Game Theory: Puzzle Motivation.

[Ali]

I've just become profoundly stuck playing Sherlock Holmes and the case of the Silver Earring and it's started me to thinking about the importance of puzzle integration. (Apologies for the lengthy post, but I'm fueled by 'that-puzzle-was-ludicrous' energy.)

I don't think I'll give too much of the game away by describing the set up of the puzzle: I want to get into an inventor's safe. The safe has a model of Noah's ark on it.

In order to open it I need to lead a series of 8 animals into the ark. The clues I need to work out this order are hidden in a series paintings and a document I recieved by solving another one of the inventors puzzles.

Having read about 5 walkthroughs I am aware of the three problems I had with this puzzle:

1. I can't tell the difference between the developers' drawing of a bear, a pig and a gorilla.

2. Only 4 of the animals have to go in in a particular order, the others are arbitrary.

And most importantly:

3. No one, not even a 'wacky' inventor, would keep the combination of his safe HIDDEN IN HIS PAINTINGS!

The puzzle is un-motivated by it's surroundings. A brilliant inventor would surely be aware that just not telling anyone the code to a combination lock is safer than hiding the solution about his house.

Likewise, locking a door with a key, and then not losing it is the best way of keeping a door locked. That's why very few of us lock our houses with complex coded key pads, levers or arrangements of crystals.

The bit I'm finding tricky is how to integrate obstacles and puzzles into a game in such a way that they are motivated by character, plot and story. I think this is crucial, and wondered if anyone had any thoughts on the matter.

[nihilyst]

Plus, puzzles should fit the scenery. I remember Still Life and that ridiculous cookie puzzle. That was so ... damn silly. Or activating a crane by solving a lever puzzle.

[ManicMatt]

Still life's cookie making puzzle.

It was the first and hopefully last time I will call my mother in for help on a puzzle in a game.

With a lot of puzzles there will always be difficulty in it being realistic, as it's not the sort of thing we do in real life. (Unless you have puzzles like fix an engine of a broken down car, find the remote control down the back of the sofa, or make cookies)

[ildu]

At the time, the cookie puzzle felt like a nice change, but in retrospect, it's quite awful. First of all, mixing amounts with encryption made it way too difficult and frustrating. Also, the lack of direction was apparent. Secondly, the puzzle was near impossible for people outside the main English-speaking countries, since many of the ingredients were uncommon to them.

I used a mixture of knowledge of the English language and the american culture with my knowledge of cooking. I knew the ingredients that go into making chocolate chip cookies before I played the game, and I knew that, for example, molasses means a type of sugar, but it was still overly difficult for me. I can't imagine anyone with less knowledge in the English language or cooking being able to solve that puzzle without help.

[MrColossal]

For me, when making puzzles, if there is no reason for a puzzle to exist I don't let it exist.

If you can't think of any real way to open a locked door other than a long convoluted process... Don't lock the door. That will "shorten" the game in a way but I'd rather spend time solving puzzles while chasing a murderer or disarming a bomb or freeing trapped alien slaves than unlocking a door. The door may just be blocking the way to all the good stuff inside the location, don't stop me from getting there.

Also, it seems people spend a lot of time trying to make puzzles where you go from easy to hard through the whole game. "The puzzles must get harder and more devious!" and they forget about a lot of simple actions that can feel like puzzles.

You have a knife in your inventory? Draw some overgrown weeds on the door [where applicable] and have the knife cut them down. Or just let the player USE HAND on WEEDS and tear them down. I didn't need to search for weed killer only to find an empty can of it but read the ingredients and find that all the ingredients are simple house hold objects I can get through other puzzles and then use the new weed killer on the weeds and they magically die in 1 second and the door is opened. Sure that adds a lot of puzzles to the game and a lot more time the player is playing the game but UGH! Quit, Uninstall, Nasty Email from Eric!

[The Inquisitive Stranger]

I'm disappointed to find that there's no real game theory in this post.

[Ali]

Pah! That just sounds like people trying to make maths exciting...

Eric, your thoughts are really interesting but I suppose the natural question is, how do you engineer puzzles that feel natural (i.e. not convoluted) which are also challenging.

Something I've been thinking about in reference to machine style puzzles and locked door puzzles is the misuse of obstacles in adventure games. The worst examples of myst-clones or games featuring alien technology tend to feature elaborate mechanical locks and the like.

