Overnight Pizza and The Consistent Unreliability of Expert Guidelines

Author disclaimer: As always, keep in mind that I am at best some random guy on the internet. This article deals with food safety - my official recommendation on this is that you only eat food that has been properly stored, sterilized right in front of you, then tested for any possible risks by a crack team of highly-trained scientists.

I’ve been thinking a lot about overnight pizza. If you will, consider this scenario:

As a classy person, you decided to spend your evening as all persons of class do: you ordered a pepperoni pizza, drank several ounces of mid-quality bourbon, and fell asleep on the couch watching Good Omens. The next morning, your sleep-fogged eyes light upon a dilemma: the last few slices of pizza sit, boxed, on your coffee table. You forgot to refrigerate them.

You check to make sure, but it turns out you didn’t stop liking pizza overnight. Your hangover is crying out for the headache-soothing effects of the grease. At the same time, you don’t like violent stomach cramps much and like dying even less. Do you eat the pizza?

The two sides of this argument are pretty well represented in the self contained environment of this tweet1:

The two sides here are roughly “common-sense” vs. “deference to authority”. I think that Richard (the video guy) is working from an experience-driven opinion - basically, that while “leaving food out” is something that can give you food poisoning in theory, leaving pizza out overnight never actually does this in practice. He’s done it a bunch of times, I’ve done it a bunch of times; we both know people who have done it a bunch of times and it’s never actually panned out that anybody got ill or died from it. For us, it’s wasted food for no reason at all.

Conor is working from the opposite side of things - he seems to think of Richard as a particular sort of uninformed hick. From Conor’s perspective, Richard doesn’t seem to be aware of food handling guidelines as sent down by the experts, and is taking his life in his hands every time he orders out because he doesn’t understand the things Conor and the science-is-good set do. Thus this is simultaneously “a dangerous food safety thing people commonly do” and “relatively high risk” (Conor regrettably does not say what this risk is high relative to; it’s almost certainly high compared to eating a fresh saltine, less so compared to doing wheelies on a motorcycle in traffic).

(Note: I’m reading a lot into what Conor said here. The general “people who don’t believe X proclamation from official sources have no good reason” stance certainly exists, but it’s possible Conor meant something different. I don’t think I’m reading it wrong, but for fairness’ sake assume any references to Conor from here on out are of my scapegoat version of him - I need a name to assign to one side of the argument, but I don’t want to unfairly beat on him based on something as easy to misinterpret as a tweet.)

This basic conflict is pretty representative of a whole spectrum of regard for what science declares as opposed to what a person actually experiences, ranging from people who will believe a Facebook post from a friend claiming essential oils cure cancer against any amount of science all the way over to someone who will acknowledge known lies from some trusted science authority yet cock their head at you like a confused cocker spaniel when you suggest this might diminish their trust in future statements from the same sources.

That’s interesting, but a little broader than I can cover in one article. What I’d like to focus on here is a much more specific concept - basically, that the Conor-esque set should know better than to take proclamations of impending death from overnight pizza at face value because those same proclamations are driven from a sort of exaggerated lie for conservatism’s sake. Since we can expect that lie in a significant range of situations, it’s worth it to break down what it is and how that happens.


Before I can break down the particular routine inaccuracy I want to warn you about, best practices demand I show my homework. To do this, I first need to cover where the official “don’t eat overnight pizza” advice comes from and why it’s generally unreliable in its usual form.

When you start dipping your toes into the murky depths of the is-old-pizza-going-to-kill me pond, the first stop is articles like this and this, essentially low-quality advice websites universally on Conor’s side and parroting advice that overnight pizza will possibly kill you. The authority being drawn on here is the USDA, whose rule on this kind of thing is as follows:

If a perishable food (such as meat or poultry) has been left out at room temperature overnight (more than two hours) it may not be safe. Discard it, even though it may look and smell good. Never taste a food to see if it is spoiled. Use a food thermometer to verify temperatures. Never leave food in the Danger Zone over two hours; one hour if outside temperature is above 90 °F.

