One of my favourite things about open data is the unexpected value
people find in it - sometimes far removed from its original purpose.
The
LIDAR data
published by the Environment Agency is a perfect example.
It's a more-or-less complete topological survey of the
landscape of the UK, collated over several years and released in
well-structured and well-catalogued chunks available for anyone
to download. Its original purpose is for things like flood modelling and
monitoring soil erosion, but that's by no means all it gets used for now.
A few years ago I became aware of
this project by Chris Gutteridge,
and his further adventures writing a
script
that combines
LIDAR and Open Street Map to generate a Minecraft copy of a real place.
Having played with this script to build more Minecraft versions of my home and my then-workplace
than I could possibly use, I wondered what was next.
Is it just a gimmick, or does it have a legitimate practical use?
At Plymouth City Council's
fourth Data Play event
in 2017 I got chatting to Leigh Cooper during the coffee break. Leigh is the founder of
Nudge Up
and a behavioural psychologist with an interest in education.
We wondered, with the event's focus that year on green space and the environment,
whether we could come up with a project to combine our respective
skills; a lesson pack to engage schoolchildren in caring for their local
green spaces. I'd recently been introduced to the world of Minecraft
mods (curated content packs that add new items to the base game) and
we realised there was potential to provide a richer biosphere for
players to explore.
Together we designed a small pilot project that could be rolled out across
multiple schools. This included a field trip to explore the real place,
and a classroom session using Minecraft to model improvements the
participants had identified as things they'd like to see.
The project itself is written up in full here:
http://www.dataplymouth.co.uk/green-spaces-minecraft
so I won't reproduce the full report in this post. Instead I want to
focus on how we set up the learning environment and what we discovered.
Virtual classrooms have this in common with
physical spaces; learners need to have confidence in the facilities, the
content and the tutor. Their questions are pretty much the same
regardless of whether the virtual space is Zoom, Teams, or
Red Dead Redemption
.
Here's how we tackled these in Minecraft.
With the Teats Hill project, this was straightforward. The spawn location was set
for everyone in a floating classroom high above the world map, with a
teleporter to take participants down to the focal area. Since we were
in the same physical classroom as the participants we were also on hand
to direct them if needed.
However at a similar project for
ThingsCamp
, where people would be joining from
multiple locations and at different times, we created a spawn point with
shelter and signposting and placed a number of teleport points to
connect different learning spaces. Above all else, things needed to be
immediately visible, self-explanatory and uncomplicated.
In Minecraft it's possible to set starting conditions for players as they join, including the resources they have at the very beginning of the session. The three most common options are:
For Teats Hill we went with the second option, with chests pre-stocked
with food and tools up in the classroom. We knew we'd be present the
whole time the students were in game, so we'd be on hand to provide
anything else that was needed.
For the ThingsCamp world we needed a slightly less manual approach,
since we wouldn't necessarily be there to refill the chests each time
they got emptied. Instead we made use of command blocks to create an
infinitely renewing starter kit, replenished every time the chest was
opened by a new person arriving.
Where only a single item type is needed, the dispenser works best; for example we
used this to give players small stacks of redstone, iron or gold
at the redstone testing ranges in the ThingsCamp world.
In a classroom space of course that's the teacher's role. In a remote
learning session this should still apply, but surprisingly there are
still occasions where I've signed in to deliver content and found I
can't access basic functionality like sharing my screen, setting up
and assigning breakout rooms, and so on. A little frustrating, but
eventually you learn to manage.
No such issues in our two Minecraft project examples, anyway.
In both cases I set up the server, generated the world, configured the
access rights and assigned my player god-tier privileges. Flying,
teleporting, conjuring stacks of esoteric resources into being with a
flourish of my typing hand ... Not a problem.
But is that all? If like me you've suffered at the hands of an educator
with an authoritarian streak paired with no-better-than-average domain
knowledge you'll be aware that no, it certainly is not.
