Change one variable at a time - good advice? - Page 9

#81: Post by **DavidMLewis** » July 29th, 2009, 7:34 pm

malachi wrote:but... in response to the true question...

1 - if you're looking to truly understand espresso, then yes - I would recommend changing one variable at a time. It's a useful learning exercise.

Hi Chris,

I'm not entirely sure I agree with you on this. The assumption behind the logic of changing one thing at a time is that the variables are independent, which as you state, they aren't in the case of espresso. So to really understand what one variable does in isolation, you have to change that variable and then re-optimize the related ones at the new setting, I think.

Best,
David

July 29th, 2009, 8:02 pm

DavidMLewis wrote:The assumption behind the logic of changing one thing at a time is that the variables are independent, which as you state, they aren't in the case of espresso. So to really understand what one variable does in isolation, you have to change that variable and then re-optimize the related ones at the new setting, I think.

The assumption is that if you change one variable, it will change the pull in a predictable way. It will either improve a result, or not. The corrolary is that if you change it the other way, it will either make it worse, or not. It answer's the question, "What happens if I do this?"
If you start out having bad luck, and you change two or three things, you won't ever be able to figure out which change had the most positive effect in the result, and which one will do what.
Randomly changing variables without any clue as to what does what is shooting blind. Ya can't steer the careening car away from the cliff's edge if you don't know which lever is the velocitator and which one is the deceleratrix, or what that big-roundy-thingie in front of you is supposed to do.
One at a time to discover what does what, and then start pairing them so see how one affects the other, and then, only after you have a fairly keen grasp of what they are, and how they interact, can you change multiple variables with any hope of improving the result any better than blind chance.

The assumption that one should change all the variables that are off presumes that the operator already knows which ones are off, by how much, and in what direction. Not to mention what level of change will effect what other variables and by how much.
That's purty danged presumptuous...

#83: Post by **zin1953** » July 29th, 2009, 9:14 pm

DavidMLewis wrote: . . . and then re-optimize the related ones at the new setting, I think.

Doesn't this presume you know how to "re-optimize"? And doesn't that imply that you already understand how the variables relate to one another?

#84: Post by **Illyfex** » February 10th, 2010, 3:00 pm

Some people approach a problem analytically and some use intuition. What works for you may not work for me. I don't weigh my dose, press into a scale or time my shots. I put just-so much coffee in the PF, tamp just-so hard, and pull the shot 'til it's done. It's a Zen thing for me.

#85: Post by **CRCasey** » February 10th, 2010, 3:37 pm

Zen and Espresso are a good way to go if you like wandering in the woods.

Otherwise a bit of method will help. This is not a fix nothing to correct everything sport.

-C

#86: Post by **Sakae** » May 2nd, 2011, 12:28 pm

Psyd wrote:It may be true that the variables are interdependent, but there are a slew of them. Changing one at a time gives you and idea, to a certain degree, of what that particular variable does. Imagine you inherited a stereo that was very complicated high-end schmata. Except, all of the screen printing had washed off the face, so all of the knobs weren't labeled. I'd start out by setting them all in the middle, (think of this as the 'golden rules' of 14g, thirty seconds, thirty pounds, 60 ml) and then start moving one knob at a time to see what it did to the sound. Yeah, I'm not ringing in my system moving one knob at a time, but I'm figuring what effect each has. Eventually, I'll tweak each into where I want it for a particular record or disc, or whatever, but moving them all a bit each way and seeing what I like is going to take a LO-O-OT longer, and I'm not going to be sure how I ended up there, and as soon as I change discs, I will probably just end up trying a bunch of knob turns until I'm happy, or tired. Knowing what knobs do what is a far handier tool to have in your tool box, and eventually you start to feel how they are inter-related. Eventually. If you stick with it and have an experimental nature.
I am a professional sound engineer. I have racks and stacks of kit with knobs on it. Heck, the smallest console I have has more'n a hundred knobs in it alone! While setting up a band, from Classical to thrash-metal, I find that I'm rarely turning more'n one knob at a time. And they are VERY inter-dependent. I do turn one and see that what I've done is what I wanted, and in the right amount, before I go to the next. Sometime I do have to come back and add a bit or subtract a bit of the previous change that I made, but I have to wait to see how the succeeding change effected the following one.
You can graduate to changing a buncha variables at once, but I'd suggest that you wait until you learn what each knobs do before you go messing about.

