Palmer's Algorithm for Good Gelato Recognition
On the methodology of iced sweet treats.
While visiting Italy for a friend’s wedding, I have been wielding daily a trick from the internet: Ada Palmer’s algorithm for spotting good gelato (abbreviated to Palmer’s Algorithm for the rest of this post). So as a fun lick-sized post, let’s analyze methodologically this algorithm.
Palmer is a historian specialized in the history of ideas and the Renaissance. As such, she loves Florence1, and, sensibly, she also loves gelato. This love has transmuted into an uncanny ability:
Friends traveling with me are often perplexed to see me stick my head in a gelateria’s door and instantly proclaim it good or bad, despite not having approached close enough to smell, let alone taste, the contents of the brilliant, alluring bins of swirling color. It can be done. There are visible signs of good and bad gelato, so today I am sharing my gelato-assessment method, applicable in Italy and around the world, and hopefully of service to you (especially to several specific friends who are going to Italy soon).
(Ada Palmer, How to Spot Good Gelato from 15 Feet Away, 2013)
Goal: Grounding “Good Gelato”
Before we get into the weeds of the algorithm she offers, we need to get a grip on her goal.
Palmer wants to find “good” gelato. But what does that mean exactly? Good is a value judgment, and so it’s particularly important to make it explicit, lest we confuse what good mean to us and what it means to Palmer.
At its core, good gelato Palmer fits the intuition behind “artisanal: use high quality ingredients, barely process them, and avoid artificial chemicals at all cost. This ideal of gelato particularly resonates with the culinary culture of Italy.2
This mean concretely a preference for simple flavors, done particularly well. And it also entail favoring fruit gelato made from water instead of cream (because the cream alters the flavor of the fruits).
Gelato is fundamentally just sugar and milk, or sugar and fruit in the case of a sorbetto, and the flavors are usually simple (hazelnut) or extremely simple (fior di late, pure milk). Thus, you can taste it very easily if gelato contains poor milk, poor fruit, or artificial chemicals (many respectable places still use chemicals to help the gelato coagulate and remain the correct degree of softness in the freezer), far more easily than you can taste the same chemicals supplementing the more full-bodied base flavors of ice cream. Thus high quality gelato is (in my opinion) better than the best ice cream because it showcases its excellent ingredients better than the ice cream does, but bad/cheap gelato tends to be worse than bad/cheap ice cream because ice cream’s natural fattyness and saltiness effectively conceals poor ingredients and artificial additives.
(Ada Palmer, How to Spot Good Gelato from 15 Feet Away, 2013)
Personally, this is what I’m looking for in gelato. I’m not that much into big mixes of flavors, or super-sweet or artificial combos, and I never liked cream-based fruit ice-cream. This means that I can apply Palmer’s algorithm directly — it targets what I aim for. Wherease if you wanted indulgent combinations of flavors or complex transformations of ingredients, you would need to pick and choose which parts of Palmer’s algorithm work for you.3
In addition to this notion of good gelato, another constraint of Palmer’s algorithm is sensorial: only eyes, no taste. You want to find if a gelateria is worth your money before ordering, thus you cannot rely on taste.
So to summarize, our goal is to detect if a gelateria is “artisanal” (high-quality ingredients with minimal processing) from sight alone.
The Algorithm, Reordered
Now we’re ready to look at the algorithm itself.
In the original post, Palmer presents what sounds like relatively independent steps (although the first one — colors of fruit gelato — is explicitly mentioned as the first thing she checks), which together combine to give a picture of the gelato quality. But I have found that a natural order between the steps emerges from repeated use, so I will focus on my slight adaptation.
Stage 1: Display
The first point when I’m gauging a gelateria is either from google maps, or outside the shop itself. I thus have low resolution and cannot make out the details necessary for the future steps.
But I can still apply a key step from Palmer’s algorithm: is the gelato is piled up in huge mounds. If it does, this is a highly negative point, as the only way for gelato to do that is to be pumped full of artificial chemicals:
Gelato is soft and fluid, and over time it will naturally flow down, like pudding. The only way to get it to stably stay in a big tall mound is either to freeze it solid (no longer yummy), or to add chemicals that help it remain solid (which can usually be tasted since there is no salt and little fat to conceal them). Thus big, tall, enticing mounds of gelato can be a warning sign.
(Ada Palmer, How to Spot Good Gelato from 15 Feet Away, 2013)
This criterion is powerful methodologically for a few reasons:
It’s clearcut: if huge mounds, no artisanal gelato.
This is much more straightforward than some of the follow-up steps, which are not guaranteeing one way or the other.
It’s visually obvious: the piles can generally be seen from outside or pictures online
No need for a fine eye for color differences
It’s directly tied to the underlying property: no artisanal gelato will have these mounds, because they physically cannot maintain such shapes
So the only cause is the artificial chemicals, and thus it’s clear that it invalidates the “artisanal” hypothesis.
So, huge mounds are an early-stopping criterion: no need to go further, you shall not find artisanal gelato in there.