Better games, like The Dig or Myst-not-clone, give you the same obstacle in a different form. You are presented with a means of achieving your goal or a mode of transportation but it doesn't work, you have to fix it or learn how to operate it.

I think it's much more satisying to fix an elevator and get to the second floor, than to try and solve yet another implausable mechanical lock.

[Helm]

Eric, your thoughts are really interesting but I suppose the natural question is, how do you engineer puzzles that feel natural (i.e. not convoluted) which are also challenging.

Hello, this is Eric speaking.

One possible answer is: making puzzles which are relatively straightforward interesting again, by stripping away layers of automation we're used to in pointless clicking games. This is sort of tall-order for AGS games, but do you have a door that's locked? Do you have a crowbar? Show a closeup of the door, actually place the crowbar and try to wedge that open. Just 'click x on y' itself can make the simplest to the most complicated thing boring.

This is all very much to ask. But gameplay makes games. Not clicking through a story.

[MrColossal]

Hello, this is Helm speaking.

For a quick at-work-eating-lunch answer I'm going to default to my long running bit of advice to people making adventure games.

Add a verb.

Current style adventure games have stripped down all the possible actions one can do in the game into "USE". Why do games even have TALK TO when USE would work just as well with characters. Use is such a generic do everything word that it ruins puzzle design. You don't need 20 verbs [turn on, turn off, push, pull, examine, look blah blah] but just try adding 1 new verb. In Spellbound I added the magic system which let me add as many new verbs as I wanted. In Automation I added KICK. In a way this agrees with Helm though I dislike the crowbar example because it's a lot to ask [as he says] and unless you build an engine around interacting with the world like that, it just ends up being a minigame.

You can keep the USE icon but limit it to just what can be done with the hands. If you create a GUI where there is HAND, EYE, FEET, MOUTH and let the player know that they don't just represent USE, LOOK, WALK TO, TALK TO.

MOUTH could be anything that can be done with the mouth. Talk, eat, blow on, suck [grow up!], bite... FEET, walk to, kick, jump... HAND has a ton of stuff it can do but try your hardest not to make it the generic verb, grab, open, pick up, press, pull...

This doesn't solve the problem of convoluted puzzles because you could still make a puzzle where you have to blow air up a drain pipe to unclog it and get a birds nest that was in there... But it opens the possibility of puzzles and frees you to think, in my opinion.

Crowbar in door: With the hand put it in the door, the crowbar sits in the door, use hand on crowbar, the character jiggles the crowbar back and forth, not enough strength to break the door down, use foot on crowbar in door, the character kicks the crowbar which breaks the door lock and the door swings open.

I don't know if that helps or is relevant to the discussion.

[Helm]

I get this: Have things that are obvious, work out in the obvious way, and some that do not.  Don't have things that are not obvious work out in not obvious ways! There's nothing I dislike more in adventure games where something is needlessly convulted. Sometimes opening the window should entail, opening the window.

But! place a few situations where something you have to do requires a few more steps to get done than usual automation in pointless click games has us used to. As eric says, he makes an 'use crowbar on door' puzzle a two-parter in a very intuitive way. It's still the obvious thing to do, it just takes more thinking of how to use an actual OBJECT on another OBJECT and not just a symbol of an object on another symbol of an object. Making people think in abstract terms is not always good for gameplay. Stuff is made from parts, with physical properties, you can look under stuff, you can manipulate their shapes perhaps, you can use both ends of the crowbar for different things, it's not just A ON B. This sort of thing is what killed the genre. More direct translation of what the player wants to do exactly (therefore, options), not just a generic USE THIS approach would help.

My solution is usually a parser. If you don't know what you want to do, you can't do anything. You can't pointlessly click everywhere.

[HillBilly]

If you don't know what you want to do, you can't do anything. You can't pointlessly click everywhere.

I like everything you and Colossal have suggested this far, but how would something like this be excecuted? Same with the crowbar example: Should the crowbar icon have two hotspots, one for each end? Or is there another way to do this?

If anyone has played the CSI games, I think that would be a nice direction for adventure games. Atleast for the whole exploring part. A more close up look on inventory items and objects, and a way to manipulate them. Of course, this needs to be done in a way that does not drive the user insane. E.g. no pixel hunting.

