Interesting concept, but 100 words is really quite a lot to get through... It's tiresome trudging through the easy words at the start, and I never got to see the interesting words before getting bored.
I've seen other systems like this calibrate far more quickly by assigning a sort of score and confidence behind the scenes. Confidence starts out low and increases over time - correct/incorrect answers rapidly adjust score at the beginning, then things settle down.
In practice this means you get a sequence of increasingly uncommon words initially, until you get one wrong, then you drop back to something easier until you start getting things right again, and eventually circle around words at your level.
Also - too many clicks per word. It's low stakes, just let me click the definition once and I'll live if I misclick (or add an undo button).
> Also - too many clicks per word. It's low stakes, just let me click the definition once and I'll live if I misclick.
This, and accept that people will have incorrect input and build it into the confidence. Even the smartest person in the world sometimes makes clerical errors, or has the wrong neuron fire at the wrong moment.
100 is too many? Thats two or three minutes at most.
I would suggest a bias in this test towards reading. More than a couple are words i know but rarely see in print. But maybe im too much a fan of british TV so i hear many of thier words without seeing them written down.
They’re also too far away. I’m on a laptop and I have to keep moving the cursor up and down just to confirm. Give each option a letter or number and let me press it to choose the answer¹.
¹ There is (was?) some service for forms which does that and it works quite well. I think it was Typeform, but I just opened the website to check and—of course—it’s now just plastered with mentions of AI so I lost interest in verifying.
> I'm guessing it's testing our susceptibility to machine-generated compliments
I fail to see the point. For one, the compliments aren’t particularly good or interesting; for another, I didn’t even read them (I just went back to check after your comment), I simply clicked when seeing green.
I suggest skipping the submit button and just showing it's correct when pressing and moving on after a sec or so. Having to click on submit twice really breaks the flow.
Also in all the words I tried I noticed out of the 4 options one is the correct one, another is the opposite of the correct one, and the other 2 are random stuff. You can basically skip any option whose antonym isn't present as well.
It'd also be a lot less awkward to go through 100 words if it had keyboard shortcuts (1-4 for the words, enter to submit) and if they fixed the layout shift jank
It estimated 74k words for me, but I feel this might be inflated; much of the time when I didn't know the answer - I could vibe guess it just as you did it. The distractor answers weren't convincing enough. For starters, when an answer was based on deconstructing the word into common English words, that ruled it out. After all, if it was, then it wouldn't have been obscure.
A tangent: writing distractors for multiple choice questions is hard. From the exams I know (excluding those whose nature precludes it, such as based on calculation or rote memorization) the only that does this brutally well is LEK (Polish medical graduate exam). It's nigh impossible to vibe guess it at more than random chance for someone outside the field.
Yeah I also got exactly 74k. Stuff like “xylologist” I guessed had to do with vegetation because of “xylem”, whereas xylophone player was too on the nose. Then again, maybe knowing xylem in the first place makes 74k reasonable.
Xylophone was still a good clue, since the etymology explains itself immediately even if you don't remember all your Greek roots. It's a wooden instrument, so that's xylo + phone, so now you at least know xylo.
Xylologist would have to be someone who studies wood. The word I had was xylopolist, which is wood + seller.
But at this point it's not really specifically about how many English words you know, just how many you can understand from etymology.
I think the test was vibe coded, because a xylologist is someone who studies wood, not someone who sells wood. I am not sure if "xylolgist" was the exact word, though.
xylo- = wood; -logy = study
Indeed from M-W: "a branch of dendrology dealing with the gross and the minute structure of wood"
66k for me, but I didn't get that word, instead I got ones like Hippopotomonstrosesquippedaliophobia, Flibbertigibbet, and Brobdingnagian... which the latter two interestingly do show up in my keyboard's word completion suggestions.
It would have been nice to have an “i don’t know” button. Instead I decided to select the first option for words I didn’t know instead of trying to figure them out. Although when I got to the final group I couldn’t resist trying to figure them out. It estimated 61k for me.
I guess, it's not too bad for a non-native speaker.
Minor feedback:
1. The correct answer for "Lethargic" is "Affected by lethargy". I think, definitions should not use words that share common root with the defined word, because:
a. it makes guessing too easy
b. it basically becomes a circular definition which is meaningless
2. Options almost always include 1 correct answer, 1 direct opposite and 2 completely random. Once you learn to recognise it, you can easily rule out 2 random options and have a 50/50 guess.
