7 Rules for Puzzle Builders – The 7 Series

1st April 2012

Single Player First Person Shooter Maps and Mods for Half-Life 1, 2 and Episodes 1, 2 and 3

As a level designer I often take a keen interest in the puzzles that others include in their maps. Half Life 2 offers a huge wealth of puzzle possibilities due to the physics engine and the gravity gun. If you ask a player what their favourite puzzle in a game was they could probably tell you immediately (I’d like to know yours by the way) but they may not be able to tell you why.

Equally they may struggle to explain why a bad puzzle doesn’t work, but in reality there are rules to a good game based puzzle and if you follow them, as a mapper you can’t go wrong.

Here’s my rules that I’ve learnt through creating maps and playtesting on how to create a good puzzle.

Don’t forget to add your own in the comments.

Number Seven: The most important rule of puzzle building is that the player must never be trying to work out what the puzzle is!
Single Player First Person Shooter Maps and Mods for Half-Life 1, 2 and Episodes 1, 2 and 3

To give an example, in one of the horrid Tomb Raider games you enter an area that seemingly has no exits. After about 10 mins of logical attempts to find an exit the player gives up all hope and ends up spamming the use button around every wall (I eventually found a hidden zip line at the area’s highest point, then threw my PlayStation in the bin!).
The goal of the puzzle must always be clear and simple and obvious. Its how you reach that goal that is important.

Number Six: Solving the puzzle should require brain work AND SKILL!
Single Player First Person Shooter Maps and Mods for Half-Life 1, 2 and Episodes 1, 2 and 3

This is often missed by mappers. A really good puzzle requires the player to first of all use their brain to join the dots on how to reach their goal and then secondly, use their skills in the game to achieve it. Many game puzzles emulate HL2’s official game puzzles, for example “follow the wire back to its source” but then simply have a button push to solve the puzzle, not hugely challenging or fun. In Miigga’s recent Magellanic Stream map we see how the player is then required to use their skill with the grav gun to drop the generator through a hole to solve this type of puzzle. Well done Miigga!

Number Five: The game mechanic or logic to solve the puzzle must be known to the player
Single Player First Person Shooter Maps and Mods for Half-Life 1, 2 and Episodes 1, 2 and 3

his rule is one that, if broken, will infuriate most players. If your puzzle requires the player to learn a new game mechanic, it’s much better to introduce that mechanic earlier in the map through some sort of training section rather than at the point of the puzzle.
So for example, say, I wanted to introduce explosive wooden crates to my map, this goes against player expectations. Crates aren’t explosive in the HL2 world normally. Unless I’ve allowed the player to learn that new mechanic in a stress free section of the map prior to the puzzle, they’re going to be pretty peeved with me when they finally figure this out. (And yes I know I’m guilty of this myself in my map Daylight with the window jump! Lesson learnt!).

Number Four: Never punish failure with death
Single Player First Person Shooter Maps and Mods for Half-Life 1, 2 and Episodes 1, 2 and 3

Puzzles are all about exploring and allowing the player to explore. As a result, punishing that behaviour is bad. The player should always be able to get back up and try again. If they’re going to fall, it’s better to have them fall into water with a ladder that allows them to climb up and try again than have them pancake onto concrete. (yep, I’m guilty of this one too!).

Number Three: Abstract puzzles need forewarning
Single Player First Person Shooter Maps and Mods for Half-Life 1, 2 and Episodes 1, 2 and 3

Place the miniature dog in the suitcase and turn it 360 degrees. Riiiiiight?
This is what I mean by abstract. We’ve seen some examples of this in the past on PlanetPhillip. Some people like these kinds of puzzles, I don’t personally but hey, each to their own right?
I have to admit, I completely missed the point in the map Station Well. I just didn’t get it.
I would recommend that a bit of forewarning should be required on these kinds of maps letting players know that they should be thinking outside the box.

Number Two: If you have to give the player a hint, your puzzle needs more work
Single Player First Person Shooter Maps and Mods for Half-Life 1, 2 and Episodes 1, 2 and 3

Puzzles should be intuitive. That means that when the player enters the area they should be able to look at the scene and establish what they need to do. No text should appear telling them where to look, no NPC’s should be telling them what to do, no cut scenes should show them important elements of the puzzle, they should simply be able to look and connect the steps required themselves.
If, during play testing, players can’t do that, then the puzzle needs more work.

