This is a paper from the Proceedings of the 8th International NDM Conference by Gary Klein, Chew Lock Pin, and David Snowden that I have seen shared in a few different places so I wanted to dig into it. I’d read some of Gary Klein’s work before, mostly on pre-mortems, so I was curious to read this one on anticipatory thinking.
P.S. I always love to hear if these emails are helpful or not and how you’re applying them or how I could improve. Please hit reply and let me know, the emails will go straight to my inbox.
Anticipatory Thinking The authors define anticipatory thinking as the ability to prepare in time for problems and opportunities.
There are three different types of anticipatory thinking that the authors are able to recognize at this point in their research though they do say that they are predicting that more will arise in the future.
- Pattern matching for individual cues
- Trajectory tracking for trends
- What they call a conditional form, which is where people are reacting to the implications of combinations of events
In this paper they discuss what problems can arise that prevent anticipatory thinking and also discuss some ways to make it better.
“Anticipatory thinking is the process of recognizing and preparing for difficult challenges, many of which may not be clearly understood until they are encountered.”
This seems like a great skill to have in our field of software dealing with complex systems and large scale incidents. The authors point out that in most cases, the ability to use anticipatory thinking is actually a mark of expertise. They cite some research from the 40s showing that Grand Masters who would look at different chess positions while determining what move to make next, would almost instantly react to a move they had in mind and either positive or negative.
The Grand Masters didn’t have to go through a long process by which they reasoned out the move. They were able to almost immediately know whether or not the effects of that move were good or bad.
It’s important to note that anticipatory thinking is_not_prediction. There is an overlap, but it is a mistake to look at anticipatory thinking as a way of predicting what’s going to happen.
They have a quote that I like and maybe I’m biased since I live in Las Vegas, but they say “we are gambling with our attention to monitor certain kinds of events and ignore/downplay others”.
To demonstrate this, they look at an eye tracking study done on new and experienced drivers. The drivers who are experienced are actively looking for hazards. They’re scanning around looking for things that might cause trouble, but new drivers don’t do that. New drivers are strongly focused on just staying in their lane and trying to keep the car on the road.
It’s not that these more experienced drivers are predicting that there will be a problem, they’re just managing their attention. The experience they’ve developed have created the ability for them to notice the patterns that could signal trouble. These signals are present, of course, for the new drivers, but they don’t yet have the sensitivity to some of these weaker inputs to make sense of it.
We’re doing the anticipatory thinking in our heads, but we’re expressing it through how we pay attention, what we pay attention to, and at what level.
Anticipatory thinking depends on our capability to prepare, and not just our ability to predict future states of the world.
Forms of anticipatory thinking
The first one is pattern matching, which is also a hallmark of expertise, because the longer you’ve been working at something, more experienced you are at it, the larger database you have in your head of patterns that you can match against. This helps you almost immediately recognize certain things. Anticipatory thinking isn’t only involving the detection of problems, but it is one of the best uses; to provide an early warning system before running into trouble.
This is often why experts may say that something doesn’t feel right or doesn’t seem right and then begin to redirect their attention even if at that moment, they cannot quite put their finger on it.
The downside to pattern matching is that it may give us overconfidence and have us be overly sure of a diagnosis and as a result fail to notice something different that may have been seen by someone else.
The next type of anticipatory thinking, is trajectory tracking. Just like it sounds, this is how people get ahead of the curve ahead of events that are going on and preparing for how the events might unfold.
Trajectory tracking is not the same thing as pattern matching. Trajectory tracking is a more difficult task. Instead of matching an input with some sort of recognized template, we need to compare what we’re observing with what is expected to happen.
The authors point out something interesting that I think can have implications for us as software practitioners, especially if we’re in an organization that does post incident review: In this area of anticipatory thinking, people are using narrative. By looking at what others have experienced and listening to their stories, we’re actually getting more patterns given to us that we can then use to determine our response to those future events as we do this trajectory tracking. More data around which we can calibrate our expectations.
As we’re looking at the trajectory of events, we’re able to see more possible places that different people or situations could end up at.
The last type is convergence, which is seeing connections between events. Instead of having this queue that the triggers us to either match or pattern or start looking and predicting the trajectory, this is where we think about the interdependencies of events and how they connect together.
This is where “we notice an ominous intersection of conditions, facts, and events.”
The authors use an example of where this didn’t happen, which was a friendly fire incident between some F-15s and a Blackhawk in the 90s.
