There's a similar game, but for guessing both the year and the location
Discussed on HN a couple of years ago too: https://news.ycombinator.com/item?id=37203511
edit: found another game like OP in the linked thread https://www.chronophoto.app/game.html
One interesting observation is how much less variation there has been in clothes for the past 20 years or so. Someone from 2005 could look completely undisguisable from someone today, by just wearing regular non-fashion forward 2005 clothes. Same goes for haircuts.
Same can't really said about someone from 1955 and 1975, 1980 - 2000, etc.
edit: Score 4695 Avg. Years Off 3.0
It's an interesting phenomenon that you can observe with other things as well like tastes in music. I think it has something to do with people having access to about sixty years of people trying out just about everything they could think off. It's all been done before at this point and it's all available in new and fresh forms. So, it's no longer about collectively picking something that is genuinely new but individuals cherry-picking whatever that they like. And it helps of course that we're not funneling media through a handful of TV channels, newspapers, etc. anymore like we used to. So people cherry pick where they get their information as well.
Probably more that’s it’s much easier to market and sell something that’s been done before than come up with something new. Also our supply chains are highly specialized for the stuff we currently make. In 3-5 years we will see bell shaped pants back again as we go from 90’s repeats to 00’s repeats which were 70’s repeats. Etc etc.
Perhaps another less depressing reason is the intersection of two things:
1) For all its negatives, online culture makes it easier to find and acknowledge different groups with different tastes that share yours, you're not as subject to having to "fit in" just with those near you. Maybe another way to say it is that fewer things are "weird" because it's easy to find others doing something similar
2) The availability of styles is not quite as bottlenecked into a limited number of taste-makers like it used to be.
For me, think it's probably more your second idea, that we're not funneling media through a handful of TV channels, newspapers, etc. anymore like we used to. This creates (for better or worse) more of a "group think" mentality, since they've all seen the same tv shows and movies, and become more in sync with their views on all things, cultural, political... So people's sense of fashion is also in sync as well as the need to fit in.
> It's all been done before
Until something new is done ;)
The sound pallette is just about infinite with what's possible.s.
I think modern music has become homogenous because true art is risky and won't pass a modern focus group.
I keep waiting for inside-out clothes to be a thing. Hipsters were getting close at one point.
back around 2002 I had some T shirts from the Gap (so, pretty mainstream) that had the visible stitching and tags look like the shirt was inside out; sort of a duplicate in that if you actually wore them inside out they still looked inside out. the only way to really tell was the little washing instructions tag on the inside.
i really liked them, was sad when they went away
They are. Lots of sweaters with seams on the outside.
I'm talking actually taking clothes not crafted to be worn inside-out.
Take your favorite pair of pants with seams and put them on inside out. Favorite button-down or polo, etc.
Is that really true? It is true for me, but I always assumed that is just because I were younger last century. In 1990 I could easily tell if someone was looking like 1980, but today I could never guess if someone looked like 2015 or 2025. I would be happy to learn that this is because fashion actually slowed down, but until proven wrong I will just assume it is because I am older and not paying attention.
> could never guess if someone looked like 2015 or 2025.
If you don't think the younger generation dress crazily then fashion didn't change. Every previous generation thought the next generations clothes were crazy.
It’s because you were younger. As a “Gen Z” person, there’s a huge difference between 2010s fashions and the fashions of today.
> I could never guess if someone looked like 2015 or 2025
It's the length of the socks
And the width of the pants
And the length of the pants vs the socks
How long are socks supposed to be now?
Apparently Gen Z prefer crew socks as opposed to millennial ankle socks.
It's easier to identify women's fashion among millennials imo:
2000's: jeans
2010's: leggings
2020's: not those
Clothing variation has a lot to do with not only fashions but fabrics.
Considered over a longer timescale, up through the late 19th century, the principle options for clothing were cotton, flax, silk, wool, leather, grasses (in some regions), and metal, with very few other options.
