Skip to the text

The Rope Can’t Tell

The Ringelmann effect

Put more people on a job and each one tends to do less.

Max Ringelmann measured it with a rope in the 1880s: 85.3 kg a man alone, 61.4 in a group of fourteen. The rope only reports the total.

Pull four times here and see what company does to you.

Scroll to haul After Ilya Repin, Barge Haulers on the Volga, 1870 to 1873.

Grand-Jouan, France, 1882.

Max Ringelmann is testing engines.

Horses. Oxen. Machines. Men.

He hardly draws a line between them.

His test is a rope, five meters long.

One end runs to a recording dynamometer.

The other end is yours.

You’re the student.

His students averaged 85.3 kilograms.

There were 28 of them, each pulling alone.

Now he adds men.

If effort simply added up, eight would pull 8.00.

Two men: 1.86

Three men: 2.55

Four men: 3.08

Five men: 3.50

Six men: 3.78

Seven men: 3.92

Eight men: 3.92

The eighth man added nothing.

That table has been quoted for about a century.

He never said which task it came from.

A German summary in 1927 said rope, and the rope stuck.

His measured rope was kinder.

Alone: 85.3 kg a man.

In sevens: 65.0.

In a fourteen: 61.4.

Nine.

Ten.

Eleven is how many Ilya Repin painted.

Barge Haulers on the Volga, 1870 to 1873.

A critic stood in front of it in 1873.

“A living haulage machine,” he wrote. Vladimir Stasov.

Then he looked closer.

“Dissembling a little and contriving to carry less.”

The tall soldier, puffing his pipe.

This one is filling his pipe as he walks,

his shoulder resting on the man beside him.

You can see them.

The rope can’t.

A dynamometer reads one number.

It has no idea who is in it.

Take your place.

Would anyone have noticed you let go?

The rope can’t tell.

Ringelmann blamed timing:

“the lack of simultaneity of their efforts.” Ringelmann, 1913.

Every man peaks, just not together.

Crews already knew the fix.

They sang.

But he’d seen something a song can’t fix.

Prisoners turning a flour mill:

“each man, trusting in his neighbor to furnish the desired effort, contented himself by merely following the movement of the crank.” Ringelmann, 1913, translated by Kravitz and Martin.

In 1974, a team in Massachusetts took the timing away.

They blindfolded each puller

and said other men were on the rope behind him.

They were faking it.

He was pulling alone, and his pull still dropped.

Ten men are behind you.

They’re still back there.

Now they let go.

It’s just you.

Take off the blindfold.

There was never anyone behind you.

They’re paint.

Every reading in this room came from your two thumbs.

You skipped at least one pull, so this rope has no reading for you.

  • Alone: no pull
  • With the crew: no pull
  • Blindfolded, “ten men” behind: no pull
  • Blindfolded, alone: no pull

Scroll back and take all four, or carry on. The research doesn’t need your thumbs.

A low number doesn’t mean a lazy man.

Stasov’s next hauler is wiping his brow,

“terribly ill and exhausted.”

The needle can’t say why.

What changes it is being heard.

In 1981, blindfolded people shouted less

when they thought others were shouting too.

Given their own microphones, they shouted just as loud in a group as alone. Williams, Harkins and Latané.

Give every man his own dial.

Now the rope can tell.

Far off, Repin left a steamer.

The rope’s replacement, trailing smoke.

New crews are on new ropes now.

The Ringelmann effect

Put more people on a job and each one tends to do less.

Max Ringelmann measured it with a rope in the 1880s: 85.3 kg a man alone, 61.4 in a group of fourteen. The rope only reports the total.

Pull four times here and see what company does to you.

or scroll down

After Ilya Repin, Barge Haulers on the Volga, 1870 to 1873

0 of 4 pulls recorded.

The Ringelmann effect

Add hands, and each hand does less.

The total usually still goes up. What falls is the share each person delivers, and the rope reports only the total.

1913

the year he finally published

A lost report

Max Ringelmann ran these tests at the agricultural school of Grand-Jouan, mostly between 1882 and 1887. Comparing people with groups was a side interest. He wanted to know which way of pushing or pulling got the most from a worker.

For decades psychologists knew him only from a small figure in a 1927 German paper, and some assumed he was a German psychologist working in the 1920s. David Kravitz and Barbara Martin tracked down the original in 1986 and found the famous table had no stated task behind it.

2

different ways to lose force

Timing, and trying

Ringelmann put the shortfall down to coordination: people don’t peak at the same moment. Ingham’s team showed in 1974 that the pull drops even with nobody to coordinate with. In 1979 Bibb Latané, Kipling Williams and Stephen Harkins named that second loss social loafing.

A 1993 review of 78 studies by Steven Karau and Williams found loafing across many tasks and groups of people. It shrinks when people think their work can be evaluated and when the task means something to them.

580,000

commits across 58 open source projects

On software teams

Ingo Scholtes and colleagues measured how much code each commit actually changed. Output per developer fell as teams grew. An earlier study that counted commits had found the opposite, so what you measure decides what you see.

30

debating agents, no more diverse than one

On agent teams

In a 2026 preprint, thirty language-model agents in dense debate gave no more answer diversity than a single agent on a hard benchmark. The authors call the link to human loafing an analogy, not a claim that models slack off. Mixing different models helped.

What helps

  • Keep time.A shared rhythm recovers what bad timing loses.
  • Sign the work.People pull harder when their part can be seen.
  • Keep the crew small.Every extra person adds less than the one before.
  • Ask why first.A low reading might be a sick man. And a number people are paid on stops being a good number.