[dannyjed]

I do the same as the others have mentioned, but I have learned to only change one variable at a time.  When I first started, I would change two or more things at a time and would be left scratching my head and wondering which one made the difference.

[tomsawyer]

Denny I see the book selling used on Amazon.  And what will you brew for the 500th batch?  You might have to start work on a new "quint" style ahead of time.

Interesting perspectives.  I am guilty of using the kitchen sink approach in the name of creating complexity.  Makes it tough to diagnose problems or determine what adjustments to make.

I'm less a fan of single variable manipulation.  Too many variables to quantitatively manipulate each one individually.  Maybe when I'm really close on a recipe, but not in the early stages.

I do need to employ some brewing software, that is something I've resisted for the most part.  Sounds like there are some nice brewbuilding features in a few of these programs, I don't think I was aware of that.

[markaberrant]

I do the same as the others have mentioned, but I have learned to only change one variable at a time.  When I first started, I would change two or more things at a time and would be left scratching my head and wondering which one made the difference.

This is pretty much what I have been doing the last couple of years.  I have been trying to hone in on 9 beer recipes (3 session beers that I brew twice a year, and 6 bigger styles that are brewed annually), starting with very simple recipes, and then tweaking.

It depends on what you want out of brewing, and where you are at.  After 5+ years of brewing all sorts of styles and never brewing the same recipe twice, I know the styles I like to drink the most.  I am now attempting to "master" these 9 recipes if you will.  For me, it has given a lot more focus to my brewing process and recipe formulation, I am much more disciplined and consistent.

[aviking427]

I do need to employ some brewing software, that is something I've resisted for the most part.  Sounds like there are some nice brewbuilding features in a few of these programs, I don't think I was aware of that.

I know both Beersmith and Beertools Pro offer 21 day free trials. If i remember correctly, the downloads were the full program. You can create a few beers in the 21 days to see which you prefer. Nothing beats the knowledge you need to create a beer without it, but it is helpful.

[denny]

Denny I see the book selling used on Amazon.  And what will you brew for the 500th batch?  You might have to start work on a new "quint" style ahead of time.

Cool...grab it!  It's a really good book and for a change he doesn't preach about decoction!  As to 500, that's 4-5 years away.  I'll settle for just being alive then!

[tomsawyer]

Cool...grab it!  It's a really good book and for a change he doesn't preach about decoction!

Just did, although I'm feeling stupid because when I announced here that the book was on Amazon there was a copy for $8, and mysteriously in the interim someone else snatched that up and I paid $25.  Stupid me.  I also picked up Gordon Strong's book while I was at it.

Hey I thought decoction was the greatest thing since sliced bread, same flavors too.

[Hokerer]

Get an idea for a beer
Search the interwebs for as many difference recipes as I can find for that style
Try to distill the basics of the style i.e. what specialty malts in what proportions do I see fairly commonly
simplify everything (I tend towards very simple malt bills at least to start with)
decide how much I want to diverge from the basic pattern I have discerned.
create a starting point recipe.
Post it here for feedback.
tweak, brew, tweak (iterate endlessly)

+1, just have to re-emphasize the "post it here" bit.  Check this thread... http://www.homebrewersassociation.org/forum/index.php?topic=4001.0  I'd never done a Dunkelweizen, won some CaraWheat, posted here discussing a recipe, and, voila!  Gold medal time.

[tomsawyer]

Hey I made a dunkelweizen with carawheat and it didn't finish in the top three even.  Must've gotten some bad carawheat.

[jaybeerman]

I like the architect/engineer approach.  When a beer concept comes to me, I put on the architect shoes and roughly sketch the details of a beer that's never been built before.  Then I put the engineers shoes on and mull it over long enough to either devise a method or call the architect back and tell him it's impossible to build.  If the later happens I keep it in the back of my mind and sometimes the solution comes to me.