Viscose rayon was the first synthetic fibre created, in 1899. It was followed by Nylon (1930s) which pressaged a slew of additional synthetics, notably polyester. Adoption was somewhat slowed by various factors, including WWII, but by the 1960s synthetic fabrics and brilliant dyes were exploding into popularity.
Flagrant use of synthetics faded somewhat through the 1970s and 1980s, with fabric blends and natural fabrics becoming more prevalent (yes, I'm aware that the leisure suit was prominent in the 1970s, but it was less so by the end of the decade).
There've been variations in specific styles since, though most to my eye have been evocative of earlier 20th-century periods since the 1990s, rather than distinctively original. (I'm far from an expert in clothing fashion, take with heaping handfuls of salt.)
Other scene-setting elements include architectural styles (notably houses in the West), automobiles, and since ~2000, the presence and style of handheld mobile devices, smartphones after about 2010. Camera styles would be a useful indicator for much of the 1900s ("Brownie" box cameras, SLRs, "Instamatics", and the like).
Well there's very little "common" cultural influence these days. It used to be that you could see a significant episode of a TV show and EVERYONE was talking about it the next day in the office. Now you're lucky if you can even find one person that has seen the show you're even talking about, let alone if they are "caught up" on it since many people prefer to buffer a few episodes while they watch other things.
Agreed, I think the last of the shows where everyone talked about was Game of Thrones.
I would say athleisure is a recent development. 10 years ago you'd see far fewer women wearing tight leggings outside the gym. Heck, even in gyms you'd be hard pressed to see the kind of tank tops and sport bras you regularly see nowadays [1]
N.B. I'm just noting the phenomenon, not commenting on its merits
[1] https://www.vice.com/en/article/women-are-intimidating-men-w...
I’ve never seen women wearing that kind of outfits in gyms in Norway or France, but I see it all the time online. Is it an American thing?
Can't comment on the USA, but I'm seeing a fair number of them in Canada and (somewhat less commonly) in Croatia.
Not uncommon in Norway, at least at few gyms I've been to after they reopened after COVID.
This cultural phenomenon is sometimes called "the Long 90s"
>edit: Score 4695 Avg. Years Off 3.0
How is that even possible? Only three years off? I thought guessing it which tens would be good enough. Getting the last digit right is impossible without blind luck? Or am I missing something?
I got 3.2. Spoilers below.
Round 1 I just guessed in the middle of the 70s.
Round 2 I was 5 years off, but I should have been able to get closer if I'd slowed down. Someone with a better knowledge of modern Middle Eastern uprisings or Arabic could have read the signs, but also the tech in everyone's hands was pretty closely dated.
Round 3 was pretty clearly the London Blitz—the uniform was WW2 era and people were piled on top of each other in a London Tube station—which narrowed it down to basically a single year.
Round 4 was another guess-the-decade shot, I was 6 years off (I guessed 1940). Had I thought about it harder it would have been unlikely to have been taken during WW2, which would have bumped my guess up a few years to be closer.
Round 5 was a bunch of protesters who were pretty clearly in the mid-to-late 2000s (cell phone in the background), and the topic of their protest was gay marriage which had a single very important flashpoint in that decade with Prop 8 in 2008.
Round 1 was the same.
Round 2 I guessed it was something to do with Arab Spring but without looking up the year of it happened I remember it was late 00s or early 10s so I put 2010.
Round 3
>and people were piled on top of each other in a London Tube station—which narrowed it down to basically a single year.
Not sure that is a knowledge I know but I only guess it was WWII, again without looking it up I only remember it as the 40s.
Round 4 I had the same thought so I guess something like 1960s because I thought it wasn't WWII.
Round 5, not from US so I only know / heard of Prop 8 when Mozilla fired their CTO Brendan Eich. I thought it happened in early to mid 2010s.
I averaged 5 years off, but my average would have been much better without one photo that I botched by 16 years because I'm unfamiliar with early 19th century swimsuit styles.
There were 2 major world events represented, that helps. Clothing style helps.