Essentially, no one along the way had anticipated a number of events affecting the code they transmit to show whether or not someone is an enemy. They hadn’t anticipated consequences of some changes in combination with some changes of their hand procedures, after which this disaster happened.
There’s a number of contributing factors there. Each factor representing a failure of converging anticipatory thinking, a failure to appreciate how problems might arise in the future.
The authors suggest that convergent thinking might require some degree of mindfulness, but regardless of whether it is a deliberate action or not, converging anticipatory thinking requires noticing some amount of inconsistencies. But in addition to noticing these consistencies for the convergent form, we need to notice connections as well.
Components of anticipatory thinking
Anticipatory thinking is made up of sense making and explanation/diagnosing. This is absolutely critical to coordination, which is something I’ve discussed beforein a previous issue. Effective teams need to be able to predict each other’s reactions and have some ability to anticipate how they’ll respond to different events.
Anticipatory thinking is more than just gathering a list of inconsistencies or discrepancies and then attending to it when it passes some threshold, but is a form of problem detection.
Instead, this often requires us to reframe our understanding of the system in order to have the evidence that we’re noticing become significant.
This ability to compare what we expect to happen with unfolding events is the basis ofcommon ground. If we didn’t have this, we wouldn’t also have the ability to be surprised by others and then repair a common ground as a result.
Obstacles to anticipatory thinking
Unhappily, the set of barriers that interfere with anticipatory thinking is fairly long.
This includes things like fixation, or explaining away inconsistencies instead of seeing their significance, or we may just be overconfident in our diagnosis or understanding of situation.
And then on top of this there are organizational factors as there always are. There might be policies that keep the information from reaching you or a big gap between the people who are using the data and those who collected it, or difficulty in directing someone’s attention. Directability being another component of common ground.
So given this, the authors did a study to see if they could improve anticipatory thinking specifically in small teams, specifically fighting fixation. In practice this could look like dismissing alarms or other signals. They’re trying to increase the chance of someone noticing and responding to a weak signal.
So they set up this study using some different military and intelligence scenarios and gave them to seven different teams, about four to five people in each team. And in each scenario, they put in some weak signals so that they could look at whether or not the team noticed them. And then whether or not that altered team behavior. At pre-established points, each team member was to stop and write some notes about what was going on as well and monitored team discussions.
They found that there was always at least one person in a group who noticed the signals that they planted and even noticed what the implications might be. About half the group would at least just notice the weak signals, but perhaps miss the implications. But no one took these signals seriously. If they were mentioned, they were dismissed. So the group never acted on these.
This helps show that common admonition to “pay more attention” doesn’t work. These people saw the signals, but didn’t always know what to make of them, or couldn’t get the group to act on them.
The challenge then, is actually to not just help people recognize weak signals, but helping organizations and teams do something about it when individuals are perceiving them.
Improving anticipatory thinking
Given the results of the study, the authors then examine what can help the problem.
One that we’ve discussed before, is the value of having someone else from the outside and a team, that diversity of perspective as well as bringing in some fresh eyes.
Another strategy is to try and overcome weak mental models. One solution they give might be to develop expertise and having an organization that values that.
They also looked at different forms of “ritualized dissent”. For example, having those people with fresh perspectives bring actual, authentic disagreement. Also, an organization could expect members to voice unpopular, but_sincere_beliefs.
These methods seem to have helped team perform better, but it’s important to note that these are not just contrived beliefs for playing devil’s advocate.
Anticipatory thinking isn’t just another new term. It is a form of sensemaking, looking forward rather than retrospectively. … We believe that anticipatory thinking is critical to effective performance for individuals and for teams.
- Anticipatory thinking is a key component of team coordination
- Advice to “pay more attention” doesn’t help anticipatory thinking
- The use of narrative can help others improve their anticipatory thinking skills
- “Devils advocates” don’t seem to help team performance, though expressed, authentic differing view points do.
- Even when weak signals are noticed by individuals_and_the potential consequences understood, it can be difficult to get a team or organization to act on them.
- Being able to anticipate the behavior and actions of others, and to sometimes be surprised, is the basis of common ground.
**Who are you? ** I’m Thai Wood and I help teams build better software and systems
Want to work together? You can learn more about working with me here: https://ThaiWood.IO/consulting
Can you help me with this [question, paper, code, architecture, system]? I don’t have all of the answers, but I have some! Hit reply and I’ll get right back to you as soon as I can.
**Did someone awesome share this with you? ** That’s great! You can sign up yourself here: https://ResilienceRoundup.com
Want to send me a note? My postal address is 304 S. Jones Blvd #2292, Las Vegas, NV 89107