Almost every font on almost every worst-fonts list is fine. Comic Sans, Papyrus, Curlz, Brush Script — these are competently drawn typefaces that got put to work in places they were never designed for, and then took the blame for the person who put them there. The genuinely bad ones are boring, and nobody argues about them.
I want to defend the famous ones first, because the defence explains what actually makes type bad.
Comic Sans was made for a product it missed
Vincent Connare designed Comic Sans at Microsoft, released in October 1994. He made it for Microsoft Bob, a piece of software with a cartoon dog that spoke to you in speech bubbles — and which, when Connare saw it, was setting those speech bubbles in Times New Roman. A cartoon dog talking to you in the typeface of the London Times. He was right that this was wrong.
He had The Dark Knight Returns, lettered by John Costanza, and Watchmen, lettered by Dave Gibbons, in his office. He drew a face from that tradition: hand-lettered, informal, cheerfully irregular. It is not an imitation of either, but it comes from the same place.
Then he finished it too late, and it never shipped in Bob at all. It went out with 3D Movie Maker instead, another product with cartoon guides and speech bubbles, and from there into Windows — onto tens of millions of desktops, sitting in the font menu next to Arial and Times New Roman.
That is the whole story of the hatred, and it has nothing to do with the drawing. For roughly a decade, Comic Sans was the only informal typeface most people owned. Anyone making a birthday flyer, a lost-cat poster, a church newsletter — anyone who wanted to sound like a human rather than a memo — had exactly one lever, and pulled it. Including, inevitably, people announcing funerals and delivering test results.
Blaming the typeface for that is like blaming a hammer for a room full of things that turned out not to be nails. The design brief was cartoon speech bubbles. It performs that brief well. It was simply the only tool in the drawer.
Papyrus is a twenty-three-year-old's sincere idea
Chris Costello drew Papyrus in 1982, aged twenty-three and just out of college. He had been reading the Bible, and started wondering what writing would have looked like in the Middle East at that time — rough surface, reed pen, irregular edges. He hand-drew the whole thing over about six months with a calligraphy pen on textured paper. Letraset released it in 1983.
Read that again as a piece of work rather than a punchline. Six months of hand-drawing, from a coherent idea, by someone barely out of school. Most typefaces are worse than Papyrus and almost none of them are as clear about what they are trying to be.
What happened next is the same thing that happened to Comic Sans: it shipped with everything, it became the only vaguely ancient or organic option most people had, and it went onto every yoga studio, herbal tea box and Mediterranean restaurant menu in the world. Then it went onto a blockbuster film's subtitles, and became a joke with a much larger audience.
Costello has been remarkably gracious about all of it, which is more than most of us would manage.
So what is actually wrong with them?
Nothing, in the drawing. The problem is ubiquity plus a narrow brief.
A typeface designed for one specific tone, distributed to everyone, will be used for every tone. The face itself carries a strong voice — that is what it was hired for — and when the voice does not match the message, the mismatch is loud and it is the typeface you can see. Nobody looks at a funeral notice in Comic Sans and thinks poor tool selection by the bereaved. They think that font is awful.
The tell is that neither face is criticised on typographic grounds. Nobody says Papyrus has bad spacing, or that Comic Sans has a weak italic. The complaint is always about where it turned up. That is a criticism of a decision, not a drawing.
The genuinely bad ones
Here is what actually makes a typeface bad, and none of it is fun to argue about:
- Characters you cannot tell apart. A capital I, a lowercase l and a numeral 1 drawn as the same vertical bar. An O and a zero with no distinguishing mark. This is the failure that costs someone a wrong digit in a serial number, and it is the one thing where the typeface genuinely outranks the setting.
- Spacing that was never done. Plenty of free fonts are a nice set of drawings with no kerning and default sidebearings. They look fine as a word in a specimen and fall apart as a paragraph, with holes after every capital and collisions on every AV pair.
- A character set that stops at the edge of English. No accents, so the moment someone writes café or Zoë the browser silently swaps in a different font mid-word.
- Weights that are not related. A "family" where the bold is just the regular with a stroke added, so the counters close up and the proportions shift. You notice it instantly and cannot say why.
- Hinting that gives up on Windows. A face that is beautiful on a Mac and a smear on a ClearType screen at 13px. Half the world is on the other machine.
These are the real failures, and they are almost never the fonts on the lists — because a font has to be popular to be hated, and popularity usually means somebody did the boring work.
The campaign, and its ending
Ban Comic Sans was started by Dave and Holly Combs, graphic designers in Indianapolis, after an employer insisted they use the font for a children's museum exhibit — which, in fairness, is close to the one context it was built for. It became a manifesto, stickers, a genuine internet movement.
A note on sourcing: reference works disagree on the year it began, giving both 1999 and 2002. I have left the date out rather than pick one.
The part worth knowing is how it ended. The Combses wound the campaign down in 2019, with Dave Combs saying the hatred had got out of hand. The joke had stopped being a joke about typography and had turned into a way of sneering at people who had never been taught to choose a typeface and had no reason to know.
The campaign against the world’s most hated font was ended by the people who started it, because the hating got mean.
The uncomfortable footnote
Comic Sans is widely used in schools, and it has two properties that legibility research actually supports: characters that resist being confused with their own mirror images, and unusually wide spacing. Those are real virtues, and they are the same mechanism that makes a licence plate readable.
That is not a claim that it is a remedy. No typeface is — the research on faces sold for dyslexia has repeatedly failed to find a measurable benefit. But it does mean the most mocked typeface of the last thirty years is doing something defensible for the readers who use it most, and the people mocking it mostly have not checked.
What a type designer takes from this
Two things, and the first stings.
If you draw a face with a strong voice and give it away to everyone, you are signing up for it to be used badly, publicly, forever. That is not a reason to draw bland type. It is a reason to be clear-eyed that reach and control are traded against each other, and that the trade is permanent.
The second is about where criticism should point. Choosing type badly is a skill problem, and skill problems are fixable with a sentence of advice. Ridiculing the tool teaches nobody anything and mostly signals membership. If a menu is set in Papyrus, the useful response is that the face is doing all the talking and the words cannot get a word in — not that the owner is an idiot.
Vigil is a quiet face on purpose, which is a different bet: less voice, more range, harder to use conspicuously wrong. That is not virtue. It is just a different trade, and it comes with its own failure mode — nobody ever loved a typeface for being sensible.