[anthony]

Granted, I take this hobby more serious than some, but this is my 2c....I approach recipes differently depending on what the expected output is. With specialty beers and specialty meads for example, I often take a proven base recipe and essentially produce the specialty product from it in ~1 gallon increments. On a standard batch this gives me a lot of opportunities to get feedback from friends and clubmembers, competition, etc. before locking down to a certain ratio of ingredients. I do this for coffee beers, vanilla beers, some spice beers, fruit beers, etc. If a beer/mead/cider is going to have some new idea or unproven ingredient or technique introduced to it, in my mind it is even more important that the foundation has been proven and that I am familiar with the expected flavors, etc. Take for example, if you wanted to brew a pecan beer ala some of the pecan beers that are popular in the south, you would want to make sure you were familiar with the base brown beer so that the pecans were the only new variable you were introducing.

To get that initial proven base recipe, I, like most others who have posted, start with an initial research step. I will look in BCS, I will taste classic examples, look at CYBI recipes for classic examples, look up old school recipes that various legends have shared online, etc. Special ingredients or techniques get their own research step as well. Once I come up with a recipe, I will brew it. The very first time I am trying to perfect a recipe I will keep the feedback loop very tight and very aggressively pursue it. When I initially rack that first iteration, I will immediately reuse the yeast to brew the next iteration, and sometimes another varied iteration. I remember spending one weekend brewing 3 batches of English Mild after I had just racked 5 gallons of it into a keg (this probably explains a little bit about my dumping comments in the other thread too). I always only tweak either one thing or two/three very minor things. After I find the iteration or two that I think is the best and most accurately accomplishes what I set out to do with the recipe, I will introduce outside feedback into the process. At that point, it becomes refinement and the pace slows down considerably.

I think the most important parts of recipe development are detailed notes (or very accelerated development schedules like my Mild example above, so that you can simultaneously evaluate different adjustments) and knowing your system well enough that you can not only consistently brew on it but replicate your results. After all, there is no point in finding the perfect recipe if you can never brew it again 

[tomsawyer]

...When I initially rack that first iteration, I will immediately reuse the yeast to brew the next iteration, and sometimes another varied iteration. I remember spending one weekend brewing 3 batches of English Mild after I had just racked 5 gallons of it into a keg (this probably explains a little bit about my dumping comments in the other thread too). I always only tweak either one thing or two/three very minor things.

I think simultaneous evaluation of iterations has some merit.  Its more like a competition evaluation that way, and you're comparing variants at similar stages in terms of aging, etc.  I'm not sure I could sit down and fill out a beer scoresheet on a batch, then make a change and in a month fill out another scoresheet, and really be able to tell if one were better than another unless there was a pronounced difference.  Whereas, side by side analysis lets you discern more subtle differences and make decisions at the "tweaking" level.

Plus as a practical matter theres nothing I like doing more than making two batches of beer in one day.

[tomsawyer]

Brewing Better Beer came in the mail today, and after reading the foreward, intro and first chapter, I had to skip to the chapter on recipe formulation for a sneak peak.  Excellent stuff!  A lot of it you've covered in your answers to this post, but just seeing it in print kind of legitimizes it for me (indepedent confirmation).  I do a lot of the stuff already, though some things at a less methodical/more subconscious level (like how much I think about brewing to style).  I think maybe my problem is one of balance, a concept that discussed at some length.  I tend to the "more must be better", where instead of using a mash to get attenuation, I'll mash low AND use a highly attenuative recipe.  Or I'll use a recipe with lots and lots of rich malts and adjuncts, and wonder why I'm not getting a smooth malty flavor.  About the only thing I am restrained on, is my use of crystal malts.  I did learn this from online sources and as a result of my general desire to brew to the drier side of any style, and those styles that are drier to begin with.

In any case, this chapter has some valuable advice and a nice plan of attack for recipe formulation if anyone is interested in that.

[malzig]

I do a lot of what's been described here, but there's one thing I do that no one has mentioned.  I like to pick up a couple of what I consider outstanding examples of the style, after I've made the beer, and see how mine compares, making changes in the recipe if I think there's something lacking or over-the-top in mine.

[Will's Swill]

I like to do this as well - buy good commercial examples of the beer I just brewed for side-by-side comparison.  I find that I am unnecessarily over-critical of my own brew.  I look for, and seemingly find, every flaw.  It'suseful to taste next to a commercial example and think, "oh, that's not a flaw, that's a feature!"