GUESS WHEN AND WHERE A PHOTO WAS TAKENTest your detective skills with our daily game, check out the archive or create your own game in the Community section.
fashion from the 80-s
red double-decker = London
driving on the left side
car looks more like 90-s
old-style signs, probably the 70-s
shot on film, not digital
English-speaking country
old buildings - city centre
You think you know the answer?
How to play
Look at the photo and try to guess the location and the year when it was taken.
Place your pin on the map.
Move the slider to select the year or simply type it in.
18602024
When you’re ready, press the button to make your guess.
FAQ

A new game is available every day at 0:00 am, local time of your device!

Yes, you can head to the Archive section from the menu and play all of the WhenTaken archive games starting from the 28th of February 2024.

Scoring system

    • 5 rounds will get you a maximum of 1000 points.

    • Each round 200 points: 100 max for Location and 100 max for Year.

Scoring for guessing the year

    • For guesses within 5 years, the deduction is 1 point per year.

    • For guesses 6 to 10 years away, the deduction is 2 points per year after the first 5 years.

    • For guesses 11 to 20 years away, the deduction is 3 points per year after the first 10 years.

    • For guesses more than 20 years away, the deduction is 4 points per year after the first 20 years.

Example: The correct year is 1990. The user guesses 1974. The year difference between the guessed year (1974) and the correct year (1990) is 16 years. For the first 5 years, the deduction is 1 point per year, totalling 5 points deducted. For years 6 to 10, the deduction is 2 points per year, totalling an additional 10 points deducted (2 points per year for 5 years). For years 11 to 16, the deduction is 3 points per year, totalling an additional 18 points deducted (3 points per year for 6 years). Therefore, the total deduction is 5 + 10 + 18 = 33. The total score: 100 - 33 = 67 points.

Scoring for guessing the location

    • The points will be awarded based on distance from the correct answer.

    • For each distance bracket, the value of 1 km will be different, meaning that the farther the guess is, the fewer maximum points the player can get.

    • Brackets:

    0-100 km: min 95 points, max 100 points, decrease by 0.05 points per km

    101-500 km: min 85 points, max 95 points, decrease by 0.025 points per km

    501-1000 km: min 70 points, max 85 points, decrease by 0.03 points per km

    1001-2000 km: min 50 points, max 70 points, decrease by 0.02 points per km

    2001 - 5000 km: min 20 points, max 50 points, decrease by 0.01 points per km

    5001 - 10000 km: min 0 points, max 20 points, decrease by 0.004 points per km

    ≥ 10001 km: 0 points

    • For the guessed distance within a bracket, calculate the points by subtracting the appropriate decrement per kilometre from the max points of that bracket.

    Final formula: Location Points=Max Points for Bracket-((Distance in km-Bracket Starting Distance)xPoint Decrement per km) + 10 points for Correct Country if (Distance>500 km)

Example: The correct location is Kyiv, and the player guesses Odesa. The location points are calculated this way: the distance between Kyiv and Odesa is 475 km, which falls into the 101-500 km bracket, so the max number of points the user can get is 95. Calculations: Location Points = 95 - ((475 - 101) * 0.025) = 90 - (374 * 0.025) = 95 - 9.35 = 85.65 points. Rounding this to the nearest whole number, it would be 86 points.

We're doing our best to collect the most accurate data to provide you with the best gaming experience - we even use Street View to find the most accurate coordinates, however, sometimes we make mistakes or can't find the exact location, so bear with us. Another reason why the location might be slightly off is that we were unable to find the exact location and instead used the city coordinates.

If you find a mistake, please, report it and suggest another set of coordinates through the form by clicking on the ⚠️ button. We really appreciate the input from locals who help us make our data more precise.

Only the daily game contributes to your stats. Community games and Archive games will not affect your stats.

We use a Bayesian scoring method to calculate ratings, which ensures a fair comparison between games with different numbers of ratings. This approach prevents games with very few ratings, like 5 out of 5 (100%), from being unfairly ranked higher than games with many ratings, like 98 out of 100 (98%).

Formula:

Weighted Rating = (Rtot  ⁄  (Rtot  + M )) × P + (M  ⁄  (Rtot  + M )) × Pgeneral

Where:

Rtot = Game's total number of ratings (fires 🔥+ poops💩)

P = Game's percentage of positive ratings = [fires 🔥 / (fires 🔥+ poops💩)]

Pgeneral = Average percentage across all games (e.g. 0.70*)

M = Minimum number of ratings threshold. (e.g. 18*)

Example #1: 5 positive out of 5 total ratings

Weighted Score = [5/(5 + 18)] × 1 + [18/(5 + 18)] × 0.7 = 76.52%

Example #2: 98 positive out of 100 total ratings

Weighted Score = [100/(100 + 18)] × 0.98 + [18/(100 + 18)] × 0.7 = 95.35%

* Note that M and Pgeneral will vary based on the timeframe selected in the Filters section. M ranges from 18-50 and Pgeneral is calculated dynamically based on the average rating during the specified timeframe.

We use a normalized play count formula that adjusts the total number of plays logarithmically. This helps provide a fairer comparison between games, especially when there are outliers.

Weighted Play Count = 100 ×  (log10(Play Count)  ⁄  log10(Max Play Count))

Play Count is the play count for the game in question.

Max Play Count is the play count of the game that has the most plays of all time.

To accurately assess a game's play count, we calculate plays using the number of rounds played. (E.g. if someone plays only 3 rounds of a game then quits, that would count as 3 plays.)

Example:

Max Play Count = 10,000

Game A: 10,000 plays

Weighted Play Count = 100 × [log10(10,000)] / [log10(10,000)] = 100 × [4/4] = 100%

Game B: 5,000 plays

Weighted Play Count = 100 × [log10(5,000)] / [log10(10,000)] ≈ 100 × [3.6990/4] ≈ 92.48%

Game C: 4,500 plays

Weighted Play Count = 100 × [log10(4,500)] / [log10(10,000)] ≈ 100 × [3.6532/4] ≈ 91.33%

Note that without normalizing, Games B and C would receive scores of 50% and 45% respectively. In scenarios where a single game is considerably higher than all the rest, these scores would not be, in our minds, reflective of how popular the games are.

💌 Feedback
Something not working properly? Want to suggest a new feature? Simply love the game and want to give us kudos? Fill out the feedback form to let us know what's on your mind.
Let us know what you think 😉Submit Feedback