[juncmodule]

I stumbled across THIS a while back. While it pushes his opinions a lot, it does have some good points and is presented in a pretty clear way. I think it could be used as a nice guideline in conjunction with Helm and MrC's ideas. Personally I think this is the best piece of advice:

For me, when making puzzles, if there is no reason for a puzzle to exist I don't let it exist.

As far as modern commercial adventure games are concerned I think they all tend to miss the mark. I have yet to play anything made in the past ten years that made me think I like modern commercial adventure games (AGS games on the other hand...). Then again, I haven't played many. I did read a rather nasty review of that Sherlock Holmes game recently though.

later,
-junc

[Helm]

Pixel Hunting is just fine if it's the sort of game that continuously rewards close examination of your environment. Pixel Hunting's only cruel when a game is mostly smooth sailing with very visible items and hotspots and just has one or two very tough ones.

My idea would be a hybrid GUI like the one Bernie's using for Hero Theorem. Take away 'USE' pointer from player, retain click-intuition with look-ats and other obvious stuff like walk-at and talk-to, but with parsered verbs for anything more complex.

[Shane 'ProgZmax' Stevens]

This is not Helm speaking.  The number 1 most annoying thing to me in adventure games is having puzzles for the sake of having puzzles.  At that point you're just adding artificial length to a game that most likely is too weak to stand on its own.  Step back and think about the story and progression without factoring puzzles first.  Is it interesting?  Does it have staying power?  Are there enough opportunities for CLEVER puzzles and events to occur?  If not, rework the idea.  I try to keep mundane puzzles to an absolute minimum when I make a game, instead replacing them with a minigame (not a slide puzzle) or an action sequence.  This makes the puzzles in your game more valuable, a) because you have fewer of them and b) because they occur sporadically and are relevant.  The constant pick up item a to use on b to get c to use on d thing is really a used up concept in adventure games and when I play one that is just a string of pointlessly mundane puzzles I delete it.  Make sure that any puzzle in the game gives the player a feeling of accomplishment by challenging their mind and advancing the story enough to make the effort justified.

-The artist formerly known as Eric

[MillsJROSS]

I agree that adding a new specific verb other than a hand icon would be nice. However, if added it should be utilized enough to make it commonplace to use it (i.e. There's no reason to have a "punch" verb if you're just going to use it for one puzzle).

As far as text parsers are concerned, let me start of with I love text parser games. However, I don't think many other people enjoy these games. They're a hell of a lot more work to deal with than the point and click icons of today (for players and developers alike). If you want to make one, by all means, be my guest. I don't think they will necessarily go over well. This could be a wrong assumption. I do think they can add a lot to a game, it really does allow the player to be more interactive within the game.

I think the thing to really avoid is Fetch me puzzles. I do it, I'm not proud of it, though. These puzzles are probably the worst. It's part of the weed killer example. You have to fetch the ingredients. But it also goes along with characters who won't give you the medicine to help the old man until you've found them some fresh strawberries. It's really easy to make them, without too much difficulty on your part.

I think a lot of the time we build puzzles around characters vs situations. Sometimes this is unavoidable, but it seems that in almost all games our main character needs to know as much detail about all NPC's than a person would usually need to know. We don't need to know about the bartender's love life! And most games fall into a formulaic, hello, need information, goodbye, type of deal. For once I'd like to see a game where the main character is too shy to ask a lot of these in depth questions, that he doesn't need to know. And I want people to stop building puzzles around non-important characters lives.

That said, I do enjoy puzzles where you might have the material available, but your character doesn't have the knowledge at hand to do certain things. The only downside to this, is we come across situations where we figure it out before the MC does, and sometimes it seems there's enough evidence for the MC to know what to do, but sometimes the MC has to hear explicitly what to do, which can be rather annoying.

You should either build your puzzles around with the plot or the plot around with your puzzles. They should be closely related to each other, and not work against each other. You can allow puzzles to force different plot lines to come out if a player decides to solve something a certain way. I have no problem with making puzzles more difficult as the game progresses. As long as the puzzles don't step away from the plot it's alright. And I just want to say, while the plot is important, don't downplay the importance of puzzles. The key is in being original with your puzzles, and no one can really tell you how to do that.

