You might have noticed that the Internet has lost its mind a little bit over a new paper published in the Matter Journal.

Now you might have read it, but more likely you've seen one of those ridiculous headlines where journalists try to just out dramatic each other for a title.

Anyway, The actual paper is interesting, and the actual paper is definitely worth reading.

It's titled Systematically Improving Espresso Insights from Mathematical Modeling and Experiment.

Now the title does give away the key things to understand.

Here, First thing is modeling.

So a group of scientists and mathematicians got together and created a mathematical model of how espresso extraction should work and should is is really the key thing here.

Now the maths involved is comfortably above my pay grade or comprehension, but I don't think you need to necessarily understand the math.

You just need to trust that the people doing it understand the mat involved.

They're using this model.

They could predict how something like extraction would change with grown setting or how extraction might change if you change the constant pressure for the puck on.

That's interesting.

So they had their model of have espresso should work now.

The model shows things that are pretty predictable.

If you grind your coffee finer, you increase the surface area, and also you make it more difficult for the water to flow through.

Therefore, for a constant pressure, you should see an increase in extraction the finer that you go.

So far, so obvious, right?

Makes total sense.

But then they tested the model in the real world, and this is where things got interesting instead of seeing a constant increase in extraction as they went finer and finer and finer.

Instead, they saw extraction begin to decrease as you went finer and finer and finer.

So it very fine grown settings.

They had much lower extractions than they did coarser grown settings, and this was, well, a little unexpected.

Now, at this point, I need a quick kind of caveat or side note on on extraction in general, What they were able to do on what many people could now do is take a cup of coffee and see how much of the soluble material from the grounds ended up in the cup of coffee, right?

And from that you can calculate a total extraction on broadly speaking, extraction is linked to quality, but extraction is kind of the end result.

It doesn't tell you how you got there.

You would have a scenario.

Let's say where you have very high extraction of a very small portion of the grounds and a very low extraction of a lot of the grounds.

Well, that might have a similar total strength in the cup of total extraction, as as a much more even extraction.

That's a little bit lower, right?

So it tells you what you got.

It tells you how much came out, but it doesn't tell you whether or not it was even the extracted or uneven extracted.

Therefore, it's not a perfect indicator off quality.

So back to the paper that drop in extraction seems counter intuitive.

But they did have an explanation at very fine grinds.

You're inevitably going to get channelling.

Water cannot pass evenly through that cake of coffee, and so some water, quite a lot of water, is gonna flow through just a little bit of coffee where channel form you'll have a very high floor it through a channel on the coffee around that channel won't see as much water.

It won't be extracted as much, and this is causing this drop in total extraction.

You've now got two competing factors in espresso brewing.

Right, You've got grind size, which determines on the one hand total surface area and the final you grind, the more coffee you should be able to extract, but the course a you grind, the more evenly it's likely that the water will flow through the cake.

And so what you're looking for is kind of a cross ever point a point.

That's the best of both.

Where you going?

As fine as possible for an even flow rate.

So far, not so shocking.

This just sounds like dialing an espresso.

But it turns out that during the experiment, this crossover point was producing his breast of brew times of between seven and 15 seconds, and this is considered an absolute no no.

In the world of espresso brewing, no one would tell you can get good espresso in 15 seconds.

So I read the paper.

Hopefully, like many of you, and I thought, okay, this sounds interesting.

Little weird.

I should replicate that, and so I could replicate a reasonable amount of what was in the paper.

They were using a Malka Nik E K 43 grinder with coffee bars.

I'd run those lying around it square mile.

They I think we're using an opera espresso machine, but they won't do anything particularly distinct to that machine's technology.

They were just brewing at six bars, flat profile, and I had an espresso machine that I could put 26 bars and that was fine on the six bars is interesting and notable.

Traditionally, we brought about nine bars of pressure at the pump on.

There's a couple of explanations for that.

The first is the kind of myth version of it, which is that nine bars was considered to be the kind of average pressure of an old school leaver shot.

Not sure how true that is, but it's a nice story.

The second version is that you tend to see a peak of flow at around nine bars below nine powers.

If you're groaning very fine for espresso, the flow rate is a little bit lower because just less pressure pushing the water through above nine bars.

Your floor.

It slows again because you compress the cake so much that it's very hard for water to get through.

And nine bars around about is a kind of peak of flow in their testing.

They had a lot of issues with nine bars, so they brewed it.

Six bars and actually, coincidentally, six bars is what John from decent espresso or recommends people to use if they don't have a high end espresso Grander.

If you've got a kind of beginner medium espresso grounder, then then lower pressures will help, not destroy the cake and create lots of channels.

So I put my machine to six bars for easily done on.

We kept the bring ratio initially from the paper, which was 20 grams of coffee in and 40 grams of coffee out.

Now erupt Gareth from square mile into this to be my co conspirator and co tester in this process, and we made ah whole bunch of espresso, and we did what the paper suggests to do right.

You start a relatively fine grind and you put 20 in 40 out on you time that shot and then you go a little bit coarser, and you do the same thing now, Ideally, you're gonna measure the extraction alongside tasting as you go through this process.

And what you would expect to see, according to the paper, is that as you go coarser, your extraction increases a little bit and then begins to decrease at that peak of extraction is your kind of finest possible grind setting for Evenflo.

Now that grand setting that you find for peak flow for 20 and 40 out?

That may not taste good.

Okay, don't worry about that.

What you're supposed to do, then, is a gesture brew recipe.

Either change your dose or change your output weight change.

Your ratio is basically to get the extraction that you want to reach what I believe they call the tasty point.

Now I confess our sample size could have been bigger.

We could have pulled definitely way more shots of espresso, but it seemed a bit redundant because we could clearly replicate the results of their testing.

