When you first start homebrewing, every batch is an exploration. You build a knowledge base of processes and ingredients by executing other people’s recipes. That experience later helps you start making your own modifications. It’s an organic kind of learning.
Eventually, you look back and realize how much you know, but there’s always more to learn. As a seasoned brewer, it might be time to augment your casual approach with something more structured. It could start with something as simple as trying out a new technique on an old, reliable recipe. If you want to get more serious, you can bring in some rigor and apply a version of the scientific method by focusing on a brewing element, making a testable hypothesis, trying it out, and then assessing the data (i.e. the beer).
What is the Question?
The first step is to pick your target: what are you trying to evaluate? You want to control the scope of your investigation. This is a matter of trimming away distractions to get to the root. Choose something specific, such as asking what first wort hopping does for a beer or wondering how fermentation temperature impacts the phenol/ester balance in a wheat beer. Constraining the question provides the structure for your experiment.
Setting Expectations
Once you have your question, you have to come up with a hypothesis. How do you expect your change to make a difference in your beer? This will help you construct your experiment and figure out how to assess the results. At this point, it can be useful to decide how strict you want to be. Seeing whether something makes a real difference to your own palate is fairly easy, but if you’re looking for an objective assessment, you’ll want to be more rigorous in defining your experimental process and in your recordkeeping. Both approaches can yield useful knowledge.
Experimental Protocol
To get value out of your experiment, you need to relate the differences that you introduced to the differences you can observe in your finished beer. You may use a previous beer as a baseline, or you can split your batch and compare the mini-batches against one another. Two things will make the comparison and the correlation much easier. Just like you had to drill in on your question, you need to structure your experiment in a way that limits the variables. You also need to make those differences easier to identify.
Focus the Change
If you introduce too many changes, it’s hard to guess which had an impact on the final beer. Reduce the possible root causes by isolating your changes to the key idea of your question. If you’re comparing hops varieties, leave the grain bill and mash process the same. If you’re looking at dry hopping, split the batch and use the non dry−hopped version as a baseline. When comparing different yeasts, attempt to hit the same pitching rate with each one. Reducing the variables is important because we don’t brew in a laboratory. On the homebrewing scale, you can’t escape some degree of variability. By keeping things as similar as possible, we have a better chance of validating our hypothesis.
Making Differences Stand Out
Even if you do control your variables, you also need to make sure that other elements of the beer won’t eclipse the changes you introduced. It doesn’t make much sense to compare 2-row malt vs. 6-row using a Russian Imperial Stout as your base beer. Beginning with the differences you expect to observe, think about how you could best showcase those distinctions. Often, a milder base beer is the best starting point.
Assessment
To recap the scientific method, you asked a question, you constructed a hypothesis, and you worked out an experiment to test that expectation. Now it’s time to collect your results to see if your idea worked. This is another area where your decision about rigor will make a difference.
If the testing is in the tasting, your options could range from a casual session with some friends to a more formal blind tasting (more on that to come next time). Remember, though, that not all experiments need to be centered on flavor. You could also evaluate metrics in the brewing process, such as lag time or final gravity.
In any case, your test should focus on whether there is a real difference between your baseline and your experiment, and also whether the results validate your expectations. If you’re lucky, you’ll get an obvious answer: yes, this worked, or no, it didn’t turn out. On the other hand, there’s a good chance that your results may be more ambivalent. While that can be frustrating, it actually supports another fundamental scientific truth that repeatability and feedback define progress. You may need to run your experiment again. If so, take the time to analyze your protocol and see what you could do better to reduce the effect of other variables and make the differences stand out.