If you guessed the 19th century, you were off by a bit more than 16 years.
And here I complain about people erroneously specifying EST all summer along. They’re only off by 1 hour.
I skip specifying Standard or Daylight time altogether, and just use ET (or CT, MT, etc). Much easier on my brain, and nobody actually cares which they're in at the moment.
4586 Avg. years off 5
I felt like I should have been able to be closer. The images wouldn't zoom on my phone though. I'm impressed by three years off but would consider it quite possible for somone who knows their history.
Score 4722 Avg. Years Off 2.6 here. Spoilers: I got two exactly right (Arab Spring, Battle of Britain) and one off by a year (dancing). The two exacts were pretty obviously tied to specific dated events, and I both love history and have a pretty easy time remembering dates. I love this sort of game, as I do trivia, history trivia especially.
I got one off by 20 years, oops. But there's two where I think you can get exact or off by 1 with some knowledge of world events, maybe it's a 50/50 shot. And there's two where the half-decade is pretty clear.
For some images, it involves some luck. But one image was WWII for example, that one is much easier to get accurately.
> Score: 4947; Avg. Years Off: 1.2
Skill issue.
It's called "stuck culture". https://www.youtube.com/watch?v=Pyi-KGwuR9c
I found it equally interesting how much you can identify a photo's time period by the film qualities.
Lindy.
The "global score distribution" (and my score) and the end was pretty good idea. Much more informative than just some high score list.
I wish it had distributions for each picture
Agree – great feature & nice for sharing.
That was fun and definitely a bit challenging to figure out—I scored 3677, landing in the top 46%, with an average of 9.2 years.
As for some (hopefully constructive) feedback: I think the year selection slider could benefit from a few adjustments. If an image is from the distant past—say, pre-1950s—it might make more sense to use decade-based precision rather than individual years. For example, a black-and-white photo had me way off that caused me to lose 20 years of my life :). It would be helpful if such images were categorized more broadly, something like “191x”.
For mid-century images (roughly 1960–1990), a 5-year interval could strike a better balance. I came across one from that era and was six years off. And for more recent images, say post-2010, a one-year precision feels reasonable.
Of course, this suggestion mainly applies if we're deducing the year based on visual cues like clothing, hairstyles, and the environment. If the game is intended more as a quiz based on the text descriptions under the images, then it shifts more into trivia territory—and that’s a different type of challenge altogether.
The score takes the magnitude of error into account. Isn't it nice that guessing the precise year is more rewarded than guessing the decade?
Hi, I made this game. Really appreciate the feedback! Seems like older images tend to be frustrating, so I think I'll use them less going forward.
Also, older photos are scored more generously, so being a few decades off won’t tank your score nearly as much as it would on a newer one.
No, don't stop using the older images! They were the most fun for me! Reasoning about the political landscape that would have led to the beach photo was challenging and rewarding when I got it right. And the London Blitz photo was also a ton of fun to see and think about.
Adding to the chorus: I like the older images and the precise year is important! The underground shelter is something that wouldn’t exist just a few years after that photo, or before.
> The underground shelter is something that wouldn’t exist just a few years after that photo, or before.
It’s probably there now, it’s the London Underground.
I disagree with all of that feedback. Might be because I got top 3%. I don't mean this in a mean way, but it sounds like sour grapes to me!
For the record, I especially enjoyed the older photos. Maybe it's just a matter of making it obvious in the UI that they penalize less, so that score maximizers won't be annoyed with them.
Yes! Games are to be fun, but for many the challenge is fun.
This is fun. One tiny bit of UI feedback - when maximising the image, I'd expect the escape key to go back to the game, and found it a bit frustrating when it didn't.
Hi, I think you should credit the photographer of these as well.
I got that London blitz photo bang on. The older photos are fun.
Great idea! The trick is to zoom into peoples clothing, and then the associated event becomes quite obvious. Eg. Photo of Disco, no crazy lapels, obviously towards the end of disco, which puts you within a year or two.
