Question: what’s your algorithm aversion? If you’re not sure what this means, here’s a use case that’s happened to all of us: you’re looking for a good place to eat, say sushi or thai, and you read some good reviews about a particular restaurant and even get a glowing recommendation from someone you respect. But, much to your surprise, the dish that you eat isn’t that good. Do you go back again or do you avoid the restaurant forever?
Based on that one sample, there’s a good chance that people wouldn’t visit the same restaurant twice. It gets worse when talking about algorithms, according to a recent article in the Harvard Business Review:
“In a paper published last year, Berkeley Dietvorst, Joseph Simmons, and Cade Massey of Wharton
found that people are even less trusting of algorithms if they’ve seen them fail, even a little.
And they’re harder on algorithms in this way than they are on other people. To err is human,
but when an algorithm makes a mistake we’re not likely to trust it again.”
In a way, we didn’t need a study to tell us this as we all rely on our gut feelings – we see this play out in the investment market all the time. The average investor has underperformed in comparison to the stock market usually because emotion plays a bigger role than facts, thinking that “I can perform better than the forecast.” The study highlights this thinking directly, with participants less likely to choose forecasts made by an algorithm than by a human, even when the algorithm was shown to perform better. As the study states, “This is because people more quickly lose confidence in algorithmic [sic] than human forecasters after seeing them make the same mistake.”
Data rising to the top
In the investment market, more success is seen by those that mix actual data insights into the decision process, in fact the term “data-driven decision” has become very popular in the last few years. According to a study of top performing organizations by Rasmus Wegener and Velu Sinha from The Bain Company:
• 53% of top performers make data-driven decisions ‘very frequently’ vs. 28% of everyone else
• 41% make decisions ‘much faster’ vs. 6% of their market peers
• 47% execute decisions as intended ‘most of the time’ vs. 15%
Obviously, it pays to keep the algorithm aversion in check. A big way to avoid the aversion is to realize that, when analyzing data, different algorithms provide different perspectives and insights into the business, as long as you’re providing the right inputs and can correctly gauge the performance of the algorithm. Just as different shoes fit different people, some algorithms are more fitting for certain problems than others.
By embracing algorithms and taking advantage of the insights your data can provide, your decisions can be made faster, more accurately, and as you intended.
To see how algorithms that provide error handling, accuracy, and stability give you confidence in results, visit IMSL Numerical Libraries.