The absence of mounds is a good initial sign, but it’s not a guarantee though
The best gelato will usually not stick above the rim of the bin, unless it has just been brought out. Many very good gelato places don’t even have an open bin, but keep the gelato in round metal containers with lids deep inside the counter. This means you can’t see the color of the gelato, but is generally a good sign, since anywhere that doesn’t show off the visuals of its gelato is usually good enough that it knows it doesn’t have to, and cares more about protecting the gelato than about showing it off. You do need to watch out, though, since some places that serve cheap gelato delivered by vans from warehouses receive it in flat bins with plastic wrap over the top, which is then unwrapped and served. So while tall mounts of gelato are a bad sign, flat bins aren’t a guarantee of quality. Metal lids pretty much always mean good quality.
(Ada Palmer, How to Spot Good Gelato from 15 Feet Away, 2013)
Stage 2: Choices
Let’s assume that there are no mounds, so you actually go inside. What’s the next phase of Palmer’s algorithm?
Personally, I find looking for which flavors exist much easier than looking at the quality (color and translucency) of the gelato itself — probably because of my untrained eye. So instead of prioritizing the colors (especially of fruit gelato), which is the official first step in Palmer’s post, I start with the steps about options:
Do they offer nocciola (hazelnut)?
Do they offer seasonal fruit flavors?
Do they offer fior di latte (flower of milk) and/or fior di panna (flower of cream)?
The first two (nocciola and seasonal fruits), come from the same proxy: they are significantly more expensive and effort-intensive to make than other flavors. This means that places which invest in these are much more likely to be artisanal and really care about the quality of their gelato.
This flavor [nocciola] is, gram for gram, usually the most expensive to produce, and to make genuinely powerful. For that reason, many gelato places save funds by offering chocolate-hazelnut flavors, bacio or Nutella, but not pure hazelnut. Others compensate with artificial or weak hazelnut.
(Ada Palmer, How to Spot Good Gelato from 15 Feet Away, 2013)
All gelato places will produce lemon, strawberry, and other popular flavors year round, but a gelato place which pays careful attention to the seasons, producing watermelon, apricot, and peach in summer, fig, apple, and pear in autumn, citrus in winter, and diverse berries in spring is another sign that the people in charge care about quality, and are therefore willing to put in extra effort to master a fleeting seasonal fruit which will only be profitable for about a month a year. This too bodes well for the quality of all the flavors. Similarly if you see a bright orange apricot flavor offered in December, safe money says that is a 100% artificial flavor, and many of the others probably are as well.
(Ada Palmer, How to Spot Good Gelato from 15 Feet Away, 2013)
Of course, a gelato place could fake it, by paying these costs and yet serving inferior products. It’s just that in practice, humans are quite correlated for this kind of thing: someone taking the pain to do most of the things the right way is far more likely to get all the way.
Another one of this type that Palmer doesn’t mention, but which I’m sure she would agree with, is whether they serve a chocolate sorbetti. That is, a chocolate gelato without dairy. This is done with dark chocolate, and is delicious iff done with high-quality chocolate. So it acts as a test of the quality of the chocolate, and yet another proxy of the amount of efforts and sourcing that the gelateria took.
The fior di latte and fior di panna are similar, but not exactly about cost. My understanding is that they’re not necessarily more expensive (like nocciola), and don’t require constant changes (like seasonal fruits), but instead are particularly revealing of the fundamental ingredients of gelato: milk and cream. Offering these flavors reveals the underlying quality of the dairy products, and as such is only really done when they can stand on their own.
These flavors, made from pure milk and pure cream respectively, are the basic form of gelato. It means they are the flavors that most clearly expose the quality of the milk, and most clearly betray the presence of artificial additives. In Italy, virtually all gelato places will offer fior di latte, and any one that doesn’t is conspicuous. Abroad, especially in the US, it is much more rare, because it exposes inferior ingredients, and few non-Italians know what this flavor is (Americans, for example, always ask for vanilla instead, because we’re not used to the idea that the pure white version of a frozen desert could be so good as to require no flavor, not even vanilla). If a non-Italian gelato place offers fior di latte, it’s often a good sign. If an Italian one offers fior di panna, that is a sign that they have put in extra serious effort into maximizing the flavor of their dairy (and cream is more expensive than milk) so also good. But if they only offer fior di latte with chocolate chips, or with flavored syrup drizzled all over it, they could be showing off their syrups, or they could be covering inferior milk.
(Ada Palmer, How to Spot Good Gelato from 15 Feet Away, 2013)
Phase 2: Fine Details
These last few steps (in my ordering) are the trickier to get right, because they demand a discerning eye about color and texture.
I can sort of perceive the color point, which is mostly about discerning if there are artificial colorants in the fruit gelato:
Banana, apple/pear, or berry flavors (frutti di bosco) are the easiest tell. If the fruit gelati are made of pure, real fruit then they will be the color that fruit would be if you crushed it: berry flavors a deep dark off-black purple/red, apple white or brownish or yellowish sometimes with flecks of peel, and banana a rather unappealing shade of gray. If, on the other hand, banana is a cheery yellow, apple a perky spring green and berry flavors are the light-ish color of blueberry yogurt, then the gelato before you is a mix of milk with food coloring plus fruit extracts or artificial fruit flavor. Pistachio similarly should be the color of crushed nuts, not bright green. The artificial fruit gelati can still be delicious, but only pure fruit sorbetti will give you the overwhelming flavor of top quality fruit gelato which tastes more like fruit than the fruit does, hyperconcentrating the fruit’s flavors and bringing them out with sugar.
