Common Cognitive Errors

I sat through a pretty interesting class last week on collective decision-making and some common mistakes and misconceptions that are made in that process.  We talked about how collective decision-making influenced the launch of the Challenger and some of the steps leading up to the catastrophic gasket failure.  The lesson there was obvious, to say the least.

In this post I'm going to attempt to contextualize a few common decision-making errors and try to relate them to Northwestern Ultimate.

Seven biases that limit effective decision making:
1. Escalation of Commitment (i.e. Sunk Costs)
2. Loss Aversion
3. Confirmation Bias
4. Illusion of Transparency
5. Overconfidence Bias
6. Anchoring
7. Sampling on the Dependent Variable

1. Escalation of Commitment
Essentially, decisions should be forward looking and not oriented on the past.  So a good example of this for Northwestern Ultimate might be a coach or captain being hesitant to change our offensive strategy, perhaps considering that in the previous year we spent a lot of time and built a lot of team knowledge of a certain system.  And so perhaps the leadership doubles down on what was not an ideal offense, using a lot of practice time resources on it and ending up in largely the same non-ideal position.

2. Loss Aversion
This is a status-quo preference.  So for Northwestern Ultimate, perhaps a captain or coach is concerned about making a big change because this could result in a rankings drop or a lower end-of-season finish in the series.  Basically the idea is centered around worry related to "What do we stand to lose".  Two re-framing solutions were offered to this, both of which I totally loved: A. Re-frame it as "What do you stand to gain?", or B. Re-frame it as "What do we stand to lose if we don't seek change?"

3. Confirmation Bias
The idea here is that we seek out information that confirms what we know.  So for example, at a strategic level, if I think "side stack is the best offense", I will seek out examples that show great side stack offense to justify my position.  At a player level, when I see a good player make a good play, I might think that confirms my belief that they are good.  But if I see someone who I think is a good player make a bad play, I am naturally quicker to dismiss it as a rare mistake.  This is certainly on display in the tryout phase for Northwestern Ultimate, and is very difficult to address.

4. Illusion of Transparency
This one is a "tendency to overestimate the degree to which our mental state is known by others".  I think this is basically a fundamental underlying tenet of team chemistry and has very broad applicability.  It seems to me like this comes into play in basically every interpersonal interaction within the team, and the better your team is equipped to navigate it, the better your team chemistry will be.  Things like team retreats seem to be good ways of helping to build a base from which to navigate this.  Also formalizing feedback mechanisms and check-ins with players feels vital for this.

5. Overconfidence Bias
This means that people tend to be more certain about things with a little information, i.e. "A Little Knowledge is a Dangerous Thing".  An obvious one here is when players put too much value on power rankings or what they've heard about a team, without having seen them play.  In the past years, going into a game, players have expressed remarkable confidence that they are outmatched or are going to lose, before the game has even begun.  Obviously thinking this is not how to set your squad up for success.

6. Anchoring/Insufficient Adjustment
Essentially this is saying that we overvalue our initial reference points of things, even if those initial references are irrelevant later on.  And also, we don't adjust as dramatically as we need to.  A good example of this is on an individual player level on defense.  For example, maybe we're playing a matchup scheme and player X gets toasted deep.  One option is I can try to coach them to adjust and build in a bigger buffer, but if I'm not explicit in the magnitude of that buffer (or if we haven't really practiced this kind of adjustment), the change that is made is often not as big as it should be.

7. Sampling on the Dependent Variable
This is sample selection bias.  The example that was provided in class was "books on management that just look at the strategies of successful companies".  The idea here is that without a full set of data, (i.e. strategies of both successful and unsuccessful companies) we aren't able to gain very much insight into what really works and what doesn't.  So for Northwestern Ultimate, this means that only looking at what frequent nationals-qualifying teams do can only go so far, it really be best to understand the key differences between those teams and teams that do not make the big show, in order to draw more functional links between behaviors.

----

One thing I forgot to mention last week was on my mind this week, so I want to put it here as food for thought:

Top Down Change Often Leads to Poor Implementation

Simply mandating that people do something and enforcing it with:
- Mandatory training
- Complaint Systems
- Tests
Does not result in successful implementation.

What works better:
- Voluntary (Opt-in) Training
- Self-managed teams (i.e. Ownership)
- Cross-Training (i.e. Breadth)
- Mentoring and Peer Mentoring
- Task Forces (i.e. Empowerment)