Official Blog of the Adventuron Text Adventure Creation System

Tuesday, 21 July 2020


What follows is a very personal opinion on the use of LOOK, SEARCH and EXAMINE in the context of interactive fiction / text adventure games.

Traditional IF/TA games typically allow the player to move around quasi-geographic locations in an environment - described in some detail initially. Nouns representing scenery and objects within a location in the environment may be interrogated to find out more information, to discover new layers of descriptions, or even discover new objects.

There are largely three verbs that are employed in this interrogation. EXAMINE and LOOK are generally synonyms, and refer to the act of looking at an object. SEARCH is an additional verb that generally means rummaging or deeply looking at or inside or atop of an object (or noun). 

Many text adventure / IF games take the approach that examining and object and searching an object are two different actions and therefore should be represented by separate verbs. Superficially, this makes perfect sense. Looking at a bag is not looking inside a bag after all.

The former is the act of looking at something (somewhat passively), and the latter is the act of digging through or thoroughly looking at the contents of something (an active action).

When the player types EXAMINE an object should be looked at, when the player types SEARCH, the object should be thoroughly examined inside and out.

Whilst EXAMINE (or the common synonym LOOK) is almost unavoidably required for the purposes of revealing new clues or sub-items that would not fit in the location text (examine and look are synonyms in Adventuron by default), SEARCH is something that really essentially means EXAMINE MORE (in the context of the object).

There is a correctness to keeping SEARCH and EXAMINE separate, but just like gameplay physics in 3d action games does not represent reality, perhaps we need less correctness for the purposes of gameplay in IF/TA.

So let’s get started …

The problem with using SEARCH is that, except when clearly signposted, the player has no idea when something should be examined and when something can additionally be searched.

Some games never use SEARCH, and some games use SEARCH for a few or just one object.
This is purely my own opinion, but SEARCH, if in your game, should be a synonym for EXAMINE (or look).

Making the responses to EXAMINE and SEARCH is padding at best (EXAMINE could fulfil the same role), frustrating at least (if the player knows SEARCH is a verb in the game, then when they are stuck they now have to scour the games for new nouns to SEARCH), and game breaking at worst (the player has played games without SEARCH and thinks that EXAMINE is the same as search, and therefore the game is unwinnable for the player).

There may perhaps be benign uses of search such as items that literally beg to be searched, such as piles of leaves or bodies, but I would still say that the problem with implementing SEARCH as a separate action even if just once in your game is that now the player has uncertainty if every noun in the game will yield progress if they try SEARCH on it too.

To make SEARCH and EXAMINE different response handlers is to invite the player to have to SEARCH and EXAMINE every object, once the moment comes in the game where they get stuck.
In the case of the pile of leaves, you could code a joint EXAMINE and SEARCH routine that yields the same result. If there is an object hidden in the leaves, then reveal it with the examine after describing the leaves.

If an object is clearly signposted as being searchable, then the separate SEARCH is redundant, tell the player upon examination that an additional search was performed if a search action is required to move the game forward.

If the SEARCH is difficult to predict that it will yield a different result to an EXAMINE, then all the more reason to not not require an input the player is unlikely to type (except via a grind pass of the nouns).

If you don’t want to code a combined handler for EXAMINE and SEARCH, then code a single routine that will point the player towards examining “Examining objects will also search them if necessary.”.

As long as the player views SEARCH and EXAMINE as the same then they will not think they have to brute force nouns when the going gets tough. The moment they find one different response for SEARCH, then they now know that they have to grind, or even worse, they are not aware to SEARCH.

If parser-based IF/TA ever has a hope of being acceptable again to a winder audience, then in my view, then reducing parser friction is the only priority.

Almost all graphical adventures have an EXAMINE feature, but not one of them (to the best of my knowledge) has a SEARCH features, because it’s redundant. Imagine there was a SEARCH verb in Monkey Island, and one object something hidden inside. Would that add or detract from your enjoyment of the game?

This isn’t saying that a game can’t be complex, but a game should not be designed to integrate TWO grind mechanics. The EXAMINE mechanic is somewhat of a necessary grind evil, given that examining nouns in a room has next to no creative input from the player. Doubling up the grind, however “correct”, has an adverse impact on game flow (imho).

I’m advocating for making SEARCH and EXAMINE interchangeable ESPECIALLY when the game has an object that must logically be searched. It’s a small precision compromise, but I wholeheartedly believe it’s the right thing to do.

LOOKING IN or LOOKING ON something is different to search in my view. Containers are clearly understood by humans and it’s entirely natural to look in a container therefore if a grind instinct was invoked on LOOK IN and LOOK ON, you will be dealing with a very small (and non frustrating subset) of nouns in the game.

I see no reason why a LOOK IN or LOOK ON can’t be a separate handler without triggering a grind instinct in the player. LOOK UNDER is something I generally don’t think is a good idea to have a different response (to EXAMINE) for as again it could invoke grinding (player trying to look under everything after they get one positive response).

I do think that we are on the cusp of a mainstream IFTA comeback, but if we want to escape the prejudice of the past with regards to parser friction, we should do some work ourselves to meet the player half way.

1 comment: