Can You Hang a Picture Straight?

weirdly accurate.

Can You Hang a Picture Straight?

Can you hang a picture straight in a crooked room?

Nearly, but on the rod and frame test a tilted surrounding pulls almost everyone off by a degree or two. In a 2010 lab version, 30 adults turning two dots level inside a frame tilted 18° were about 1.9° off on average, against 0.6° with the frame upright. This test lets you find out how far a whole crooked room moves you, and which way.

This test turns the rod-and-frame test (Witkin & Asch, 1948) into five rooms. A picture hangs 6–14° crooked on a nail; you turn it level by dragging round the nail or with the adjust bar, then hang it, and a plumb line drops from the nail to show how many degrees off you were. Room 1 is straight, a warm-up and your baseline. Rooms 2 to 5 lean 12°, 18°, 18° and 24°, two each way, and get busier: a plain lean, striped wallpaper, neighbouring pictures and furniture, then the lights down to one spot. Your score is the average error over the four crooked rooms, the pull shows which way and how far the tilt moved you, and the result places you among the 30 adults in Docherty and Bagust’s 2010 study.

It is not a balance, inner-ear or eye test. The comparison is a rough guide in both directions: your phone’s straight edges stay in view and make it easier, while a whole tilted room with stripes and furniture pulls harder than the lab’s single frame. ‘Level’ here means level with your screen, so hold the phone upright; tilting it or lying down changes the result. With only 30 people behind it, the percentage is rounded to the nearest 5 and kept between 5% and 95%.

What the test measures

One thing: how far a crooked room can bend your sense of level.

The task

  • The rod-and-frame test (Witkin & Asch, 1948), redrawn as rooms: turn a picture on its nail until it hangs level
  • Five rooms in a fixed climb: straight (a warm-up), leaning 12°, 18° with striped wallpaper, 18° with neighbouring pictures and a console table, then 24° with the lights down
  • Each picture starts 6–14° off, either way, and the lights go off between rooms so you never see the new tilt arrive
  • Drag round the nail, or use the adjust bar (0.2° a tap, hold to keep turning) or the arrow keys

Your result

  • Your average error in degrees over the four crooked rooms, with your straight-room baseline beside it
  • The pull: which way and how far, on average and room by room, the tilt moved your ‘level’
  • Roughly what share of the 30 adults in a published lab version you hung it straighter than
  • A hanging report with each picture as you hung it, a shareable card and a link that challenges a friend

Before you start

  • About two minutes, five rooms
  • Hold your phone upright: level means level with your screen
  • Works on phones and computers; sound is optional

Rod and frame test FAQ

Is this test free? Do I need to sign up to see my result?

Yes, it’s free, and there’s nothing to sign up for: no account, no email, no payment. Your result appears on this page the moment you finish, and you can share it or play again straight away.

What is the rod and frame test?

The rod and frame test is a classic experiment on how people judge upright, described by Herman Witkin and Solomon Asch in 1948. In a dark room a person sets a luminous rod to vertical while a tilted frame surrounds it, and the question is how far the frame moves the setting. This version swaps the rod for a picture on a nail and the frame for a whole leaning room.

How straight do most people hang a picture in a tilted room?

About 2° off. In Docherty and Bagust’s 2010 study, 30 healthy adults aged 20 to 60 turned two dots level inside a white frame tilted 18°, wearing video glasses in a dim room. Their average error was 1.95° with the frame tilted one way and 1.94° the other, against 0.59° with the frame upright.

How is the percentage on the rod and frame result worked out?

The percentage compares your average error over the four crooked rooms with the 30 adults in Docherty and Bagust (2010), treating their scores as a bell curve with a mean of about 1.9° and a spread of about 1.4°. It is rounded to the nearest 5% and kept between 5% and 95%, because 30 people can’t support finer claims. The straight room is left out and shown as your baseline.

Why does a tilted room make a straight picture look crooked?

The hanging report explains it after the fifth room, next to how hard each room pulled you and in which direction. Knowing the reason first would change how you turn the picture, so the page keeps it until you have hung all five. It also explains why the last room turns the lights down, and how your phone’s own edges affect the result.

Is the rod and frame test a balance or inner-ear test?

Not in this form. Lab versions use video glasses or a dark room so the tilted frame is the only thing in view; a phone in your hand can’t replace that, and this game says nothing about your balance, inner ear or eyesight. If you often feel dizzy or unsteady, talk to a doctor.

Why did I get a different result the second time?

Five pictures is a short run, so the average moves from one try to the next. Holding the phone at a slight angle, sitting or lying down and the size of the screen all shift it too. The tier names — under 1°, under 2.5°, under 5° and more — are the site’s own bands, not clinical cut-offs.

Where the numbers come from

The percentage compares your average error over the four crooked rooms (12° to 24°, 18° on average) with 30 adults who turned two dots level inside a frame tilted 18° (Docherty & Bagust 2010, Table 2), treating their scores as a bell curve: mean 1.945°, and as spread the average of the two reported SDs, about 1.41°. The tier names are the site’s own bands. The site does not publish its own players’ data.