-MillsJROSS

[Radiant]

I agree that adding a new specific verb other than a hand icon would be nice. However, if added it should be utilized enough to make it commonplace to use it (i.e. There's no reason to have a "punch" verb if you're just going to use it for one puzzle).

I have quite an innovative GUI design for this. I'll post a module in a week or two when I find a couple of hours to work it out.

[CodeJunkie]

To be honest, I don't mind a simplified interface if it suits the game.  The Discworld series is one of my favourites, but it only has USE/LOOK for the mouse buttons, and an inventory.  It didn't make the game easy by any means since the inventory had about 30 items at times, and so most of the game was inventory based.  I really didn't mind since there are too many items to solve the puzzles by brute force in any short time, but it wasn't obvious what to do with them.  The Curse of Monkey Island is similar too.

Don't get me wrong though, I didn't mind the verbs in MI1 since they were all used a few times each, but I don't think it's the only interface that works.  Having to move your mouse back and forth between the verbs and the scene is tiring, and having your hand sprawled across the hotkeys isn't that much better.  I get tired of the standard 4-verb interface in most AGS games as well, and pressing the right mouse button to cycle to the icon you want isn't really that pleasant, especially since the icons are different for every game, and often too ugly to recognise instantly.

The only parser games I've played are text-only, except for The Lion's Den (OROW).  I'm not that good at them, but they're a lot more fun, and more personal in a way.  In fact, I think the best interface is the drop-down menu to suit each object.  It means you don't have to think about the stupid verbs which hardly apply to anything, such as 'open/close' and 'give'.  I saw a tech demo or two with this in, but nothing longer.

As far as puzzles go I haven't got much more to add.  I like MJR's point about conversations, and I think there should be consequences for what is said in a conversation.  Every game just seems to have conversation topics that are built up and then exhausted during the course of the conversation, until talking just results in 'bye' each time.  What order it is done in doesn't matter one bit, one topic doesn't affect another in the slightest.  I want to play a game where I save before an important conversation in case what is said is pivotal to the storyline, or at least slight details throughout the game.  I want to be able to put my personality into a game I'm playing by actually choosing the pathways I want, not reading pages of linear script.  Of all the aspects of the game, the NPCs are the only things that can think, and talking to them should be exciting.

[Czar]

For me, when making puzzles, if there is no reason for a puzzle to exist I don't let it exist.

If you can't think of any real way to open a locked door other than a long convoluted process... Don't lock the door. That will "shorten" the game in a way but I'd rather spend time solving puzzles while chasing a murderer or disarming a bomb or freeing trapped alien slaves than unlocking a door. The door may just be blocking the way to all the good stuff inside the location, don't stop me from getting there.

Or just put another and more interesting way around it, just like they did in Grim Danfango.
Or Day of the Tentacle. Or even Monkey Island 2.

Do you rember those? When I first played the GF demo I was amazed with the whole climbing up to the ledge and entering the offices through windows. It was such a great and atmospheric thing.

The same thing happened, if I recall right, in Day of the Tentacle(oh you Tim, you). Hoagie was going through a chimney to get to the attic. When I was a kid and was watching my uncle playing this game, having Hoagie entering the attic window was one of the memories that got me remembering that game about 10 years later.

Also the time when you had to break in the kitchen in MI2 was great. The way they did it was in true adventure form. It was always screw the laws, and if it didn't go the ordinary way, it would go the fun way.

[Helm]

SimB, I agree with your post a lot.

[jetxl]

Let me tell you a thing or two about a thing or two.
To be honest, I don't care about "dynamic conversations". Heck, I always press through dialog when I can. I don't talk that much in real life eighter.

I'd rather had a designer focus on dynamic story and puzzle solutions then on dialog. Why do I have to solve the slider puzzle on the box if I can go to a shop and buy a hammer and a screwdriver?

And since when do adventure games hate FedEx puzzles! Plot coupon X is out of reach, but as soon as you got it you know that you need to give it to person Y to solve puzzle Z. It's a great way to get some pasing in the game, and it gives people a desire to get the item, thus playing your game obsessively.

Last, I would like to say that games are set in Game World, not real life (I sometimes forget it as well). Game World is like Movie World. If it's not about the story, character development or setting atmosphere then don't use it at all.
Real life soully excists out of events that are left out of the Movie World.