Home » On The Priorities Of Graphic Depiction 3: People (NPCs)

On The Priorities Of Graphic Depiction 3: People (NPCs)

The story so far…

This is the third article in this series, and the second of a set of mini-posts that I’m going to be writing and publishing as quickly as possible, something I’m calling a mini-blitz. My normal publication schedule will resume at the end of the series.

Each post will examine one of the specific image categories nominated in the first post of the series, dividing them up into strata of commonality.

The goal is to define a set of policies, processes, and principles that use the game value of the result to define how and how hard it is worth searching for a particular image within each category / strata combination.

The first mini-post dealt with Objects. Today, the subject is NPCs.

People (NPCs)

Compared to the category of Objects, People are relatively straightforward.


Mundane NPCs are your generic crowd scene, or representatives from such a scene. These can be divided into four subcategories:

  • Crowds that say nothing more than “the location is crowded.”
  • Crowds that are important because of the common activity being depicted.
  • Groups that are important because of some collective common feature.
  • Individuals with whom no significant interaction is expected.
    The Location is crowded

    The crowd are superfluous window dressing in such images; the importance is actually dependent on, and attached to, the location itself, and (if anything) the crowd is likely to be obscuring that, though they would impart a greater realism to the scene. Such images should be evaluated as depictions of the location, which will get dealt with in a later mini-post.

    Depictions of activity

    Shoppers in a bazaar or marketplace, for example. This category is a half-way house in which the activity of the crowd and the location tend to be of equal significance, but an image of rioters would also fit into this category with no location value being evidenced.

    Such activity depictions tend to be plot-significant, so these images are immediately possessed of Game Value. Depending on the nature of the crowd, they may be either easy to come by or extremely hard to find. “Medieval rioters” would be quite difficult, for example, although you might succeed by thinking outside the box – a still of villagers from Frankenstein, for example, might fit the bill.

    Common features of a group

    The King’s Guards. The palace courtiers. The Riot squad. An invading army of orcs, or androids, or martians. An ordinary group of people all suddenly wearing the same badge, or mask, or whatever. A demon horde.

    No way any of these would ever be relevant, is there? Quite often, it’s the mere fact that there are a group of them that is the most important fact to convey; the actual appearance of the group is secondary in such cases.

    This can get tricky, because you might be able to find an image of a representative member of the group but not of an assembled group. That puts you in the position of displaying the individual as a representative member, or of editing the image to insert additional copies of the individual which are then manipulated to create diversity to the required level. I’ve worked it both ways, depending on the Game Value of the group, but in general, the first would be the most acceptable compromise if an actual group shot can’t be easily obtained.

    Individuals of no plot importance

    The final subcategory deals with another form of window dressing, the movie-set extra. These exist for no other reason than to add verisimilitude to a scene, but they can be useful in hiding inappropriate content from the viewer in a visually-arresting way. It’s just as much work to insert a small dragon into an image as it is to paint out an air conditioner, but the scene-plus-dragon is likely to be the better result.

    I’ve also used this technique to cover street signs, fire hydrants, and to replace inappropriate (modern) vehicles with something more era-specific; the general term I use for the procedure is ‘time-shifting’.

    For example, you can take a modern image (without too many people) of a British village’s historic town center and use a few bits of window-dressing to set the image in a medieval era. All you need to do is recognize the possibility of the altered image when viewing the source, and plan what you are going to need to correctly “dress” the scene.

    Beyond such purposes, though, rent-a-crowds have very limited Game Value.

So there are some functions of appreciable Game Value for images of mundane (in plot terms) people, and some of negligible value. This makes your intent in choosing to display such an image (assuming that you can find or make one) critical. You need to have a specific purpose, and you then need to select images or image elements that achieve that specific purpose. Anything more that you might get out of the image is a bonus.

Clearly, some of these will be more easily-obtained than others. As a general rule of thumb, the greater the Game Value of the resulting image, the more difficult it will be to find the right image. This simple relationship means that you should search until you find something suitable if a search is warranted at all.

Ideally, you will find two or three images to choose between, but there is often a degree of luck associated in finding anything at all. Therefore, when undertaking such a search, I don’t take the first result, but continue until either I have enough such options, or the search extends beyond what I consider the Game Value in prep time – at which point I choose between whatever I have found. That might be one image, or a pair of images, or even more – if the plot value was high enough that I had continued the search after achieving that ideal result (it has happened).