Might have gotten lucky but zooming in on clothing got me to 1.8 years off on average (top 2%).
Pinch to zoom didn't work for me on Firefox for Android, and neither did double tapping.
Settings > Accessibility > Zoom on all websites
Thanks!!!
Long press the picture, open image in new tab. Then you can zoom.
Yeah Firefox has trouble with a decent fraction of web apps nowadays.
Not their fault of course, with people not testing with Firefox.
It doesn't work for me either (Android+Chrome)
Me either. Squinting at it didn't help.
I just went by feel of each and got 2.4 year average, also top 2%. I guess it doesn't do higher than top 2% then.
> obviously towards the end of disco
What is the end of disco?
Disco is probably the only genre whose "death" (as in very quick drop from being mainstream) can be traced to a very specific date: July 12th, 1979.
Interesting. Can we organize an event to blow up reggaeton records (or trap, or latin urban, or whatever they call it now - sounds all similar to me) and hope for it to go away?
Personally, I'd prefer for a young white male demographic to not riot against music primarily associated with certain marginalised communities anymore.
The person you’re responding to is an autistic wheelchair-bound lesbian woman living in Ghana.
What now?
Personally, I’d prefer internet commenters argue in good faith instead of emotionally parroting 2010s social justice identity jargon. You could have responded with something op was missing about said music (interesting history, examples of artists innovating there, etc) instead and added some value to the thread.
What a sane and normal response! Did I not already share some interesting history? Aren't you directly replying to that?
But okay, here's another one: that same queer and black community from Chicago that definitely didn't feel great about ~50k white men burning their records developed a new (now super popular) genre on top of the ashes of disco (both metaphorically and literallly).
Its name is a reference to a nightclub in Chicago called The Warehouse (https://en.wikipedia.org/wiki/Warehouse_(nightclub)), which was frequented primarily by gay black men in the early 1980s. People all over Chicago were looking for the type of music that was playing in the Warehouse, hence the name: 'house music. So, not only was that day in Chicago fundamental to the death of disco, but it was also fundamental to the birth of house music. The more you know!
I'd be suprised if an autistic wheelchair-bound lesbian woman living in Ghana would either confuse reggaeton with trap or be overly bothered by a local excess of either sufficient to want to detonate and destroy any physical media given much of it would be immortalised as digital patterns in the cloud.
I can respond with music from women in
* Nigeria: https://www.youtube.com/watch?v=tvY31eN3gtE
* Zambia: https://www.youtube.com/watch?v=H2lvgKDpiSA
* Australia: https://www.youtube.com/watch?v=JLQ4by3lUJo
* The Gambia: https://www.youtube.com/watch?v=PtmmlOQnTXM
and suggest there's some interesting cross cultural threading in all these examples.
I'm from Spain and here that music is pretty much everywhere you go. Not associated with any specific community at all.
I'm not sure why that would mean it doesn't go higher than top 2%, it would only imply the scores you ended up with were in the same percentage bracket.
Confirming this, it looks like you indeed get top 0% when you end on a perfect score of 5000 with an average of 0 years off.
I presume OP meant towards the end of the disco era.
Tried plugging it into the new o3 model out of curiosity and it got 0.2, i.e. it got everything exactly right save for one that it missed by a year.
You know, there are easier ways to cheat at the game? The photos are from the internet, a reverse image search will surface all of them.
Why would s/he come here to boast about cheating on a game? S/he was obviously testing the AI.
I read it as more of an experiment to gauge the capabilities of AI than an attempt at cheating.
I'm impressed that AI is that good.
Is it good, or are those photos already in the training data that the model was trained on?
Would be interesting to see how well it would fare on five photos from different years, using pictures that have never previously been uploaded anywhere.
I would bet it would still do well. You can give a frontier LLM photos that it has never seen before and it will reason about them incredibly accurately, since is is trained on, well, a lot of other similar photos. You can easily try it yourself, give o3 a photo you have taken yourself of a random place, it will crop, zoom and analyze different sections of the photo until it gets most of the time the exact coordinates of where the photo was taken.
