Contacting Orcasci

US +1 866 616 7366
US +1 917 387 4334
UK +44 20 3289 8375

+1 866 616 7366


Founder's Blog

For more insight into our founder's wide range of activities, visit Mark Turrell's blog

Book List

Check out our Reading List

Entries in network theory (5)


*New* Lectures & Workshops on "Science of Spread" available

Interest is growing in understanding how ideas, behaviors, and products and services spread. We have now designed a variety of Learning Formats to introduce you to the Science of Spread™.

Spread is a specialist discipline, a new discipline, one that can have a huge positive impact for all organizations. Even if some of the techniques can be guessed, it is complex and broad enough to require specialist insight and knowledge. Our Learning Sessions are designed to provide individuals and organizations with the knowledge they need to design effective spread programs, based on solid scientific understanding and real-life cases.

Yes, you can keep muddling your way through. But, how much better would it be to know what you are doing?

Give us a call if you are looking for a workshop or lecture for an internal event, a conference or a university course, delivered by an experienced and engaging speaker.

See the document below for more details.


Why Mushrooms hold the Secret to Contagion Ignition

- “Follow your peers"

- “Keep doing what you are doing until there is a compelling reason to change"

These are two very important, deeply simple rules of thumb, or heuristics, that help us as humans living in a complex world survive every moment of our existence. We use these heuristics as individuals to work out what we should do in a given situation. There are lots of heuristics (another one is the crucial "don't bump into things") and research indicates that mostly we use a 'best fit and satisfice' method of heuristic selection. That means simply: run down the list of heuristics, and stop when you have found enough to make sense to keep on living.

These two heuristics therefore explain all kinds of behaviors, from queuing to massing in the streets (or not – as 'not massing' is also a behavior pattern that can be followed).

Now an interesting question to work out is how anything new gets started. After all, if everyone keeps doing what they are doing, why would anybody ever change?

One might start at the concept of deviance, whereby one or more people – let's call them 'people' to be simple – exhibit deviant behavior from the rest. They might do this for any number of reasons, from being mentally ill to misunderstanding what people are supposed to be doing. Either way, irrespective of their inner workings, their behavior may deviate from the normal pattern, and in this way they are 'deviants'.

One might think that 'young people start things'. Well, there is some truth to that, but note that by and large very few things get picked up by anyone else, and so the fact that occasionally a young person may start a trend would not be a compelling explanation.

A phenomenon that actually allows you to study how things really start (before we get into mushrooms) is a standing ovation. Simply put, standing ovations do not and cannot start at the back of the room. This is the simple secret to contagion in its very early stages.

Standing ovations cannot start from the back. They cannot start from a mile outside the auditorium. They actually can ONLY start from the front of the room. And they rarely ever start in the middle of the performance ("it's just not done").

The reason for starting at the front is obvious. A person trying to kick off a standing ovation from the back is totally ignored, not because they misunderstood the performance, but because not enough people can see their actions. In contrast someone who stands up at the front, at the appropriate moment, is noticed by lots of people (even if they are just trying to escape the theater early to pick up their valet car or coat before everyone else).

A standing ovation works like momentum spreading through a network. In this case, each person functions as a 'node' in a network. The network shape of an auditorium has rows and rows of nodes, stretching back, and up if the space is big enough. Now think of each node having either a passive or an active state. Everyone starts off in a passive state. The seating in the theater creates the structure of the network, but it takes some energy to agitate the nodes to turn them into active nodes, i.e. those that take part in the standing ovation.

Now that we have a concept of active and passive nodes, in the first moment – let's call it Ignition – one node will become active. As we are studying a standing ovation, by necessity (and design) this node needs to be in the front few rows. Which node? We'll cover that later, that will be the piece about mushrooms – and maybe the randomness of quantum physics. Whatever the reason, all of a sudden, one node is activated and we have ignition.

The activity spreads through the network in something I'm calling Pick-Up Waves. Generally there seem to be two distinct Pick-Up Waves, 1 and 2. After that, there will be a new phenomenon to look at, but we're not there yet.

Pick-Up Wave 1 starts due to the heuristic we used at the beginning of this piece - "follow what your peers are doing". It also uses a variant of the same, potentially deviant, behavior that sparked the initial node into life. In both cases, more nodes get activated, typically in some kind of proximity pattern to the original ignition (front rows, but not always next to the ignition node). Not necessarily masses of nodes, but enough to form a wave that is visibly spreading.