I have a copy of the shorter Oxford English Dictionary from 1970 which I inherited. It is two massive volumes and is only shorter in comparison to the full dictionary which is 12 volumes (more in more modern editions).
My shorter OED contains 163,000 words (compared to the 600,000 words of the longer).
According to this site I know 71,000 words... Let's test that against the OED. I should have about 43% chance if knowing a word picked at random.
In my totally scientific test (ha) I chose 50 words at random from the OED and discovered I knew 29 of them for a score of 58% which is more than two sigma from 43%, this disproving the hypothesis.
I forgot what that was now, but it was a fun experiment.
I also got something around 70-80k with 95/100 correct words (I don't know or use most of these words, but the later sections have a lot of words with Greek or Latin origin, which made them easy to guess). One of my wrong words was a misclick in the first section, which I think dragged down the estimate quite a lot. You may have done something similar. I assume they use a simple formula where early misses cost you a lot and late misses cost you very little.
Your method of sampling could be improved further, unfortunately at the expense of ease of use. If the dictionary was sorted according to difficulty, then you could use stratified sampling.
It seems like the right answer is usually the longest of the choices, I managed to get a few just by picking the longest. It would also be nice if there was a "I don't know" instead of guessing and skewing the results by getting it right, though maybe thats accounted for
These were likely all AI generated, or at least the alternatives were. I made an app a while ago as well, and afterwards realized AI often wanted to make a very covering answer for the correct one, making it often longer than the others, thus defeating the idea of the quiz in the process.
If the goal is to actually calculate how many words we know, then you should include an "I don't know" option. Sure, some people will choose to guess to inflate their score, but some of us will be honest because we legitimately want to know our scores.
If you force me to guess, then I'm going to guess. Not only does that give me a 25% chance of getting it right at random, but as others have pointed out, it is very hard to make a multiple choice question that isn't guessable by an astute enough test taker. I think I knew 80 - 85 of those words, but I scored 97, because those questions were very guessable.
Also, reiterating everyone else's comments with respect to the UX needing fewer clicks, and also the definitions not being exact or precise in many cases.
I only got 4 wrong as a non-native speaker. Okay, I'm widely read in English, but among LLM-generated definitions it's just too easy to spot the right one.
The UX is awful - I bailed out at 25/100 JUST IN LEVEL ONE (BASICS)
Might I suggest adaptive difficulty? After getting 10, 15, 20 correct in a row it should scale up the difficulty immediately, rather than waiting for 100 in the basic level 1...
These should maybe be checked through. Many are the second or third definitions, and some even reference the word in the definition e.g Lethargic: exhibiting lethargy
Good fun! At first I was scared of having to answer 100 questions, but when the words got more sophisticated it turned to be more engaging. Also, the result is good for self-esteem! :) Many thanks to the author!
I wonder if the test is calibrated to the fact that some answers are just well guessed? I am not a native English speaker, but I speak 3 languages overall and have basic notions in Latin, and I have to admit it helped a lot in "deciphering" a few words that I didn't know at all. And in at least 2 cases I just guessed correctly.
In addition to how much fun it was, it has potential pedagogic value for teaching sampling based estimation.
It would have paired well with an exposition of vanilla Monte Carlo and the benefits of stratified sampling.
Although stratified sampling is good, one can do better in this case by using adaptive sampling, where one uses a runtime (Bayesian) estimate of vocabulary to maximize information gain per question -- preferrentially sample from those strata where the current strata specific estimate has higher variance.
This was fun! And it told me I know 55k words which made me a little happy.
I'm not sure exactly how you did this, but I think you asked an LLM to come up with the wrong options. Two things to consider:
1. While the LLM can go r good options, they won't be always hard to guess. I wonder if instead you can have the LLM generate very close words (or skip using an LLM entirely) and put those as the options.
2. If you will generate options with an LLM, make sure you are mindful of its inability to shuffle things around. The correct answer was overwhelmingly the first or second option in the list. You should ask the model to give the options in a uniform order (say from true meaning then decreasing amount of replayability), then manually shuffle them so that the probability of which option (A, B, C or D) is always 25%.
You are a person of few words, or perhaps just a mysterious one. Quite intriguing.”
—-
This sounds more like a cute assessment of only getting two words right. And what do you mean “new words”? It wasn’t until eighty-odd words in that I actually got a word I didn’t know and had to guess by ruling out multiple-choice options.
Some of the definitions offered are slightly short of what I expect. Like for "Obsequious" it offers "obedient to an excessive or servile degree" which isn't wrong, but it misses the expression of a sort of noisy eagerness in that servility.