> it will crop, zoom and analyze different sections of the photo until it gets most of the time the exact coordinates of where the photo was taken.
I will give it a try myself too. But all my photos that I take with my iPhone have date and GPS location embedded in the metadata so I will first have to transfer the pictures to my computer and strip them of metadata.
You can just take a screenshot of them. Not the cleanest way to do it but probably the fastest if you don't already have special tools on your phone.
> are those photos already in the training data that the model was trained on?
The photos are of course in the training data. Most humans too, would have seen a large fraction of these photos before.
But associating it with a year is where the magic happens.
You are dismissing a technology capability that was science fiction a year or two ago.
4580. Helps if you recognise the current affairs. Gay marriage, Arab spring, London Blitz came up. Leaving me with a party with a mix of drab and 70s disco clothing so I guessed 1981 and was close enough!
The one that got me was some people at a beach in black and white.
I recognized the Arab spring and the London blitz.
The gay marriage one I couldn't put a date, but noticed a flip phone on the picture so I just guessed the highest number before iPhone release year and wasn't too bad.
The disco scene was clearly a 70s Pic so I just guessed smack in the middle of the decade.
The beach black and white I was the most off. I faintly remember that it was a big discussion point during the 20s that women should be allowed to wear "revealing" swimsuits,so again I went for a year in the decade. I was off
Score 4753 Top 2% Avg. Years Off. 3.0
I didn't really recognize any of the current events - or anything at all - I just ... kind of knew ...
In fact, it had less to do with the actual content of the photo than the qualities of the photo itself ... I sort of felt like photos looked like that around 1990 ... and I was right.
A fun (hard) addition to that would be the color photography by Sergey Prokusin-Gorsky because the picture quality is so out of its time (~1900) https://en.m.wikipedia.org/wiki/Sergey_Prokudin-Gorsky#Galle...
Oof what! They are like a modern phone camera almost. Much better than 80s film cameras. A lot of work must have gone into that.
I love it. Just too bad you can't zoom in the picture on mobile. Had to open the picture on a new tab.
I could zoom in okay on Chrome for Android but I always enable zoom override in accessibility so it might be that.
Usability on mobile Firefox was not great.
In Firefox, you can now go into Accessibility options and enable zooming on all sites which should help, it did for me.
This is great to know, thanks! Will hopefully have it working for tomorrow's challenge.
Zooming in worked for me in Arc on iPhone
I pinched the page and it zoomed ok
(I used Brave on Android)
One thing I realized was: the presence or absence of smartphones is an easy way to determine whether something happened before circa 2009 or after. Another: indoor smoking.
WhichYear 4/20/25 4311 pts 5⃣ avg. years off
6⃣ 1⃣ 5⃣ 2⃣
Wordle opens (or used to) a mock modal dialog with a copy button for sharing the results. I feel like opening the modal rather than requiring clicking the 'share' button probably leads to a lot more shares?
I think I'd save a longer and just put "WhichYr.com" as the top line, bump the date and score to the next line.
I've got 4861 (top 2% ~3 years off)
If someone likes additional challenge then Teuteuf Games' WhenTaken has both guessing the year and location: https://teuteuf.fr/
They also got Worldle which was covered on hn - https://news.ycombinator.com/item?id=30367906
Yes, the extra challenge of having to guess the place really adds some fun in WhenTaken. Direct link to it: https://whentaken.com/game
You know, this game pretty much demonstrates how fast cell phone addiction happened. You can tell the year by how many people are staring at their phones, and how big their phones are. You don't see nearly as many people reading newspapers in old photos as there are people gormlessly staring into their phones in contemporary photos.
Thanks, got very lucky in the sense that they were recognizable events mostly, except the ladies on the beach, so was 1.2 years off.
Agreed. The beach goers threw me too. (Spoiler) I was off by 20 years.
Likewise, 29 years out for that, everything else within a couple of years.
