Moving Sofa Problem Solved: Why Gerver’s Sofa Wins

Quick answer: The moving sofa problem asks for the largest area of a connected, rigid, two-dimensional shape that can pass around a right-angled corner in a hallway of width 1. Jineon Baek’s 2024 proof shows that Joseph Gerver’s sofa attains this maximum: about 2.2195 square units, a number known as the sofa constant. Baek’s 119-page proof is an arXiv preprint; as of November 2026 we found no confirmed journal acceptance.

Status checked November 21, 2026: Baek’s proof is available as a 119-page arXiv preprint. A June 2026 interview published by the Korea Institute for Advanced Study reported positive expert reactions and that peer review was not yet complete.

The challenge is not just to design a sofa that fits. It is to prove that every larger shape must get stuck. Leo Moser posed the problem in 1966; Gerver supplied the best candidate in 1992; Baek supplied the argument that no larger sofa can pass.

What Is the Sofa Constant?

The sofa constant \(\mu\) is the greatest area a connected, rigid, flat shape can have while still turning a right-angled corner in a hallway of width 1. Baek’s result identifies it with the area of Gerver’s sofa:

\[ \mu = 2.21953166\ldots \]

If the hallway has width \(w\), scaling every length by \(w\) multiplies area by \(w^2\), so the maximum is \(\mu w^2\). For a hallway 2 units wide, that is about \(2.2195 \times 4 \approx 8.878\) square units. This describes the best possible shape; area alone cannot tell you whether a particular couch will fit.

Start With Easy Sofas

Sofa shapeAreaWhy it works
Unit square1Slide it to the corner, then slide it down the other hallway
Half-disk of radius 1\(\pi/2 \approx 1.571\)Slide into the bend, rotate through 90°, then slide out
Hammersley’s sofa (1968)\(\pi/2 + 2/\pi \approx 2.2074\)Two quarter-disks with a block between them, minus a half-disk notch
Gerver’s sofa (1992)\(\approx 2.2195\)18 boundary sections (15 curved, 3 straight), derived from conditions for a locally optimal shape and motion

Hammersley’s telephone-handset shape is a clever idea: cut a semicircular notch out of the middle so the sofa can hug the inner corner while it turns. Joseph Gerver improved it in 1992 with a shape whose boundary has 18 sections. The gain is tiny: \(2.21953 – 2.20741 \approx 0.0121\), about 0.55 percent more area. A small numerical improvement needed a major proof, because the hard part was ruling out every larger possibility. Gerver suspected his sofa was optimal but could not prove it.

Upper Bounds: How Big Could a Sofa Be?

To prove a sofa is the largest, you need to show nothing bigger fits. Hammersley proved that no sofa has area larger than \(2\sqrt2 \approx 2.83\). In 2018, Yoav Kallus and Dan Romik used rigorous computer-assisted methods to lower that bound to about 2.37. That number is an upper bound, not the area of any known sofa.

ResultWhat it established
Gerver’s construction (1992)A sofa of area about 2.2195 fits: a lower bound.
Kallus–Romik (2018)No sofa exceeds about 2.37: an upper bound.
Baek (2024)No sofa exceeds Gerver’s area: the bounds meet.

Jineon Baek’s Proof

On November 29, 2024, Jineon Baek, then a postdoctoral researcher at Yonsei University in Seoul, posted “Optimality of Gerver’s Sofa” on arXiv. In 119 pages he proved that Gerver’s sofa has the largest possible area, 2.2195 and a bit, closing the gap completely.

  • The idea. After narrowing down what an optimal sofa must look like, Baek built an auxiliary quantity that bounds a sofa’s area from above. He proved that Gerver’s sofa maximizes this quantity and that, for Gerver’s sofa, the bound equals its actual area. That closes the gap between “this shape fits” and “nothing larger fits.”
  • Computers. The final proof avoids a large computer-assisted search; it uses mathematical arguments plus numerical calculations that can be done on a scientific calculator. Earlier computer experiments helped Baek develop the approach.
  • Recognition. Scientific American listed the result among the top 10 math breakthroughs of 2025.

Does This Help With Real Furniture?

Not directly. The problem is about the best possible flat shape, chosen together with its path. Choosing a shape and a collision-free path at the same time is a problem in geometric optimization and motion planning, which is why mathematicians find it interesting. It is also a lesson in how long a one-sentence question can stay open.

Common Misconceptions

“The answer is a rectangle.” No. The best shape is curved, with a notch for the inner corner.

“It was solved by computer.” Computers gave the earlier 2.37 upper bound. Baek’s proof does not rely on a large computer search.

“Every sofa with area under 2.2195 fits.” No. 2.2195 is the maximum for the best shape; many smaller shapes, like a long thin rod, will not turn the corner.

“It is about real furniture.” The sofa is rigid and two-dimensional, and the hallway is exactly one unit wide. Real couches tilt and squash.

Watch: The Moving Sofa Problem in Video

If you learn better by watching, these videos are worth your time.

The Moving Sofa Problem (Numberphile). Dan Romik explains the problem, Hammersley’s sofa and Gerver’s sofa.
The 60 Year Quest for the Perfect Sofa (Wrath of Math). A walkthrough of the history and Baek’s proof.
What’s The Largest Sofa That Can Fit Around a Corner? (Up and Atom). An animated introduction to the problem.

Check Your Understanding

Five quick questions to see how much of this has stuck. Pick an answer and press Submit; you will see a short explanation either way.

Answer each question and press Submit to see your result right away.

1. Who formally stated the moving sofa problem in 1966?
Found a bug? Let us know!
2. What is the area of a half-disk of radius 1?
Found a bug? Let us know!
3. What kind of result was Kallus and Romik’s 2.37?
Found a bug? Let us know!
4. If the hallway is 2 units wide instead of 1, how does the maximum sofa area change?
Found a bug? Let us know!
5. What did Baek prove in 2024?
Found a bug? Let us know!

Frequently Asked Questions

Has the moving sofa problem been solved?

Yes, according to Jineon Baek’s 119-page proof posted in November 2024, which shows Gerver’s sofa, with area about 2.2195, is the largest shape that fits. Expert reactions have been positive; as of November 2026 the paper is an arXiv preprint and we found no confirmed journal acceptance.

Is 2.2195 an exact value?

No, it is rounded. The sofa constant is 2.21953166 and continues; it is defined as the area of Gerver’s sofa.

Will every smaller sofa fit around the corner?

No. The sofa constant is the maximum for the best possible shape. Many shapes with smaller area, such as long thin ones, cannot turn the corner.

What is the sofa constant?

The sofa constant is the largest area a sofa can have and still turn the corner of a hallway of width 1. If Baek’s proof holds, it equals the area of Gerver’s sofa, about 2.2195.

What is Hammersley’s sofa?

A 1968 shape by John Hammersley, made of two quarter-disks and a rectangle with a half-disk notch removed, with area pi over 2 plus 2 over pi, about 2.2074.

Did a computer solve the sofa problem?

No. Kallus and Romik’s 2.37 upper bound in 2018 was computer-assisted, but Baek’s proof avoids a large computer search and needs only calculator-level numerical checks.

Keep Exploring

Sources and Further Reading

Related to This Article

What people say about "Moving Sofa Problem Solved: Why Gerver’s Sofa Wins - Effortless Math"?

No one replied yet.

Leave a Reply