Where a perishable food is defined as:

Examples of foods that must be kept refrigerated for safety include meat, poultry, fish, dairy products, and all cooked leftovers. Refrigeration slows bacterial growth.

Note that the USDA’s standard here is much, much harsher than “overnight”; any food that’s been out more than two hours has to be chucked. Thus if you or Conor have ever had any guest go back at 8:00 PM for a second slice of a pizza that arrived at 6:00 PM and didn’t slap it out of his hand, you and he stood just as much in defiance of the USDA’s guidelines for this sort of thing as any overnight-pizza eater.

Pizza being included as a “cooked leftover” is telling here - we right away know that the USDA has created a catch-all category. Essentially, they don’t trust you to know the difference between a slab of microwaved fish you bought from a guy in an alley and a home baked saltine cracker in terms of contents and perishability, so they’ve told you to throw out all food regardless of ingredients and cooking method at the 2 hour mark regardless.

This is where strawman-Conor and his set stop reading - there’s an official word on this, so it’s a done deal - if you go any further than that, you are a person who fails to “trust experts” in a way that implies personal shortfalls they don’t view that much different than, say, being anti-vax or believing that video cameras steal your soul. But what they are essentially accepting is a lie-through-wording-choice.

Remember, we want to know if it’s safe to eat such-and-such food after such-and-such time; the USDA tells us no, but what they actually mean is something like “some foods might be a significant risk after several hours, but we can’t trust you to sort out which ones, or to keep close track of your cut-off time, so we are saying two hours is the limit for virtually all foods to keep your dumb ass safe and our smart asses covered”.

I think most people know and recognize this kind of bureaucratic lie/overstatement/overcaution as what it is at face value, but it’s not really fair for me to stop there; for one, I have to justify that this probably doesn’t apply to some perishable foods as defined by the USDA (or if it does, that they can’t prove it does) to show at least one exception to their rule. Plus at least some of us sort of want to know whether we can eat that pizza - this hangover isn’t going to cure itself.


Between toppings, sauces and crust choices, pizza is an unusually diverse food; for the purposes of this discussion we are imagining a flat-stock sort of pizza as might be ordered at a large franchise pizza parlor: average crust thickness, cheese, some sort of basic tomato-based pizza sauce, and pepperoni. I acknowledge this is going to miss an awful lot of edge cases, but the existence of things like sashimi pizza means I have to draw the line somewhere.

First, we can set a sterility baseline pretty easily: Foodsafety.gov seems to set the highest internal temperature needed for food safety at 165 F; luckily, baking almost always meets this standard. So right away we have a sterile baseline - i.e. the red-hot pizza they slid into a box is at that time about as safe as any food can be.

Then we have to look at individual ingredients. Bread is easy, as even the hyper-paranoid USDA acknowledges that bread can be stored at room temperature for days:

Commercially baked breads and rolls can be stored at room temperature for 2 to 4 days or 7 to 14 days in the refrigerator. Bread products retain their quality when stored in the freezer for 3 months. Any breads containing meat or hard cooked eggs must be refrigerated within 2 hours.

The USDA here says “breads containing meat” are subject to their ultra-conservatism, though; the idea seems to be that the meat could infect the bread, or at least that being wrapped in bread does not significantly protect the meat. So we need to move up the chain immediately to the pepperoni. Luckily this is pretty easy, too, since pepperoni is a dry-cured sausage; the whole point of that class of food is that it’s shelf-stable. I can be corrected here, but it’s hard to imagine a super-compelling reason why it would lose so much of its preservation value from being cooked that it would drop from “weeks and weeks” to “less than ten hours”, especially as cooking it dries it out somewhat and water content is one of the main drivers of bacterial growth in the first place.

That leaves us with cheese and tomato sauce. Both have been linked to food poisoning outbreaks here and there - it’s important to see why. In the case of cheese, it’s almost always related to flaws in production and storage on it’s way to you. See here - note that the big risk factors are either being made with raw milk, post-milking contamination, or both. Improper storage would make this problem worse as bacteria counts might grow as the cheese works it’s way through the supply chain to your local pizza place, but all of that would be stopped by the cooking process. Any toxins accumulated over that time would still potentially be present, but there’s just not enough of a record of “problems caused by home storage” that I can find to indicate this is a big problem, even in the short term. (If you can find something showing this isn’t true, please show me so I can amend).