But wait- not all pick-up waves are positive! It would be a mistake to assume that one person standing up will always encourage others to do the same. It turns out that nodes can activate or remain passive. If they remain passive, this is effectively a form of 'anti-action', and the resulting pick-up wave is one of non-action. This is an important point, as the rules of following still apply, it is just a 'negative follow'.

Pick-Up Wave 2 is similar to Pick-Up Wave 1, except that there are more people now standing (or not, for a negative pick-up). The positive Pick-Up Wave has more mass, and there will be some 'node jumping', with people further back and on different sides of the room standing up. Each, by the way, may be standing up for their own reasons. And each may deny that they stood up because they saw other people doing it. Nevertheless, now we have a full blown pick-up wave in force.

Have you ever noticed that all of a sudden the room goes from a spotty standing ovation to everyone and their granny standing up? (apart from granny sometimes – more on that later). At the ignition phase and the pick-up waves, the driver of human behavior was primarily emergent, totally voluntary, with a good degree of random activity thrown in. After the second pick-up wave, the story changes, and the system goes from being emergent to structured. Now a new rule takes over: "I need to see what is going on".

There are now sufficient people standing up to make it nigh on impossible for most other people to be able to see. And so, even if they did not think the performance was that great, they find themselves standing up, en masse, in a wave that propagates to the back of the room. It is worthwhile saying, just for completeness, that the wave does not keep going indefinitely. Obviously there is a wall between the auditorium and the street, but it does highlight that there are constraints to how far any wave can propagate.

Note that this system is practically non-voluntary. This is made even clearer by the fact that those people who do not stand up instead end up in isolation. Usually they have a very good reason not to stand, such as a broken leg or being in a wheelchair (hence the granny reference above). At the same time it is obvious that they are not forming part of the group behavior, and in some way they are 'bad' as they do not conform to the new normal behavior.

Thus we can use the standing ovation to work out how things spread, quite simply, a micro-second at a time.

Which leads us finally to mushrooms, the secret to ignition.

For some reason, mushrooms always tend to sprout up, seemingly out of nowhere, in the same place over and over again. It might not be in exactly the same place, more 'thereabouts', but close enough for you to remember.

Mushrooms are fungi. In order for the fungi to sprout as a mushroom (note: I am not going into advanced biology here, it is a metaphor), the spores need a substrate to grow on in addition to enabling nutrients (soil). They sprout out when there is a catalytic reaction (heat and sunshine), and out pops a lovely mushroom. The key element in the case of ignition is that the spores are on the substrate, appropriate for the mushroom, and the nutrients are in place.

A standing ovation starts from the front of the room because, in the world of biology, that is the substrate and best nutrient environment for the wanna-be mushrooms. People who like a performer might book a ticket earlier, guaranteeing them a good spot. They might be willing to pay more money. They may even 'prime' themselves during the performance that the evening was amazing, so they are ready to sprout…

We just don't know which mushroom – or now we go back to 'node' – will activate first. We just know it has to be one that sits in the front of the room.
There is a certain spookiness to the random behavior of the nodes at the front. We know that one will ignite, but have no idea which one. In the same way we really do not know how the Pick-Up Waves will work, positively or negatively, and which of the nodes will activate. We do know that there is an imperative to have practically everyone else follow in the end, irrespective of desire, but these first batches in the path to contagion are random.

I'll work more on random and the spooky nature of random connections in other papers (I have already written some on the spooky nature of abundance). Suffice to say here that it does have 'knowable' properties, especially if you look at the picture from the whole perspective and not fixate on any individual node.

Hopefully this paper / blog post will be useful to you. It is a simplified way of looking at how things like the Wikileaks release of the State Department cables can spark revolutions in Tunisia and Egypt. And hopefully it will get you started looking at the world in a slightly different way.

Try it, next time you are in an auditorium, clapping or being in a standing ovation. And you'll start seeing the world with a new lens.


Changing Network Shapes

It is a good day for insights on networks today. I'm sitting towards the back of a women's group, a coffee morning at a school. When the group came into the room there were rows and rows of chairs set up. After thirty minutes, the room has been entirely reconfigured, and this has inspired me to think more deeply about networks, network structures, network shapes, and network nodes.

I created a little network spread diagram to illustrate. Each frame is a moment in the network, a snapshot of structure and behavior. In the diagram, the black curves are chairs, and the red dots are people, with the arms indicating the direction they are facing.