...split from Basket Overdosing; time for a serious re-evaluation! by moderator...

One variable at the time is a sage advice, however another_jim has a point to make, which is why I believe that I, as a novice, should start deliberating over an advice posted somewhere on the forum, that I need to examine quality of my shots, before I change grinds, weights, baskets, coffee beans, etc.

I prefer to create a prioritized table of various variants I wish to try, benchmark where I am in the process, and then test my changes progressively in different settings. If nothing works, I return process to the benchmark, focus on next variable, and go through the trials again. It is tedious (but loving) work, which sometimes might not succeed in cases in which several variables need to change at the same time to achieve desirable outcome (which is I think another_Jim is probably refering to). In some cases friendly barista who is willing to share his secret is probably invaluable source how to get one out of messy situation.

#87: Post by **asicign** » May 2nd, 2011, 3:17 pm

This discussion reminds me of the time I worked on linear programming optimization of an oil refinery. There were hundreds of independent variables,such as the type of crude to purchase, which products to create, and how to set up the various process units to optimize the yield of the product slate selected. What was interesting to me in that exercise was that in such a complex problem, there are multiple maxima (in this case profit), and you might find a solution that looks optimal, but by forcing a few constraints, a new better maximum might appear. In the refining business, many of the relationships are well understood, so it's possible to model the effect of a given action, so you could devise perhaps a few hundred equations to work against perhaps double that number of variables.

I'm not particularly familiar with the science of espresso preparation, but the number of variables, from bean selection and how it is roasted, grinding, and pulling a shot is daunting. Especially problematic is using quality of a shot as the criterion for optimizing a single variable. You might be honing in on, say, brew temperature, but perhaps everything else isn't anywhere close to being optimal, so the effort isn't productive. Unfortunately, with real gear, you can't vary everything at once like you can do with software.

#88: Post by **another_jim** » May 2nd, 2011, 3:46 pm

My point is simple. If you take a short step versus a long step, you aren't just changing how far you move your feet, but the dozens of other muscle groups that balance you; if you did not, you would fall on your face. High school physics will tell you the mano variables of espresso prep require the same sort of co-adjustments.

If you change the dose, you need to adjust the grind, in order to maintain a comparable flow rate (i.e. the brew ratio rate -- change of shot weight per gram dose weight per second). You should know how much of an adjustment turn on your grinder requires per gram of dose change to maintain the identical brew ratio rate. You should make that adjustment automatically whenever changing dose.
The average temperature of the brewing process throughout the puck will drop if the puck is deep and the water temperature remains the same. This is a more controversial adjustment, since the extraction zone while brewing is a thin slice of the puck travelling down from the puck's top to its bottom over the course of the shot. Early in the brew cycle, the lower parts of the puck actually absorb coffee solubles rather than giving them up. This slice by slice brewing compensates for depth in puck changes somwhat; but I still find I need to go up about 2 degrees with a high dose versus a low one to maintain the same balance of sours and bitters. This is again a co-adjustment I make automatically

#89: Post by **michaelbenis** » May 3rd, 2011, 4:09 am

Curiously I read this just after doling out the classic "change one variable at a time" advice to a relative newbie.

I entirely agree with Jim that once someone has a got an understanding/feel for the different variables, they will appreciate that there is an interplay at work, but I also agree with Psyd that to get there and adjust the web effectively (rather than getting tangled in it) they need to get some sort of understanding in the cup of how each parameter works when changed individually in order to appreciate how they also interrelate.

The walking analogy is a good one. Anyone with a little walking under their belt will recognise its accuracy; but someone starting out on the path (sorry: couldn't resist the pun

) will need to simply concentrate on moving one leg in front of the other without falling over before they get into the subtleties of what happens when you vary your pace....

In short, I still think it's essential advice to stop a newbie from simply getting terminally confused, while reminding them that further on down the road....

Cheers

Mike

#90: Post by **another_jim** » May 3rd, 2011, 12:17 pm

Yeah, I agree. It's important to distinguish between learning and doing. But this is being erased by modern design Nowadays, people expect their gear to have enough brains on board that experience is no longer necessary. I find this terrific for all the chore-like stuff I have to do which aren't hobbies; but for the stuff I enjoy doing, I'm much happier if there is a payoff to increasing skills.

One variable at a time may work for learners, but as an operating guideline for experienced espresso pullers it is both wrong and stultifying. We do learners a disservice if we tell them that this approach is anything more than training wheels.