Probably didn't help that I'm not American so the cultural references aren't there.
As someone with interest in fashion and social politics, I got that photo year within 2 years difference. Many photos here are largely identifiable as political events but certainly helps to have a fashion background as others have commented in this thread.
I have a tendency to date films and photos by their quality. Thats why I was 16 years off on the 1934 one. I got the first one exactly right, even though Im born in 1990.Some B&W cameras 80 years ago took better pictures than most cameras today.
As an aside, this is a great website to browse old B&W photos. Some of these scans are really breathtaking. https://www.shorpy.com/
I wanted to play more but couldn't find any way. Refresh page, no new game link, tried entering URL again. Checked but no cookies to delete. That's one way to ensure I never visit the site again.
Hi, I made the game. Appreciate the feedback! Working on a mode that lets you play more right now.
Sorry for being so hard on it. I looked at the intro page again and see now that it's not what I was thinking, where you have a large database of good photos with known dates it can choose from, instead each day your curate (I assume) a few photos for everyone to try and compare results (for the same photos). So it's not just a matter of letting someone play again, because currently they would get the same five photos. I took it to be something like the classic "Am I hot or not?" with a database of tons of photos.
To be fair it's a daily challenge. Try again tomorrow. But yes, I would love it if there were more game modes (similar to GeoGuessr)
I like things I can improve my skill on. This works by getting a stream of feedback as I study what I did and the errors I made, then try again with changes. Having only 5 rounds a day (assuming it resets each day) greatly limits this. This limitation seems more like an attempt to make me use it in a certain way (make it part of my daily routine). That's off-putting. I wanted to sit there grinding over dozens of photos and watching my skill improve in real-time.
4878, avg 2.4 years off. Way better than I thought I'd do, I think I just got lucky. I agree that it's a lot like chronophoto.app, but that also has freeplay. This has a nicer looking UI but I think a worse feature set. In particular the scroll-to-zoom and freeplay on chronophoto are nice to have.
Edit: I had seen the photo in the subway before, or something very similar. It's an interesting decision to use photos from big events that people might recognize. (Not inherently bad IMO, just not what I would have intuitively chosen)
Thanks! This is all super helpful. I’m hoping to have something like Chronophoto’s zoom working by tomorrow’s game, and a freeplay-style mode is in the works too.
Do you think it's more satisfying to guess based on clues like tech, clothing, and photo quality, or when you recognize the historical event or era (assuming it’s something a little less recognizable than that subway photo)?
I think both are good, it's nice to have variety. It was very satisfying as a user to think that the photo of all the people might be of the Egyptian revolution in 2011, then end up getting the year exactly right. Some photos are inherently gonna be easier to tell than others. Wartime photos can have a fun game of "figure out which war it is based on the equipment". I think the problem with the subway photo is that it's a very memorable situation, to use subway tunnels as shelters. But there are plenty of wartime photos that aren't gonna be that way.
I also like trying to figure out where the photo was taken, even if that isn't strictly part of the game. So I'm glad that the photo caption is there.
This is exactly the same game as chronophoto.app, as far as I can tell, except with way better UX and as a “daily challenge” type game. Well done, anyway! It’s a fun game.
Here's another one for guessing ethnicities https://hbd.gg/play/
Neat thing! I'm gonna add it to my round of daily games.
I'd suggest changing "Share results" to some variation of "Copy results". Share looks like it's gonna open some sharing menu and deter people from playing (and hence promoting your game)
WhichYear 4/20/25 4142 pts (top 22%) 7⃣ avg. years off
4⃣ 6⃣ 1⃣ 2⃣
When I press "All Times Stats", I see "Games Played: 2". But I only played today. Off-by-one error?
I love this idea so much and would make this one of my daily games.
FWIW: The jankiness of the slider feels quite jarring, given it’s the sole interactive element. It makes the thing difficult to stick with. On my iPhone 12, it lags way behind and the numbers update at like 10fps.
I really hope this keeps getting daily images added.
