Look fellas, no hard feelings. No one needs to thump chests around here. People are set in their ways and will go to great lengths to justify them. I just don't think ad hoc rounding, and the mindset that goes along with it, is going to help improve the quality of your beer.
So, where do you draw the line? We have variables in the quality of the grain due to age, malting, grind, etc. We have potentially inaccurate scales. We have many factors. How precise do we need to be? Whole ounce? Half ounce? Quarter pound?
How about hops? 1/4 ounce?
This is a good question. I think we all can agree that some measurement is too fine . . . no one is measuring nanograms of hops or pipetting their water to be accurate to the milliliter. But on the other hand, what is too course?
For me, its a judgement call. One part in 50 is probably a decent generalisation, but its still variable. I tend to measure to the oz on grains, although Im fairly forgiving of errors in my 2-row. But grains with stronger flavors or darker, I will measure more accurately. Particularly when the total amount is less than a pound. If a recipe calls for 3oz, Im likely to measure that down to the tenth, at 12oz, I may only worry if Im over or under by 2 or 3 tenths. If I need 8 1/2 lbs, 8lbs 9oz is going to be fine. Hops, Im likely only to worry about tenths of an oz, even though I may be measuring sub-oz quantities, because I dont think we can control the product that well.
Now, Ill note these variations in my logbook, ("a bit extra" or "a tad short") so if they do appear to have made a difference in the end-product, Ill know what I did for good or bad. Frankly, I want some variation between batches. Learning what does and doesnt make a difference is part of what makes a good chef, and I expect it will be part of what makes a good brewer. When cooking, I havent made a recipe as written in 20 years, and I take the same approach when brewing.
The first thing I do with your recipe is to enter it into my software, truncating what I think is unnecessary precision. Then, given that my process is likely different from yours, Im going tweak it to get the numbers that matter -- OG, IBU, inline with what the recipe lists. Then Ill likely tweak it further based on my own tastes -- I like a more bitter beer, so maybe Ill up the hop content, or shift some bittering charges earlier. Maybe Ill darken it or substitute some redwheat for part of the plain. Then finally, Im going to tweak it based on what I have in my closet -- If you call for 1/2lb 60L and I have 80L, I may try using it, possibly dropping an oz or two. Then Im going to print that and make that.
I just got a call from the store I placed an order with. Apparently they were out of one of the yeasts I ordered. Is that going to make a difference? Mabe, I told them to substitute another, similar yeast, Im quite confident that the beer I make this weekend will come out fine, maybe with a slightly different character, but it will still be tasty and I am sure I and my friends will enjoy it.
Can the answers to these questions vary between batch size? Measuring to a whole ounce of hops and quarter pound on grains might be fine on 20 gallon batch? But what about a 2 gallon batch?
Absolutely it depends on batch size. Any measurement needs to be accurate in proportion to its total amount.
I think you're short changing yourself if you start using rules of thumb that can lead to variations between batches, especially if you share the recipes with others. I know I would taste a beer and say that's a pinch too hoppy or too chocolatey. What do you do? How do you finely tune a recipe that started out with grains rounded to a quarter pound and halves of ounces of hops?
And in turn, I suspect you are shortchanging yourself by being too slavish to a recipe. If you dont have variation between batches, how do you learn? When I open a commercial brewery, Ill worry about consistent output, until then I enjoy the fact that each batch turns out a bit different, and my notes of "a tad over" and "sub 1lb for 6row" are sufficient for me to learn from.