English has this weird dichotomy where most of the words in a typical sentence are Germanic, while most of the words in the dictionary are French.
Fun fact: according to a quick count by AI using web search, the previous sentence contains 21 words of Germanic origin, 2 of Latin origin, 2 of Greek origin and 1 of French origin. Also the etymology of the word Germanic is Latin, while that of the word French is Germanic
Norman French due to the Norman invasion of 1066 resulting in Old English evolving into Middle English. You can see that in the words for animals vs meats (cow and boef/beef, sheep and mutton, etc.) where the Germanic people raised the sheep and the Norman aristocracy ate them.
A lot of the more common and simpler words are Germanic, as is the grammar (e.g. compound words like cupboard).
Depends is bratwurst a German word or an English one? You will hard pressed to find an American that doesn’t know thr word and what it means. You can buy them at just about any grocery store and they are a staple of many restaurants.
At some point the word becomes both. Sourced from its mother language and maybe even still meaning the same thing in both, but no less an English word than any other at this point.
It also had "weltschmerz" in the list, but I think I have only ever heard "ennui" used in English. They are both foreign words, but I would not have thought of weltschmerz as a loan word. Then again, maybe I am not reading the right texts.
French english speakers usually have a quite good vocabulary. Getting to the point of speaking english is a milestone that's quite difficult for french speakers though.
Not from Latin but through French - the direct use of Latin in English is generally restricted to technical jargon and legal terms (that English often also share with the French.)
Latin isn't really any sort of parent to Old English afaik, even though the Romans ran Britain for a while.
Presumably it's a random batch of words since you can run the test again. I wonder how much the word selection affects the outcome. I got 66,750 with 20/20/15/17/14.
I'm curious how the difficult is chosen because "obfuscate" was included in the hardest difficulty but I would not consider that to me a difficult word.
Also I found that some of the definitions were not completely correct.
It could be based on things like word frequency. I'd expect obfuscate/obfuscation to be less common outside of programming and RPGs (Vampire the Masquerade).
There isn't a "correct" way to incorrectly render the interrobang as 2 separate characters. The name was never supposed to suggest a certain ordering instead of just being both at the same time. The name "interrobang" just sounded better than "exclamaquest" (or any of the alternatives Type Talks readers submitted).
This one's much better. Shorter, faster, adapts to one's level, gives an out for being unsure, largely doesn't bother with definitions (except the occasional verification challenge), and even mixes in some fake words to ensure you're not BS-ing.
Fun game! I did worse than many others here, only 69.9k estimated words. But then English is my second language, so I'm pretty pleased with the result!
Really interesting, but I would love to be able to express honestly when I just guessed. This way the result would be much more scientifically sound. Four answers have a 25% chance of random correctness, which is a bit high in my opinion. I think either adding a "I don't know" or a confidence level (Known/educated guess/wild guess) would help.
Got a bit boring then suddenly very hard with some really esoteric words at the end in the ‘grandmaster’ level. It’d be nice if it got progressively harder without levels.
Some definitions were not great and alternatives a little silly at times but on the whole seemed pretty accurate.
Also probably needs calibrated as 96/100 was projected to 77k words, what would the estimate be for 100/100?
I like this but it should be all operable with keyboard to be faster ie up down and 1234 for options and if its righht you just move on, maybe show synonyms in the success ui.
Depending on what you consider an "English" word, anywhere from 0% to 100% of words are English words. I've definitely seen accoutrement and ziggurat in English, and quite often.
Of course, the line is very blurry. I've used accoutrement(s) in English many times, but I've never considered myself to be speaking English when I use it. It's like joie de vivre or c'est la vie.
I got 70,750 which is much higher than I expected.
The early words were obvious.
However, a lot of the later questions
I could only answer because they were multiple choice.
If I had to actually come up with a definition,
I suspect my score would be much lower.
When there are two options that describe exactly the opposite of each other, it will be one of them. Reduced a bit the fun - but then again, for some words I understood what they are dealing with, but not whether positively or negatively.
Sandi Toksvig, the current host of the BBC program QI (Quite Interesting), previously hosted by Stephen Fry. She's also been on a number of other BBC TV and radio shows.
I suspect Sandi Toksvig, one of the hosts of QI. One of the 'success' messages is "quite interestng!".
No offence mean to anyone, but the whole exercise feels very QI : superficial 'understanding' of a large range of things (for example words) without much of a connection between these words.
