Noisy Decisions

Kahneman’s new book focuses not on bias but noise in decisions

Bias takes the spotlight in decision-making. As it has done on these pages where I have focused on different biases in decision making – so what is noise? Noise is variability and inconsistency in decision making and Danial Kahneman believes in many business and society situations this can be a bigger problem than bias. Let’s review.

Apologies to the non-sporting folk amongst you but I will return to a sporting scenario to outline the case for noise in decisions. This is also because sport is very good example of some decision-making scenarios because the decisions are on view to thousands, and often many millions, of people who have a vested, and often very emotional, interest in the outcome of these decisions. By the sports people themselves, and particularly by the officials, normally the referee.

If any of you watch rugby, you cannot but know the focus on protecting players and particularly from head injuries. Anybody who also watches rugby will probably have internalised the referee decision-making protocol. Let me enlighten you non-rugby viewers. Rugby is a contact sport with highly athletic and beefy men (or women) running at each other to try to collectively put an oblong ball down behind a line. Sounds crazy? Actually, a whole lot of fun (yes, I was a rugby player in my younger years). In contrast to American football rugby players do not wear any protection (thin protective caps are an option). Clashes of heads are therefore common – over the years there has been a realisation that concussion and repeated concussion can have dramatic impacts on cognitive functioning and can lead to neurodegenerative diseases.

Hence sports bodies have begun to take head injuries very seriously. Rugby now has a head injury assessment. A player with suspected concussion must leave the field and go through a head injury assessment conducted by independent professionals, before being allowed to continue playing – or not. Gone are the days when we shook our head, poured some cold water over the back of our heads, and proudly pronounced, ”I’m alright lads”, and continued.

But the bit we can’t miss in any rugby match is the referee’s decision-making protocol around contact to the head. When players are tackling, trying to stop, other players in their tracks, they will throw themselves, in a controlled fashion at the other player. Mad when you first see it. But you can’t do this in any way. The head is out of bounds, and you must make an attempt to wrap your arms around the other person. You can’t just barge into them with your shoulders. This makes the tackle much safer, for both players. This has long been the case.

Now it becomes more interesting. Relatively recently World Rugby, the international governing body introduced new rules. The most dramatic is that contact to the head in the tackle gives an automatic red card. That’s it. Hit the head and you’re off for the rest of the match and face a potential short-term ban after the match. Designed to minimise head contact. But in rugby players are moving at high pace and also dodging and weaving and ducking and it becomes quite easy to unintentionally hit a person’s head. The old rule was if there was intentionality or malicious intent you may get a card, but how do we judge this fairly and consistently?

In the above situation the referee would be left to either give a red car at every minor nudge to the head or make some decision which could be different from case to case and certainly be different between referees. These decisions are big decisions because sending a man off will lead to dramatic influences on the match.

So, what happens in rugby, in contrast to football notably, is that there is a clear decision-making protocol. And referees go through this live on the field so every television spectator can hear the decision-making process with the on-field referee talking to the video referee off field after watching repeated slow coverage.

The protocol is this:

1. Was there contact to the head? If not, nothing, play on. If so, move on to part two.

2. Was this with force from the tackler? If not, nothing, play on. If so move on to part three.

3. Are there any mitigating circumstances? For example, ducking from the attacking player? If so, yellow card, if no, red card.

Once we get to stage three any fan of rugby knows there will be either a red card or yellow card. Now why is this interesting for us in decision-making and in other decision-making contexts?

Well, it shows a clear method of making decisions and therefore increases consistency of decisions across the field, reduces discrepancies, but also make it clear to all parties involved that there are no biases and personal judgements. The protocol is always the same. It increases quality of decisions and increases fairness. Obviously, it does not remove all biases. Each stage can be classed differently on situation (e.g. when is force forceful) but it removes a lot of the variability.

---

A clear red card

---

This is an example of a protocol to remove noise from the decision-making system. Noise is also the name of Daniel Kahneman’s most recent book, just out a few months ago. Daniel Kahneman, for those who don’t know, is Nobel laureate and one of the godfathers of cognitive bias, particularly in decision-making in economic scenarios. His other work has always focused on biases in decision making but this is different concept.

What is the difference between noise and bias?