Wrote this only to remind myself to keep practicing my writing. Persons criticizing the stodgy language or lack of polish will be prosecuted; persons pointing out the copious bullshit will be banished; persons noticing this disclaimer’s unoriginality will be shot.
By Order of the Author, Per G.G., Chief of Ordnance
Also, I am at most 50% serious about all of this.
Humor and jokes have experienced a great surge in popularity, which can attributed to the creation of more efficient delivery systems, like the Internet. However, the field of humor analysis is still notably underdeveloped. We posit a hypothesis on a formula describing the humor of a joke as a product of factors that depend on the speaker, the joke’s content, and the listener, in the hopes that this will inspire society to think about why it considers something to be funny, instead of taking it as a given. As evidence, we use material from professional comedians, and math puns that have made the rounds in mathematician subculture folklore.
One common criticism against the field of humor analysis is that explaining a joke ruins the joke. Thus, much like the goto statement, detractors say that analyzing humor should be considered harmful. This misconception is misguided and often forms the bases of most criticism against the field, so we would like to briefly explain away these worries.
First of all, humor analysis is the study of humor in general. This involves making grand, sweeping statements about why we consider jokes to be funny, without leaking or exposing any information about the jokes themselves, and this indirect analysis is what lets us make progress while minimizing joke destruction.
Secondly, it is arguable whether explaining a joke is harmful at all. Suppose a listener hears a joke, gets it immediately, then hears an explanation of the joke. That listener’s enjoyment is not ruined by hearing that explanation, because the listener already has an implicit representation of that explanation in their head. Now, suppose a listener hears a joke, and does not get it. Then that listener has so far received zero entertainment from the joke, and hearing an explanation can only increase that entertainment. And in some rare cases, explaining the joke can itself be a joke. This is known as meta humor, and is outside the purview of this post. For further discussion, see 
We admit that this argument for humor analysis is informal, and that there are other complicating factors, To be frank, we will not give a shit, and will proceed with the analysis regardless.
It is a well known fact that including unneeded mathematical notation makes text more persuasive and legitimate. Thus, we formalize some notation that will be used in later sections.
\(J\) will represent the set of all jokes, and \(j\) will represent a joke from that set. \(P\) will represent the set of all people, and \(p\) represents a person from that set. \(H: P \times J \times P \rightarrow R\) denotes the humor function. We assume humor depends on the person saying the joke, the joke’s content, and the person hearing the joke. This will often be written as \(H(p_s, j, p_l)\), where the subscripts indicate “speaking” and “listening” respectively. The output is a real value, denoting how funny \(p_l\) finds joke \(j\) when it is delivered by \(p_s\).
Lest people worry about drowning in notation, we have endeavored to explain all further equations with English as well.
Proposed Humor Formula
We propose that humor is the product of three factors: the inherent funniness of the joke when delivered by the speaker, the amount of education needed to understand the joke. and the amount of sleep deprivation the listener has.
More formally, let \(I(j, p_s)\) be the inherent humor joke \(j\) has when delivered by \(p_s\), \(E(j)\) be the amount of education needed, and \(S(p_l)\) be the amount of sleep deprivation the listener has. Then\[H(p_s, j, p_l) = I(j,p_s)E(j)S(p_l)\]
We spend the rest of the post explaining the reasoning behind these terms.
The Inherent Humor Term, \(I(j,p_s)\)
Clearly, different people will find different jokes funny. Given the wide variability, it doesn’t seem reasonable to talk about inherent humor. However, professional comedians are a counterexample to this line of reasoning. These people make a living doing stand-up or creating comedy sketches that a wide range of people enjoy. They can do this because most of their jokes are inherently funny. Behind the scenes, show hosts like Jon Stewart and Stephen Colbert spend hours crafting jokes that are likely to work on a wide audience (that have high inherent funniness), then spend more hours rehearsing their material (improved delivery makes a joke funnier. Note this is why this term depends on the speaker as well as the joke.)
