Cloud Data Viz and Analytics Health Check
Uncover the fitness of your Cloud Data Viz & AnalyticsGet my free score
The Gambler’s Fallacy and Business Decisions
With the Snowflake Summit in Las Vegas quickly approaching, I’ve had gambling on my mind. I’m interested in the intersection of psychology and probability, so I’m here to warn you about the Gambler’s Fallacy.
Here’s the scenario
You have to flip a coin 10 times, and bet on the outcome of each flip (i.e., will the coin land heads-up, or tails-up?). You start flipping, and the first 9 flips are all heads.
It’s time to place your bet on the outcome of the last flip – what do you think it will be? Here’s your three options:
- The coins should even out heads/tails, so we must be due a tails. Bet on tails.
- Heads are on a streak, so it is more likely to be heads! Bet on heads.
- There’s still a 50% chance of heads and a 50% chance of tails! It doesn’t matter what you bet on.
If you said 1, you need to keep reading because you’ve just fallen victim to the Gambler’s Fallacy! Answer 3 is correct. The 10th coin flip exists totally independently of the previous flips. The coin has no memory! It still has a 50% chance of landing on heads, and a 50% chance of landing on tails.
Sure, if we flipped the coin ten more times, we would expect the coin to land on tails about half of the time, but longer streaks can happen. The logical part of our brain tends to want to ‘even things out’ in a less random way than the universe does, which is when the fallacy arises.
What is the Gambler’s Fallacy?
The Gambler’s fallacy comes into play when we have many independent events. Indepent here just means that the outcome of one event really has no connection or influence on the outcome of another event (like two coin flips). The fallacy is simply the belief that an event that has not happened for a while is more likely to occur in the future; or conversely, that an event that has happened repeatedly in the past is less likely to occur in the future.
P.S. If you said answer B, then you’re falling for the ‘hot hand’ belief. It’s basically the reverse of the Gambler’s Fallacy, and leads you to believe that, if an event that has happened repeatedly in the past, it’s more likely to occur in the future.
Why does Gambler’s Fallacy matter?
So, you’re wondering what Gambler’s Fallacy has to do with you and your business decisions? Let’s look at some examples.
The Gambler’s Fallacy is common in the world of investing, where traders and investors assume that stocks will turn around soon after a downturn or a winning/losing streak. Unfortunately, this assumption often leads to bad decisions, resulting in hidden costs or misinformation. An investor might say “this stock has been increasing for too long, it must go down soon, I’ll wait to invest,” when in reality it may keep going up and they’ll miss out on the benefit.
There is also evidence of this fallacy appearing in loan applications. A 2016 study found that loan officers were more likely to reject a loan that’s sitting in front of them if they had approved the last loan that came across their desk. If they had just approved two loans in a row, the chance of rejecting the next one was even higher! Each loan application is independent of the one before it, so the previous decision(s) should not have an effect on the current loan at hand, which means the loan officers are falling for the fallacy.
Another example from a slightly different angle: employees might architect a conversation about budget with their managers so that they don’t ask for two or three big-budget items back-to-back. Maybe they know that their boss doesn’t tend to approve more than one ‘ask’ in a row, and so they try to design their requests to first predict, and then balance out, approval and rejection.
What other scenarios can you think of where you may be wrongly associating events that are truly independent?
Spotting Gambler’s Fallacy
In business environments the Gambler’s Fallacy can be harder to spot than in games of chance. The mistake arises when we expect one event simply because another event has occurred several times, as if there were a pattern at play. By blindly assuming that past trends will change or continue, companies may miss out on opportunities, fail to prepare for potential risks, or make poor decisions.
What to do about it
It’s important for companies to evaluate each situation and business decision on its own merits to determine if it is independent from other events. If this is the case, and the events aren’t dependent on one another (like successive coin flips), then past events shouldn’t be used to inform future decisions. When in doubt about whether an event is independent, do a bit of research, collect more evidence and data, and then proceed. By doing this, companies can make more informed and successful decisions, and avoid falling victim to Gambler’s Fallacy.
Author: Elizabeth Tofany, Data Visualization Developer – Astrato