We saw a peak of extraction and a kind of course of grand setting than expected.

It was like a 15 2nd shot.

At that point, it didn't taste great But then we played around with the recipes, but undeniably we saw this this peak of flow, peak of extraction kind of thing happen in the paper.

I think that peak of flow is about 1.7 as a setting.

On the Ikea, we were 1.81 point 85 Somewhere there, I do think that really matters.

I can't speak to the perfect alignment of barbers or coffee, all of those kind of questions.

But there was definitely a grand setting that produced higher extractions despite having a fast flow.

There is more to talk about here, but before I do that, I do need to talk about the sponsor of this series, which is square space.

There are lots of different reasons to build a website.

Maybe you want to share your passion project.

Maybe you wanna share your hobby.

Maybe you need to set up an online store to sell the things that you've been building and creating.

I would strongly recommend squarespace.

It's what I use, and I'll tell you why.

It is incredibly easy to use.

It's quick, and I can take one of their templates and very quickly adjust it, customize it and make it look exactly how I want it.

Toe.

Look, I'm sure like me, you want your digital presence to be as beautiful as possible.

But maybe you don't have a massive budget to hire Web designers from.

Frankly, you don't need to.

You can sign up for a free trial of the link below on Dhe for 14 days, spent a little time building something and see how easy it is to create something stunning, something that's truly representative of you.

And when it's time to launch use code, James Hoffman and he can get 10% off any domain or website to start with.

Thank you so much to square space for sponsoring this video.

So we've replicated the results.

This seems to be true.

Is this a big deal?

Is this flawless work to what Have any criticisms?

Well, maybe so.

Firstly, this is for them one coffee and I've done it with just one coffee.

Now they would argue that they've done other work on grand settings.

I think the freezing paper in 2016 was a good example, and they saw that most arabica ground very similarly and so they would argue that this protocol should work for any Arabica coffee, but that's definitely something worth testing and having more people experiment with and test.

And secondly, they talk about flow through the cake as if it's either homogeneous and even all uneven.

And if you use a decent espresso machine for any amount of time, you'll see that there's often a point in an espresso at transitions from being Maur, even to less even right.

It might be that the start of the shot is pretty even.

But once a channel opens, then it stays pretty opener and your shot deteriorates pretty quickly.

In the decent, you'll often see on the screen for a fixed pressure, a reasonably fixed flow rate and then a sudden increase in flow.

That seems to be the point at which the kind of resistance of the cake has given up.

Channels have opened.

It's falling apart.

It's not providing the same level of resistance on the longer you brew the worst.

The shot will taste with the decent.

You will definitely see a delay in that uptick of flow if your grinder is capable of more uniform grand settings.

Typically, flat birds do this better than conical birds.

But it's not an absolute rule.

If your technique is better in terms of prepping the bed, you will have a more even flow for longer seemingly now.

I did put this point back to one of the lead authors on the paper, Dr Chris Hendon, on Dhe.

He acknowledged that there may be a change in even less, but he pushed back and said, Just because you're seeing relatively stable flow doesn't necessarily guarantee that that's coming from a truly even extraction, truly homogeneous flow through the cake on dhe, that's a fair point to now.

Interestingly, the paper does touch on the real world application of this in cafes.

In fact, they did.

It's, um, testing in a cafe in Portland.

That's that's kind of cool.

So the recommendations of the benefits would be like this.

If you're a cafe during 20 grams in 40 grams out in 28 to 30 seconds, that's not wildly unusual.

That's pretty common.

Andi is an opportunity here.

In theory, if you followed their protocols, you would have won much faster shots.

You'd be brewing down at 50 and seconds, literally half the time, which is just wild.

But maybe interesting, and you would probably be having a lower dose of ground coffee, Let's say 15 or 16 grams, but a much higher extraction in that 40 grams of liquid out.

You may end up actually with the same sort of amount of soluble material in a coffee cup.

Then you would have done from the 20 to 40.

If you extract 20% let's say off 20 grams of coffee.

Then you've got four grams of soluble material in an espresso to dilute into Millikan to a cappuccino or latte aeroflot.

Wider.

Whatever.

We got about four grams off that puck is dissolved in the coffee liquid.

If you took your dose down to, say 16 but you could get your extraction up to, say 25%.

It's a good bit higher, very high, but in theory you'd have four grams of soluble material in that cup, and the taste may well be better, or at least similar certainly in milk drinks.

So you've cut your kind of usage dramatically you would end up with, and this is where a lot of people sort of get a bit upset or less interested.

You don't it?

With weaker espresso bread at 15 and 40 out.

Even with a high yield, that's going to be a mom or dilute espresso.

Once you delete that with milk, you couldn't tell that's gonna be irrelevant.

But if you're a straight espresso drinker, you're gonna have shots 89% strength.

And many people like thick, gooey your espresso at the 10 12 or even higher as a percentage of strength.

Now they do acknowledge that the fact that these very fast flowing shots that are high extraction maybe lack a little bit of complexity compared to a more traditional espresso.

They suggest actually blending different stars and ratios, which they admit is not hugely practical in a cafe but might be in more industrial sort of espresso extraction.

It's an interesting idea.

I in my tasting the best shots that we had I thought were perfectly complex.

I thought they tasted quite cleanly and clearly off the coffees in the blend.

I didn't feel like they were lacking this espresso.

They just didn't have incredible texture, but the texture was by no means bad.

I think you know, we were getting shots of some nine and 1/2 percent strength on those were very enjoyable.

That does bring us back to one more thing, which is that they highlight repeatedly in the paper a paper where they are measuring things.

Pay a lot of attention to things, having a very scientific approach.