipe always comes back to the trial and error. It comes back to the experiment.

Merriam-Webster defines an experiment as an operation or procedure carried out under controlled conditions in order to discover an unknown effect or law, to test or establish a hypothesis, or to illustrate a known law. For scientists, an experiment is an attempt to answer a question. This is where the vaunted scientific method, oft-misunderstood cornerstone of the modern world, comes into play. The method consists of a series of steps that the experimenter must follow for it to be an official experiment:

Question: What is the question youre seeking to answer?

Hypothesis: What do you think the answer is? You must be able to prove your hypothesis true or false.

Prediction: How will you know that your hypothesis is right or wrong?

Test: This is your experiment.

Review and analysis: Look at your results and see whether your hypothesis is true or false based on your data.

Now lets apply the scientific method to brewing. Perhaps we want to find out if late-hop-only (hop-bursted) beers exhibit the same bitterness as traditionally hopped beers:

Question: Do late-hop-only beers exhibit the same bitterness as traditionally hopped beers?

Hypothesis: Beers with only late additions will have a noticeably less bitter flavor than beers bittered to the same calculated international bittering units (IBUs) with a start-of-boil bittering charge.

Prediction: The less bitter beer will be discernible in a taste test focused on evaluating bitterness.

Test: Brew two beers from one mash. One boil kettle will get a standard 60-minute addition with enough alpha acid units (AAUs) to create 30 IBUs. The second boil kettle will get a late-hop-only charge at 5 minutes with enough AAUs to create 30 IBUs.

Review and analysis: Does a blind triangle tasting confirm the hypothesis? How many judges got it correct? What were their comments?

This may seem pretty straightforward, but the details can sink any experiment. For instance, if we change the word flavor in the hypothesis to level, then it calls into question the prediction. What does level mean? A quantified IBU? Great! But that would require fancy equipment to measure. On the other hand, flavor allows us to use a taste test. And in the end, isnt that what you care about?

In the professional world, scientists spend countless hours designing and redesigning their experiments to make sure that what theyre testing is what they think theyre testing. It takes weeks of careful strategizing for scientists to make sure theyve got the right plan in place. On the other hand, the experimentation that well be doing is going to be simple kitchen sciencewithout the rigorous controls and designs that are required if youre releasing, say, a cancer treatment. Its somewhat scientific, but its not Science with a capital S. Its about the beer.