Also laggy on my iPad Air.
Nice game! Good UI experience. When sharing you get 4 different scores (points, top %, avg error, total error) I think. This is a little confusing, better to share one score (maybe plus total error, maybe) so people can more easily compete. 4 scores is too confusing.
Very fun, but I'm horrible at this
WhichYear 4/20/25 3045 pts 13 avg. years off
4⃣
This is interesting concept, if there was more variety it could be some like timeguessr.
Looks like I also have to play "guess the timezone of the creator of this game".
I enjoyed playing it yesterday. It's now 11am today and it's still not open again. I have no idea when it will be.
Hey thanks for playing! The game resets at midnight ET. I'm working on adding a countdown to the next challenge right now. Apologies for the lack of clarity on that
Hey, small bug to report: When you share your stats, the date seems to be based off local time, not the day of the test. So like here in PT, it's 4/23, but I just did tomorrow's quiz (because it releases at 9pm PT = midnight ET); the share text should say 4/24 but it still says 4/23. Or put another way, I shared two different results with my group chat today that both said 4/23.
You don't need to apologise. You created something cool and my family enjoyed playing it.
I love this!
Small bit of feedback that – after expanding an image – it'd be nice to more easily collapse it and return to the game screen. Maybe by pressing escape on the keyboard and/or clicking on the image.
Fun, though I'm apparently terrible at this -- avg 14 yr off.
UI note: the statistics tab shows games played: 2 when I only played once, and the lifetime points appears to be double my one game.
Definitely the most polished game of this sort that I’ve seen. Nice work!
5.2 years average off. The first was a dance club in the 70s, I guessed 78, but it was 79... There was one where nudists where taking off their clothes, I guessed 50s but it was 1934!!!
Ya, I guessed the 50s too! So early!
3121 pts, 14 avg. years off.
Decent try considering most of the options were USA specific.
What do the last emoji numbers mean when you share your results?
Below my points, and how many avg years off, I got a dartboard, then "6", "3", emoji of a blind man walking, then "6".
I think those indicate how far off that guess was. I have a dartboard when I guessed the year exactly right, numbers 1 and 3 matching these years off, and a blind man for >10 years off or around there.
Yes, I confirmed it with the results from other friends. Very unintuitive, though. Just saying "Accuracy:" before the emojis, would give a good enough hint of what those mean.
Using the blind man with probing cane emoji to signify losing in a game is really disappointing
4920 here. Waiting for contenders ;)
More seriously I often wonder how this works in the brain when there is no obvious historical clue. Is it akin to guessing a numerical quantity (like when seeing a crowd) ?
If you grab statistics per photo rather than the whole thing, and maybe do rough client geolocation, you can turn this into a sociological study and publish a paper about it…
Fun game.
Suggestions:
Make the logo clickable so it goes to the homepage
On the sharing stats page, add a link to start the game, e.g. if I send the link to someone to challenge my score, there needs to be an easy way to start the game.
Guessing the year each photo was taken can be surprisingly accurate when you observe changes in technology, fashion, or even product developments visible in the background.
I don't understand this site's "come back tomorrow" part. Wouldn't it be better to let people play as long as they want?!
That was fun. I got a score of 4174, and an average of 6.2 years off. Clothes and hair styles were quite helpful in 3 of the pictures.
I'll definitely take a look at it tomorrow.
I guess I'm getting a bit paranoid, but whenever I see a game like this, my first suspicion is that it's just a trainset annotation UI in disguise.
Nah seems like pictures choosen for this purpose? Or do you mean they will sneak in a picture with an unknown date every now and then to good estimators?
Something like this, pretty much.
I doubt why they need to, these photos are all from newspaper which already have the date.
Eh, we already had a discussion a few days ago about people using AI to effortlessly geolocate photos. That's the "where". I could see a model that also provides the "when" as useful for a lot of purposes.
I created a similar game a few years ago for guessing the year of an object in the Met’s collection. Would love any feedback if people want to try it.