(Ada Palmer, How to Spot Good Gelato from 15 Feet Away, 2013)
Another related point that is not given its own step in the post, but mentioned in passing, is whether cream is added to the fruit gelato in general. This is both about the non-alteration of fruits, and also because pure fruit gelato are more expensive and demanding, so they signal artisanal care (once again).
This matters even if fruit isn’t your thing: making the gelato out of pure fruit is more laborious and expensive than using flavor extracts, so a gelateria with a brilliant dark frutti di bosco is one that is definitely trying to produce the best, and thus also likely to produce a superior chocolate, crema, etc. Now, sometimes mixes of fruit with dairy can be good, so a blueberry-yogurt-colored frutti di bosco isn’t always a bad thing, but the pure fruit ones are more difficult and more expensive, so they are always a good sign, even if the opposite is not necessarily a bad sign.
(Ada Palmer, How to Spot Good Gelato from 15 Feet Away, 2013)
Then… there is the step that I cannot for the life of me ever get right: the translucency of the lemon.
A small gelato place may not have any of the more telltale fruits, but lemon is pretty much always in stock. Is the lemon an opaque, creamy white that looks rather like the white cream-based flavors? If so, it is milk mixed with lemon extract. If, on the other hand, the lemon is translucent white or subtly yellowish off-white, so the edges of it are almost transparent like the transparent outer edge of an ice cube that’s in the process of melting, then it is just water and fruit extract. This again is a bit more difficult and expensive, because it requires better lemon juice to taste good, and is harder to make stay firm, so again it means the gelato makers have put in more effort.
(Ada Palmer, How to Spot Good Gelato from 15 Feet Away, 2013)
In practice, this is the same kind of proxy than most of the ones above, just much harder to catch. I just don’t use it that often, because I can’t reliably see it (I do not eat nearly enough gelato, probably).
Applying the Algorithm in Ambiguous Situations
If you’re lucky, you will find a gelateria that checks all these steps, and you can go there with eyes closed (Perchè No! in Florence, recommended by Palmer, is such an example that we tried with my wife on our honeymoon).
But what do you do if it’s only a partial validation?
My own line is that I never go for the gelateria with huge mounds. Then, if I want to go for dairy gelato, I will focus on the steps about nocciola and fior di latte (maybe a chocolate sorbetti), as these are more representative. And if I want to go for fruits, I’ll focus on checking for at least one non-standard seasonal fruit, and that the colors are neither exploding with neon brightness nor dulled with dairy.
If all else fail, Palmer proposes the following contingency:
Despair not! You can still have a delicious experience at a mediocre gelato place, you just need to choose your flavors appropriately. Usually if a gelato place is mediocre, it is working with inferior milk, and may have to put additives in the gelato to make it stay soft overnight because they can’t afford to make a new batch every day. These problems can be tasted easily in pure, simple flavors like fior di latte or the fruit sorbets, but you can choose flavors that conceal them, and thus still have a good experience. Chocolate is a reliable fallback in almost all circumstances. Another thing to look out for in mediocre gelato places is a complicated flavor mixing two or more flavors, like tiramisu. A place local to me here has very disappointing sorbets, but respectable chocolate, tiramisu, and pistachio, and remarkably good creative original flavors like root beer or “Elvis” (chocolate, banana, peanut butter) which are well balanced and conceal the mediocre undertones nicely. Lemon is also a good fallback. Whenever I have a rough travel day in Italy, I go to the train station gelato place and get a nice cold lemon, and even the terrible gelato they have in a train station is still miraculously curative after a rough day
(Ada Palmer, How to Spot Good Gelato from 15 Feet Away, 2013)
Appendix: True Cones
This is not in Palmer’s algorithm per se, but I’ve added a last step about choosing between cones and cups.
By default, I’m a huge cone afficionado. But there are two types of cones: what I call cardboard cones, and waffle cones.
The algorithm is simple: never get a cardboard cone, and if they offer a waffle cone, you can choose this if you like cones.
Note that this is not useful to judge the quality of the gelato. I almost think that it is inversely proportional to the quality: a lot of good gelateria I tried offer cardboard cones, as if the owner were pranking the tourists.
See this previous blogpost on her recent book about Renaissance historiography. I notably recommend her blog post series on Machiavelli, which is also about the city Machiavelli cared about most, Florence.
It also fit say, the culinary culture of Japan. But it’s not the only option. For example, French cuisine emphasizes transformation, as seen in the stocks and the mother sauces for example.
Note that Palmer herself recommends going in the “multi-flavor”/max-processing direction if the only options are “bad” gelato, because it hides the inferior quality base products.