The Education Term, \(E(j)\)
Fig 1. Prior work on the link between education and humor. 
Consider the following example jokes from math.
Q: “What’s purple and commutes?”
A: “An abelian grape!”
Q: “What’s sour, yellow, and equivalent to the Axiom of Choice?”
A: “Zorn’s lemon!”
The first joke requires knowing abstract algebra, since it is a play on “abelian group”. The second requires knowing some foundational mathematics, as it is a play on “Zorn’s lemma”. Both of these jokes have terrible inherent humor. (Should you disagree, we advise getting a second opinion immediately.) But, because they use obscure concepts from math at the university level, they still demand a chuckle from people “in the know”. We call this the phenomenon privileged shared knowledge. People with privileged knowledge know something that the average person does not know. A group with privileged shared knowledge knows they all know something the average person does not know. When someone delivers a joke, it is implied that person knows the background required to understand the joke as well. Thus, listeners “in the know” have privileged shared knowledge with the speaker. This triggers a deep seated tribal mentality, where the tribe is everyone who knows abstract algebra, or everyone who knows about the Zorn’s Lemma. Once included in this tribe, the listener gives the speaker and joke much more leeway, and will consider what the speaker says to be considerably funnier than it actually is.
As further evidence, consider this especially extreme case.
It’s an old joke that a mathematician is a device for turning coffee into theorems. However, it’s also true that a comathematician is a device for turning cotheorems into ffee.
The first joke is slightly funny on its own, and is further elevated because it mentions “mathematicians” and “theorems” in passing. However, the second joke has no inherent humor at all. It is only funny if you know a little bit of category theory.  Then, as if by magic, it suddenly becomes humorous. Not only do you need to know category theory, you need to know that there is an old joke that mathematicians are coffee -> theorem converters. By this point, anyone who has all the requisite background information must find it funny.
This also explains why people feel “left out” when they “don’t get” a joke. Because they do not have the privileged shared knowledge, they are excluded (or “left out”)) from the tribe. They are forced to rely on the inherent humor, and unfortunately jokes like these math puns have terrible scores in that department.
As a side note, this observation seems naturally extendable to Internet memes and in-jokes. For memes, the privileged shared knowledge is knowledge of the meme itself. For in-jokes, the privileged shared knowledge is the context in which the in-joke was first created. We leave finding examples of this to the reader, but suggest Reddit threads as a good starting point.
The Sleep Deprivation Term, \(S(p_l)\)
The final term we propose is the tiredness of the listener. When especially tired, people are more likely to find things funny than they normally would be. One possible explanation for this is that it is a subcategory of the privileged shared knowledge, where living through the context of no sleep is the common information. However, this does not accurately why a joke that is funny at 3 AM is no longer funny the next morning, as everyone involved should still have the shared context. Somehow, the joke becomes less funny, even when said from the same speaker to the same listener.
Therefore, humor must depend on some temporal structure. We propose that sleep deprivation makes jokes funnier because the brain is less able to judge a joke due to stress and fatigue. When we are told we are about to hear a joke, our brains start expecting humor, and primes our mental state with a baseline level of funniness. As the brain hears and considers the joke, further adjustments are made to this value, and eventually we decide on the humor level and laugh/facepalm appropriately. When tired, the listener’s brain gets less computation time, giving less time to adjust the baseline value. In this scenario, the final humor output relies more on the primed state, explaining why jokes can be funnier deep into the night; the joke itself was awful, and we simply did not have the time to realize it.
Note this also suggests that especially good jokes are not as funny when tired, since our brains have less time to adjust the internal humor value upwards. This lines up with our anecdotal experience, and therefore must be true.
The great surge in viral videos and memeing, as fueled by the Internet, has made humor analysis an exceptionally useful field. Should the trend continue, we believe considering questions like this one will continue to be useful in the future. We hope this incredibly dumb post inspires future work, which will ideally be much more insightful.