I tried playing again, but how?! I am presented with the last play stats, but see no way to play again.
I can share, or get all time stats. That's all.
This is a great way to gather labeled training data for a neural network that can guess in which year a photo was taken.
A photo archive with actual correct date labels is even better (the thing which this game is built on.)
ChatGPT 4o guessed perfectly 8 in a row!
Score: 4246 pts 6 avg. years off
I am not american or native english speaker or never have been outside Asia
Nice idea. I was 6 years off on average. Probably something the new AI models might be very good at.
7.2 years off on average. First 3 were 3, 1, and 1 year off respectively. Last two blew my momentum.
I love this stuff. I was always good at guessing song, movie or photo years.
Very clever idea! This was fun!
I did OK, didn't try to hard but was surprised at how good people are at this!
This is really fun. My score was 4307 (top 12%) with average 5.8 years off.
Man I am in love. The global leaderboard is inspired!
4902, 2% off. I like how you've created a guess distribution
This is fun! Thanks for the post.
I got a score of 3740 and an average of 10 years off.
Styles haven’t changed that much since jeans and a tshirt
ripoff of https://www.chronophoto.app ?
it's really fun/simple so can't feel too bad
I feel dumb, I can't figure out how to play again
I created a similar game a couple years ago for guessing the year of an object in the Met’s collection. Would love any feedback if people want to try it. https://davidfdriscoll.github.io/met-chronoguesser/
Peaked bell-curve: normal
No ad while using DuckDuckGo/iOS.
Very clean.
WhichYear 4/20/25 4939 pts (top 1%) 1⃣ avg. years off
2⃣ 1⃣ 1⃣ 1⃣ 1⃣
whichyr.com
Tried this, it is very addictive game, good job team !
WhichYear 4/20/25 4215 pts (top 17%) 6⃣ avg. years off
3⃣ 1⃣ 8⃣
whichyr.com
o3 got a SCORE AVG.4989 YEARS OFF 0.2 TOP 0%
Does it have only one set of photos per day?
would be so fun if it has a multiplayer mode
tried this game, it is very addictive, good job team !
Now is there anything which will take this, run it through ML and then find a date for all of my scanned family photos?
This was fun!
This was fun
Score
3433
Avg. Years Off
12.2
Score
4266
Avg. Years Off
5.2
geoguessr clones/copies are so hot rn.
Cool idea, m8. I ended up just looking at clothing and fashion choices. Although black and white photos do tend to represent historical era (ie, holocaust, ww2)
[dead]
Another for year + location https://whentaken.com/
Holy, I was off by 7 years and 427km on my first guess and I don't want to play anymore because I've surely peaked
I just got one that was off by 2 years and 24.2 meters, but it was kinda cheaty because it showed Barbara Bush on the South lawn of the White House.
Seems like they're "season photos", so a lot of Easter and Holy Week celebrations today.
I have to say I got almost perfect a Holy Week celebration in Sevilla in the 1920's (I'm Spanish, so only some hundred meters away but kinda wild guess for the year and only was two years off), and pointed Mexico instead of Guatemala for another, but nailed the year (1981) for a grand total of 902 out of 1000 points.
That was fun!
Ha, I just got that same one! I managed to guess the year (1/4 chance), and only 15.5 meters off.
lol nicely done
I only immigrated to the US in my adult life and I can't say I recognize Barbara Bush at a glance, so I was _also_ off by a few years and a few km on that one :( but still a lot of fun!
Oh, my first guess was off by only 1 year, and my final score was 3742, and my Avg. Years Off was 8.4, so it seems I have a pretty good historical memory, hahahaha.
The whole collection is great:
https://whentaken.com/teuteuf-games
Where do these sites source the photos from?
This is a nice idea. Only issue are the ads - the ad on the bottom right is bleeding over the photo, there are two video ads playing simultaneously... Makes it annoying to play
I had to close it - the ads are all over the place.
https://pointguessr.com is another one, it's like timeguessr but with real-time co-op functionality