YouTuber John Green has been working on finding a way to quantify the number of names the average person knows. IIRC, his estimate is around 5k, which seems both crazy and reasonable at the same time.
This reminds me of a learning resource that I can't find again: you start with an assessment of how many words you know and then you get new words in context with every session (and maybe some spaces repetition). It was mostly from newspaper articles and catered for every level of English. It was a website (ca 2013), not an app. Any ideas?
It might be nice if you could unlock a "hard mode" or ability to the first 1-3 levels after a first run. I scored a little over 81K and considered playing again because I like quizzes, but doing another batch of (to me) easy words seemed like a waste.
Super fun, got 70,250. Friends have always lightly ribbed me for having to go home and look up words i've used. Those remaining 100k words must be really obscure.
One suggestion would be more convincing decoy choices, some were pretty silly. But I have no idea how they come up with them.
Open any technical textbook in an area slightly outside your domain and you will quickly disabuse yourself of the notion that majority of words are obscure. Most complex words are just technical/jargon not archaic or forgotten.
As a fluent native speaker who has read thousands of books and sometimes reads dictionary entries for fun, a number of these definitions are actually slightly off.
"Verbose," for instance, is defined as "Using more words than are needed."
That's not exactly wrong, but it's kind of misleading. "Verbose" explicitly means using a large pile of words, drowning the reader in far more words than are strictly necessary.
"More words than are needed" could be as limited as "used a three-word construction in a sentence where it could have been one."
There are many more like this.
Please, I beg all of you - don't use LLMs to generate linguistic slop that claims to be linguistic education.
Literally when I got to advanced and beyond just picking the longer and more complicated looking answer was the right one. I think this test is extremely flawed.
I got 97/100 (80.5k) by picking the answer that has no relation to the word. Most of the incorrect answers bore some relation to the word, whether that be phonetic or a similarity to a root word.
Yeah I got 75k~ and did something similar ... most of the expert and grandmaster ones had at least 1 or 2 obvious incorrect answers, then it was a 50/50 so I usually went for the thing that sounded either closer to the root of the word or completely left-field
The two tests give me widely different results, probably because the sampled words aren't perfectly representative and so the results should have huge error bars to account for this sampling error.
Not a very good test. Too easy to guess many of the words, and the words seem to follow a theme. For example my list had five or six that had to do with speaking too much or too little (verbose, lugubrious, and a few others in that vein). And many easy words were placed late in the test (e.g. zeitgeist, facetious being in the expert and grand master categories?).
And it didn't even tell me at the end how many words I know!
There is a similar variant of such a test where you just go down a list of words of increasing obscurity, ticking the ones you are familiar with. If you do this once or twice, you can get a fairly good estimate of the actual number of words you know.
apparently 54,000. Seems like it is including even fictional words though in this test (like from fiction novels). Ironically I scored higher on the expert words (18/20) than the "advanced" words (11/20)
You may know more words than you think, many are shared with French
and other Romance languages, particularly the more esoteric ones (see what I did there?). Taking another recherché example: palimpsest - very similar in English, French, Greek.
Gaikwar - which I was able to guess was a former Indian state seems irrelevant as an “English” word especially given it seems to derive from a name that I have to assume is native to the region.
Vibe coders need to be forced to spend one day learning basic CSS before they're allowed to use an LLM to make a website and the internet would be a lot more pleasant as we move forward with slopification.. It doesn't have to be sloppy, and doesn't take all that much studying to at least be able to steer an llm in the right direction to make something look nice. At this point everything is just the same 3 colors and a centered flex column with weird spacing.
Fun fact: there's a test you can do called wordsum which correlates extremely highly, like .71, to IQ. It's just asking you 10 vocabulary questions. It turns out knowing advanced vocabulary correlates really well to IQ.
I don't know if I can get behind .71 implying "correlates really well" ... that's the issue I had recently with talking with GPT, it was evaluating my logical reasoning ability based on the vocabulary I was employing. You don't need fancy words to be intelligent.
I've seen other systems like this calibrate far more quickly by assigning a sort of score and confidence behind the scenes. Confidence starts out low and increases over time - correct/incorrect answers rapidly adjust score at the beginning, then things settle down.
In practice this means you get a sequence of increasingly uncommon words initially, until you get one wrong, then you drop back to something easier until you start getting things right again, and eventually circle around words at your level.
Also - too many clicks per word. It's low stakes, just let me click the definition once and I'll live if I misclick (or add an undo button).