Common People, from the standpoint of Game Value, are NPCs who may be named, but whose identity as individuals is subsumed to some other factor. That factor may be a personality trait (“angry young man”) or a social trait (“fop”) or a profession (scientist) or whatever.

The interaction with one or more PCs gives these individual NPCs significant levels of Game Value, and this was one of the drivers of the high Game Value score in my initial insight into the subject (the chart shown in the first post of the series)..

Naming such individuals gives the option of addressing them by name, having them introduce themselves, etc – it makes interaction easier. But it is often unnecessary. The rule of thumb I employ derives from the anticipated interaction between the NPC and the PCs – if this is such that it would be reasonable for the NPC to offer their name, or if the PC is likely to be directed to speak to the individual by name, then I name them.

Searching for such images is usually a case of searching for depictions of the “other factor”, possibly with additional qualifiers. Having a broad vocabulary helps. When searching for “Brazilian Deckhand 1930s”, the actual image chosen after a recent search was found using just “Tropical Deckhand”.

Because you will expect to make multiple searches, you need to be decisive. Use any tools on offer (image size especially) to narrow your search down. Any ‘contenders’ should be opened in new tabs and keep going until you have five or six of them, then winnow down to the most satisfactory image.

What’s more, amortized effort is again a consideration – the same NPC may reappear multiple times in the campaign. It’s worth spending a little extra time on such image searches because there will be virtually no effort required for subsequent appearances.

In terms of availability, people are one of the most commonly photographed subjects. That tends to produce a lot of images for you to choose from. Specific restrictions bite into that ubiquity – some more than others. Some searches will yield a lot of results, some few, and it can be unpredictable. So budget your time expecting trouble and take advantage of it when random factors align in your favor.


This sub-category generally indicates the need for an image with multiple specifics, but not a real, named, individual. Luftwaffe Captain with a scar, Riverboat Gambler with cane, Monstrous Hulking Porter, Beautiful Blonde Concierge… you get the idea.

The one thing that members of this sub-category always have in common is that you expect there to be considerable interaction between this NPC and one or more PCs, either now or in the future. That in turn means that the character that the image is to depict will play an important role in the plot, and that it is all the more important for the players to be clear about the character that’s doing the talking..

All the advice in the previous section still applies. Both the Game Value of the image and the difficulty of finding a match have increased, but in proportion, so the same standards of results apply. The increased plot value / interaction level goes on top of that, so this is a search in which it is worth taking your time and being a little more exhaustive.

I’m going to reference this sidebar again in the final part of the series, but it’s an important search tip for right now: you will often get better results, faster, if you do your image search and then write your descriptions, etc.

In the Adventurer’s Club campaign, we set a high standard for such images in terms of the amount of personality conveyed by the image. Sometimes, those results are achieved by searching for the personality profile we want (and being less selective with respect to other character traits), but more often, we will search for some specific desired quality and cherry-pick the results that show the most personality regardless of the makeup of that personality – then we can pick between the population of the resulting short-list.

The image above could be any government or business figure. The emotional content is subject to interpretation, but the image itself displays a lot of personality; it has impact.


There are two types of Unique individual. The first is someone real, often doing something specific – Stalin making a speech, Lincoln tipping his hat, Churchill looking stoic, etc. These are no more difficult to find than Specific images, but have much higher Game Value by virtue of the baggage and reputation that they carry. This is even true of images of individuals that the players won’t necessarily recognize – for example, James Buchanan – at least until you provide a relevant biography.

The other type is of an individual whose specifications are sufficiently distinctive that the likelihood of a successful search plummets. As noted earlier, the Game Value of such an image tends to rise proportionately, which means that you can justify spending quite a bit of time and effort generating a custom image, which is quite likely the only way that this sort of image search can succeed.

In both cases, the Game Value is about as high as it can get. These are always important characters in the adventures in which they appear (otherwise there would be no point in such distinctive characters appearing).

Depicting important characters is therefore about as important as it can get.

To close out this mini-post, I thought that I would repeat an image first shared about a year ago, of Brother Simon, the Pacifist Dalek, which is an example of the second sub-type of unique character.

Next: Monsters and Encounters

Print Friendly, PDF & Email


marchape is an entertainment website, strongly connected to the media markets.
Our contributors create highly enriched and diversified content, with the main goal to serve all readers.

View all posts

Add comment

Your email address will not be published.