So when we last left our heroes they were working on the Big Push on The Coming Storm in Berlin…
Sorry to have not updated, but the good news is that we were working hard on The Coming Storm so that we could get a new draft out, particularly to play testers and reviewers. We passed that milestone two weeks ago, and we’ve been resting before picking it up again, now focusing on art direction for character portraits.
Whilst in Berlin we checked in with one of playtesters via Skype and got some excellent feedback. It confirmed something we suspected – we had too many NPCs for the narrator to manage comfortably in play.
Now that situation can evolve if you are trying to create something that can be used as a sandbox setting. You want there to be a lot of conflict between NPCs for the PCs to get mixed up in, and you would like to represent shades of grey amidst the conflicts. You want to give the narrator rich characters to work with, because part of the joy of a pre-published campaign book is to take the hard work away from the narrator. But inch-by-inch you keep adding more and more help, until it becomes counterproductive, because the narrator cannot manage the task of understanding all the plots.
We looked at relationship maps for Game of Thrones and Venture Brothers and agreed that was the maximum number of NPCs we could realistically hope a narrator could cope with.
Now we provide a lot of advice about managing a large cast – in essence start with a few NPCs and introduce more as you need – but looking at these two examples we realized that we needed no more than 60 and we had 120. That is just intimidating for most narrators, even with good NPC and plot management tools.
So we cut half the NPCs.
Now the first 30 or so of that was straightforward. We cut ‘colour’ characters who did not fulfill a role in any of the plots. Our playtesters said they were happy to use their own creativity for these folks, and we needed to give them room to do that by not overwhelming with our cast.
Next we channeled Peter Jackson (or even Pixar) and merged some characters. That showed us we were on the right course, because many of the characters who assumed multiple old personas became far richer and had a stronger role in the setting.
Finally, we again made sure that NPCs had a role in the plot – and we reduced some to just names and a line or two if they had a role in one of the clans – but were not exemplary.
Doing this revealed something else – we needed some ‘introduction to the setting scenarios’. Borderlands has a great example of this – the first scenario walks the PCs through the setting, showing rather than telling them what is going on. The players meet the key NPCs and discover the key plots and conflicts. We realised that The Coming Storm needed something similiar, to show vs. tell to help new players find all the NPCs plots in the setting that they might want to run with.
Now The Coming Storm is a sandbox that you and your players can wander around. We have material covering 1618-1625 but the idea is you could discard the suggested adventures in most years (although you probably want to keep the great events like the Great Winter or Battle of Dangerford); but some playtesters wanted to just play the pre-written scenarios as a campaign without that build-your-own path element. And we didn’t support that option as well.
So for those two reasons we added some additional scenarios. In fact, we dug out and re-edited scenarios we had run at cons to, wait for it, introduce the setting. One to introduce adventures in the Red Cow clan, one to introduce adventures with the neighbours.
Those changes fulfilled ‘show not tell’ an helped make 1618-1623 something you can just play straight through, if that is your preference. We still make 1623 and 1624 much more freeform, although we come back to more events driven episodes in 1625, because by by 1623 you likely have so much divergence (if playtest is anything to go by) and published material would struggle to make sense if it was too detailed. 1625 is plotted just for the great events that happen in that year.
Now sometimes you get a creative moment where parts of a puzzle snap together. With a reduced cast list it was possible to think about using those two scenarios to introduce a number of plot lines and key players, and some foreshadowing of later events. So those two elements: reduced head count of NPCs, and introductory scenarios really help make this product gel and the whole just hums along better for that change.
Such is the value of playtesting and listening to criticism.
We still give advice on how to manage NPCs and we will have a lot of visualisations – the playtesters have to suffer more than you because they don’t have those as art needs commissioning first, but the list is more manageable. To some extent making the playtesters suffer helps to expose the weak points. And the cut material will get recycled somewhere (perhaps even as Wyrm’s Footnotes additions)
But there were a few more changes that anticipated so this post got delayed.
We’re focusing on art now – of course we will tweak the material all the time, right up until layout I would guess – but this really felt like hitting the key milestone on getting ready for that next phase.