Is the leader - teacher, tutor or whatever - good at reading the room?
Are they paying attention to mood, engagement and activity? Do they know
enough about what they're doing to command the required level of
attention and collaboration? Participants may not be fooled by a string
of buzzwords - you just try convincing a room full of gifted and
talented ten-year-olds that you know as much as they do about Minecraft,
let alone that you know more and hence they should be listening to you!
One of the dangers of using a game environment for learning is that gaming behaviours tend to surface. The focus of our Teats Hill session was meant to be exploring and improving the park, but predictably there were minor disagreements among participants leading to some trash talk in chat. Luckily we were able to spot this and remove the offenders immediately for a short time-out and a talk about how not only is that not appropriate in a learning session, it's also not okay in gaming as a whole. Did the lesson stick? We have no way of knowing, but we can hope.
We had expected, given Minecraft's popularity with this age group, that all the students would be fairly expert. As it turned out we had one who hadn't played much before and spent the early part of the session finding their feet. What's interesting is how quickly they got to grips with the mechanics and started advising fellow players on how to make things happen, to the point that when we found ourselves struggling to keep up with requests for resources we were able to allocate them the higher- level privileges to support their friends. We were a little worried that perhaps the responsibilty would make the session less enjoyable for them, but they were very happy to become the group's 'foreman'.
In complete contrast, another student was not only a very experienced
player, but also deeply familiar with the modified environment we'd
created and wanting to work at an advanced level with some of the
in-game technology.
Luckily we had anticipated this possibility and had some assets
prepared, a few elements half-built and additional materials in caches
around the map for them to find.
Given their level of expertise it might have
seemed to make more sense to give them admin rights, but that wouldn't
have worked nearly as well. Skills do not always equal leadership, or
support, and in this case it was far more productive to allow them to
work separately at a higher level.
As the people in charge, of course we could have insisted they stay
with the group and stick to the assigned activities. However that would
have made the session far less engaging for them, and increased
the chances of them becoming disruptive.
A slightly different scenario was our explorer - another independent
soul who wasn't engaged for long by the set activities and decided to
roam the map to see where they could get to. They quickly found
themselves outside the limits of the area we'd modelled for the session,
which meant there were only featureless clay blocks after the first few minutes.
When they eventually found themselves completely lost with no landmarks
to guide them, that's when they asked for help and we were able to
teleport them back. They went on to repeat this exercise in several
different directions! We noted for next time that we'd make sure to
generate the game world with caverns, mines and caches so
explorers would have tasks and resources to occupy them, as well as regular
teleport or transport terminals to help them get back to the group.
It helps to scope out all the things that could possibly happen and
plan accordingly, but even that doesn't necessarily prepare you for
the things that come up. For example, nowhere on my risk register was
"Tutor accidentally throws bucket of lava over student" ...
Our preparations fortunately did include instant respawn
safely up up in the floating classroom, along with the game rule 'Keep
Inventory On' ensuring that everything they'd made or found up to that
point would be saved.
This is a point that's relevant to all risk planning - the actual cause
isn't necessarily the thing you need to suppress, and often these can be
outside your control. Medical emergencies, for example, or natural
disasters. What matters is how well you're equipped to manage or
mitigate the impact.
Most of our findings are documented in the project report, but to
recap the most important one: Minecraft is a legitimate and effective
tool for delivering a lesson, and all the more if the game world has been
modelled and populated with clear learning outcomes in mind.
The next point would be, this isn't just an excuse to play - for students
or for tutors! There was a great deal of preparation involved in
designing the behavioral study, the field trip activities, and the
questionnaires for educators, students and parents.
The field trip is an essential component, helping to ground the students
in a sense of the real place and what already exists. The in-game
session by itself wouldn't have been effective, although it would still
have been fun.
We'd love to work with schools and other organisations to
deliver this session to more groups - if that sounds like something
you'd be interested in,
please get in touch.