Jay Barney describes a coin flip exercise to make the point that innovation might be modeled as an outcome of pure luck. If so, how can firms manage such processes? The exercise is simple:
- Distribute coins to the class and have them flip.
- Those who flip “heads” remain standing, “tails” sit down (unless everyone gets a tail – then they remain standing)
- Repeat until one person is standing & pass all coins to him/her
Discussion focuses on several key points (Russ Coff’s slides emphasize real options):
- What capabilities/skills did the winner have? Make a big show of trying to find out how the winner did it (it’s all in the wrist, etc.). Often the winner will have flipped 5 in a row or more (a 3% probability?). People will laugh since they know it’s luck.
- Is it possible that innovative companies are just lucky? We don’t see a lot of repeat innovators and, if it is luck, even these might be explained.
- Selection bias is a problem if we try and draw conclusions by only looking at winners. In a population (like the class), the probability that someone will flip 5 in a row is rather high. We can only identify causality if we study the whole population.
- If it is luck, how should one manage investment? This is a nice lead in for portfolios of strategic investments/real options or superior expectations/forecasting.
Contributed by Jay Barney
The attached PowerPoint file contains a list of what to prepare before class (slide 1) and the slides for class, including discussion and wrap up slides. It is best to go over the directions in detail in class as, unfortunately, students often do not read the directions very carefully; the verbal overview also gets them thinking about setting up a divisional vs. a functional structure for the task. It really is worth stressing the fact that they need to set up a structure. Choose an external quality control group at the beginning of class — Purposefully pick students who are quick and pay attention to detail for this task, as it will have to be completed in a short period of time.Overall, what tends to happen is that both groups improve from trial 1 to trial 2, however the functional group improves by a much greater amount and generally has fewer QC errors (i.e. words used repeatedly). 
Reprinted here with permission of author Jeffrey Barach along with my PointPoint slides I use to administer the case.
Several factors might account for high prices including:
Here are 
The essence of this exercise is simple. Teams must build a device that will catch an egg dropped from 25 feet (e.g., a stairwell). The trick is that they must build it from items purchased in an auction. As such, items that are easy to use (e.g., an old pillow) are very expensive while items that are hard to imagine a use for (a brick) are cheap. Profit is determined by 1) succeeding in the task of catching an egg and 2) having the lowest cost function. As such, this demonstrates Barney’s (1986) notion of superior expectations in strategic factor markets.
The key points are that managers deal with complex problems, and often use cognitive frameworks to help constrain the problems. These frameworks tell the managers what information is needed; however, unless the managers understand that all frameworks have limitations, these frameworks can fall trap to cognitive biases.
Carpenter Strategy Toolbox 