Tomato sauce is similar. The average tomato sauce has a PH in the ~4.5 range; that’s acidic enough to be pretty unfriendly to bacteria, even if we wouldn’t expect it to keep us safe forever. That being said, about the closest I can come to a document case of tomato sauce related food poisoning is this 20-year old who died from eating spaghetti. His case is not a lot help though, since he had apparently left the spaghetti unrefrigerated for five days, and since the toxins they found were in the noodles, not the sauce.

Taken altogether, we have four basic components in play here, none of which seem to have any significant history of food poisoning where short term improper storage at home were at fault. All four ingredients are further sterilized just before our 8-12 hour improper storage experiment begins, basically making it a best-case scenario for unrefrigerated foods. This is of course an as-near-as-I-can-tell; nobody studies this, since everyone defaults to the USDA’s “if you lose eye contact with any food for more than a shake of a lamb’s tail, burn it in an industrial furnace” food storage standards.

Even taking the “food as a whole” approach, we don’t find anything, with all the documented cases anybody talks about having nothing to do with post-sale storage. So where straw-man Conor, the USDA and several dozen shitty SEO websites tell you eating overnight pizza is unsafe, it’s a lie at least in the sense that they don’t know it is; there just doesn’t seem to be any evidence at all or any studies backing them up.

The strongest point I could make here is actually one that nobody will accept - you already knew eating overnight pizza was fine. People do it all the time; millions of people eat pizza a week and some not insignificant portion of them are eating it the next morning. If it were a significant danger, we’d be body-bagging dozens of college students a week - you’d not only hear about it on the news, you’d be within a couple degrees of separation from someone it had happened to. This kind of anecdotal evidence should be rejected where carefully gathered statistics say it’s wrong, of course. But those statistics don’t seem to exist here.

Richard-from-TikTok and I have our own possible failure states in this argument (say, rejecting good data) but strawman-Conor also has his. When an authority tells him something he isn’t preregistered as disagreeing with, he believes it, full-stop: to do anything less would be to potentially fall in with the anti-science set. In a small but real way, this hands off an awful lot of un-earned control to groups much more concerned with protecting themselves than any other motivation.


I think this kind of thing matters a lot. In the specific case, I want to eat pizza the morning after ordering it. My track record clearly shows I’m not going to remember to put it in the fridge. Taking the <personal opinion> USDA’s bullshit mega-overcautious-standard </personal opinion> as true here would mean I’d be depriving myself and wasting resources, which I don’t want to do.

But in the general case it matters more, because the implication here is that there’s a whole class of expert opinion that in general can’t be trusted. If the USDA is less than honest here, we know why: they don’t get in trouble for you throwing away good food, but they potentially get in lots of trouble if you eat bad food on their recommendation. Their incentive in that situation is to be almost comically conservative - they’d tell you not to eat anything but fresh-baked rubbing alcohol if they could get away with it, but since they can’t they throw up the harshest standard they can while still having any chance of anyone taking them seriously.

In the USDA’s standard this is probably a net benefit - you basically know they are lying at all times, but from a utilitarian point of view the precautionary principle probably works as well here as anywhere. But we already know situations where it’s a disaster, i.e. the FDA saying there’s exactly one very slow way of testing that’s acceptable, handed down by god, and that only this way can ever be considered no matter what the risk-reward tradeoffs may be. In that case, us accepting lies-for-the-sake-of-professional-caution as truth and not demanding more killed millions.

For obvious liability reasons, I can’t recommend you eat food that’s been sitting out; my official “don’t-want-to-get-sued” recommendation is, as mentioned above, that you don’t eat any food that wasn’t sterilized in front of you and tested for any possible danger by a trained team of scientists. But I do recommend you look at professional recommendations where cover-your-ass perverse incentives are in play with at least a little suspicion. Losing a delicious leftovers breakfast is bad enough by itself, but there’s more than just pizza at risk here.

1

The Pizza video comes from Richard Tyree’s Tiktok here.