The Marshmallow Problem & The Candle Problem

Being part of a team-building exercise or being tested…both exercises drive me bonkers.  However, a well-designed test/problem, with a convincing structure and valuable conclusions, intrigues me.  When I learned about the Marshmallow Problem, I was immediately reminded of the Candle Problem.  I learned both from TED talks.

In Tom Wuject’s presentation on the Marshmallow Problem (MP), he emphasizes the context of collaboration in solving the MP, and in Dan Pink’s illustration of the Candle Problem (CP) he focuses on the issue of intrinsic vs. extrinsic motivation.  There are similarities in solving these problems, but one particularly important condition in driving the test results may be counterintuitive.  I love counterintuitive phenomena, especially those that resonate with my own predilections.

Did the Indians have a planning meeting before building this fabulous structure?

Did the Indians have a planning meeting before building this fabulous structure?

In the Marshmallow problem, teams of four are presented with:

  • 26 sticks of spaghetti
  • 1 yard of tape
  • 1 yard of string, and
  • 1 marshmallow.  The goal is to build as tall a structure as possible, with the marshmallow sitting on top.

Typical participants go through the motions of: talking about design, jockeying for power in driving the team, devising some plans, and then, as time runs out, assembling the materials and sticking the marshmallow on top.  Most of the structures collapse under the weight of the marshmallow.

Among the professionals who have participated in such a workshop, the worst performers are the recent graduates from business school.  They lied, cheated, stole, and thought excessively highly of themselves. Hey I am just reporting the outcome here.  One of the top performers was…ready? Make a guess?  Yup, my favorite group: children.  In particular, recent graduates from kindergarten.  They did not attempt to seize any power.  They just plunged into the task.

The children usually do not plan first.  They “simply” do.  The doing produces a prototype – already with the marshmallow on top — from which they can make improvements.  This is exactly what is promoted in the Knowing-Doing Gap book (which I summarized earlier).  The best performing group is “architects and engineers;” their structure was on average 39”.  The overall average for all participants was 20”.  We should all feel reassured.  And the next best group that performed pretty well was administrative assistants.  What Mr. Wujec pointed out was that the well-performing groups utilized their specialized & facilitating skills.  I think the kindergartners performed well because they didn’t have preconceived notions of design or social behaviors.

When Wujec introduced a new element (that counterintuitive condition) into this problem, offering $10,000 as an incentive, the result was fascinating.  All groups failed!  This outcome dovetails with the Candle Problem.

In the CP (which is applicable to either individuals or teams), you are presented with the following materials:

  • a box of thumbtacks
  • a candle, and
  • a match book sitting on a table. The goal is to attach the lit candle to the wall without dripping wax onto the table.
A corner of the Getty Center, an architectural feat.

A corner of the Getty Center, an architectural feat.

In general, people take 5-10 minutes to solve this problem.  The canonical solution?  You empty out the tacks, tack the empty box to the wall, place the candle in it, and light it.  The key is the box, and once we let go of the “functional fixedness” of the box as an obligate container of tacks, the solution is obvious.

Now, Mr. Pink gleefully described a twist (that counterintuitive condition) into this experiment:  the researcher told one group to solve the problem as fast as possible and that their time would be established as a norm.  To the other group, the researchers offered a monetary reward.  The group with the offered incentive took 3½ minutes longer.  Longer.  When researchers divided participants into three groups, and offered small, medium and large rewards respectively, guess which group performed the worst?  Yes, the one with the highest incentives.

The candle problem study has been replicated over 40 years, across different cultures.  In fact, tests with variations on the incentive all have lead to poorer performance.

Another twist:  instead of placing a box of thumb tacks on the table with other items, the tacks are taken out of the box.  So, visually, you see four elements present on the table now: loose thumbtacks, empty box, candle, and matchbook.  In such an experiment, external incentives work.  Why?  Because the solution is much easier to derive.  Mr. Pink’s message:  When tasks are mechanical and repetitive, incentives and rewards work.  When even rudimentary cognitive skills are required, monetary rewards inevitably fail to optimally motivate people.  “Rewards narrow our focus,” and when the tasks require us to think broadly or out-of-the-box, narrowing our focus is detrimental. 

I highly recommend viewing these two talks; they’re a great 30-minute investment.

Why does the business world still obsess with awarding monetary incentives?  Yes, most people like money, but businesses need results that bring profits. When people are given autonomy, instead of incentives, to produce results, they inevitably perform better. So, there is a profound disconnect here. Why doesn’t the business world use results-oriented model?  I suspect it has much to do with not being able to let go of the 20th century invention called “management,” which is all about control.  This is part of what this blog has been about.

Have a good week of practicing losing some control.  Till next time,

Staying Sane and Charging Ahead.

Direct Contact:


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s