Chair network development

In this scenario, the chairs gradually get moved into new patterns, based on the actions of the people. This itself may be an interesting observation - the chairs are really 'infrastructure', inactive but necessary components of the network. They become engaged or active when a person sits in a chair.

Some positions are not conducive to spread. In the second frame, there are two groups who are not facing each other. This is quite unfriendly from a human perspective, and very uncomfortable if you have to keep turning round and stretching your neck muscles to listen and talk. This human need begins to change the shape of the network, both the positioning of the infrastructure of the network (the chairs face a new direction), and the network agents or people.

In terms of network nodes, the scenario illustrates the difference between the infrastructure of a node (in this case, a chair), and the active component of a node (in this case, a person). A chair cannot spread a message, but the chair creates the shape of the network. A person can exist outside of the node infrastructure, but that makes them less able - as a communication node - to participate in the action. For example, a seated group uses eye contact within the group, but a person standing to the side is less a part of the group and therefore can have less impact or involvement.

The scenario of the coffee morning provides a good example of prediction around network change, and also the risks in making predictions. When I made my original design, the last frame was a prediction of network development. As the three new visitors arrived, there seemed to be a logic that the circle would be the logical progression. And yet, while I had been face down drawing my diagram on my iPad, the network shape had changed:

Chair network development alternative

Instead of a single common circle, two groups had been created.

There are several reasons why this might be (and I would write more down but I have to concentrate a little on the discussion taking place in front of me). There are node preferences - the people know each other, and the separate group is more of a common group that is different to the other first group. It is harder to get to speak in a large group, and as people like to chat, there might be a good logic to split into smaller groups.

There is a lot more to this, and it is always tempting when you are writing (ok, when I am writing) to turn a blog post into a white paper or a chapter of a book. I'll resist temptation this time round... and stop.


Chinese Whispers - Message Transformation due to Spread

I whisper something to you, you to your neighbor, and so on, until the message returns to the start. And usually, the message is barely recognizable from the initial point. The act of spread has transformed the message.

This is going to be a big topic within 'spread'. Right now, as I develop some thoughts during a snow-bound coffee morning, early-stage concepts that spring to mind include:

  • Message Fidelity: Basic concept of the extent to which a message changes as it moves between nodes.
  • Message Morphing/Variability: The amount and type of change the message experiences as it spreads.
  • Message Amplification: The extent to which the message becomes 'more important' within the network, like a wave gaining amplitude.
  • Message Hijacking: An existing message propagating through a network could be taken over by another message. The original message is somehow perverted with the new message, accidentally or deliberately, and the new merged message continues to propagate, albeit with a new underlying message.

We have done work on this before, and I now need to hunt through the PDFs of scribbled notes to uncover our past ideas on this topic. In the meantime, at least the concept is exposed once more, and to a larger audience.


Environmental Conditions and Network Function

I was driving through the snow yesterday in Berlin, and as I was slowing ploughing through the traffic, my mind drifted back to speedier times, summer times, when the roads were all clear and the traffic sped through the streets with relative ease.

It occurred to me that the theory on networks and propagation through networks might be missing out on a core concept, that of the importance of environmental conditions on the network itself.

A road network is a network. Clearly a snow-bound road is not the best network state for propagation. The state of the network overall impacts the ability for the nodes (call them 'intersections' in the road case) to connect together, and for the 'packages' (call them ’cars’) to move between the nodes.

Some network conditions will be favorable to spread. For example, good visibility may allow the packages to see nodes more easily, faster, and possibly identify attractive nodes that are farther away and that may allow spread to take place more easily. Conversely poor visibility would slow down network propagation, reducing spread, and potentially grinding the network to a halt (anyone who has ever experienced London in an inch of snow will know what I mean, or O'Hare airport for that matter).

Taking the concept further, it is possible that portions of a network experience different conditions from the norm. A network as a whole, or as an average, may be stable and 'average', but portions may be faster or slower. The road network of Germany may, as a whole, be fine, but local weather conditions may make a part of the network very slow. This would bring down the average, but give portions of the network a very different dynamic in terms of spread potential.

Concepts like these will help us better understand the real dynamics of networks, and help us develop new insights into how things spread. So maybe my ten minute extra delay due to snow was totally worthwhile. Perhaps (and this is giving me another idea for another day) slowing down spread can lead to higher quality outcomes. Maybe I'll work on that on my next journey ;)