Number One: Only use new puzzles!
Single Player First Person Shooter Maps and Mods for Half-Life 1, 2 and Episodes 1, 2 and 3

If you can’t think of a new puzzle to go into your map, and you think about putting in a puzzle that appeared in the official HL2 game or its episodes, for goodness sake stop!
This is a waste of time. You’re probably thinking of adding a puzzle because you feel there should be a puzzle in your map. I can tell you right now, that a puzzle that everyone knows the solution to, isn’t any fun. I personally would leave it and keep on thinking or just leave out a puzzle altogether, I’m sure your action scenes will make up for it. If I have to go find that stupid orange gate wheel again I’m going to scream!

Well, that’s it

I hope this has been useful for the mappers out there and interesting for players.

Please let me know your favourite puzzles you’ve seen in mods so far and what you’d like to see more of in future.

Till next time…


  1. Nice article James, although I disagree on some points.

    7: I suppose there’s no arguing this point, although your example might not be the best one. I think you would agree that the puzzle in that case was to escape the area – that much was clear, it’s just that the solution was a little obscure.

    6: I don’t necessarily think a puzzle should require skill. My favourite puzzle is in a Miigga map. It’s just a bar that needs to be inserted into a hole in the wall. This allows the player to jump onto the bar and thence onto something else. It was simple and elegant, even realistic. No skill involved at all.

    5: Yes and no. I loved your window breaking puzzle in Daylight. The issue might be that it had never been used in a mod before BUT it followed normal, real world laws. And this for me is important. If a puzzle follows naturally accepted laws, doesn’t break any known HL-universe laws then I feel it’s acceptable. Now, whether your puzzle breaks
    HL-universe laws is debatable. I’ve never crashed against a window before in Half-Life 2, but it wasn’t much of a stretch of imagination for me.

    4: I’ve been arguing the same thing for years, although Portal and Portal 2 breaks that rule a hundred times. I suppose it’s all about context and setting.

    3: Whilst I agree forewarning players of abstract puzzles is nice, I think by definition any abstract puzzle will always be weird. That’s probably why people like them.

    2: There’s a little known feature in Minvera, where if you take to long to solve a particularly difficult section, if you look in the right direction text will appear on the screen as a help tip. It’s an interesting idea that just needs refining. I feel that something like this is a built-in difficulty leveller. Imagine a puzzle that has no interaction from the game for the first 5 minutes. At that point an NPC appears and gives a subtle hint, another 5 minutes and another NPC appears with a less subtle hint, and so on until 25 minutes go by and the NPC appears and shouts at you for being so stupid and gives you the solution.

    1: Couldn’t agree more. Even if they made it a little more interesting that alone would be sensible, but to present it as is, is just lazy.

    Lazy mappers are bad mappers

    1. Jim Partridge

      7: I guess this point is about having fun. It’s no fun when you don’t know what the puzzle is. Tomb Raider is no fun because it refuses to let you in on what the hell you should be doing.

      6: Miigga’s bar in the wall puzzle took skill!!!
      You had to pick up the bar and carefully put it in the hole. That was a fiddly bit of gameplay because it didn’t snap nicely into place when you got it near it. Go back and play it again and tell me that doesn’t take a good bit of skill with the grav gun.

      5: Well I just think that there are rules to the gameworld. Only if the player has no other options available to them should you break them. With the window jump I put the player in a space where there was no other option.

      4:Yeah that’s odd, I thought about Portal actually. I’ve pancaked so many times in that game… I still think death should not be used as punishment.

      3: Nope! I disagree with you. Abstract is annoying as hell unless you’re told to go abstract.

      2: I do like the idea of a hint after a while but that should never excuse the initial puzzle from being intuative to the vast majority of players.

      1: Glad you agree with this one. We should start a campaign… to ban certain types of puzzles lol.

      1. Mman

        Tomb Raider is no fun because it refuses to let you in on what the hell you should be doing.

        That depends on the game; the first one pretty much always gives you a good idea what you need to do, whereas later games are much more variable in that regard (and given you mention ziplines, which are only in the later games, it seems you were referring to them). I mostly agree with the article itself though, I don’t think giving hints is always a bad thing, but “hints” are abused so much to completely ruin the challenge of even already easy puzzles that designing puzzles to try and avoid them is probably a good idea. I especially agree that it’s pointless to throw in puzzles that are identical to those of the actual game/prominent mods.

  2. Oh, I’ve just remembered…

    to use their skill with the grav gun to drop the generator through a hole to solve this type of puzzle.

    I just used the crowbar and smashed the boards. What did you do?

    1. That’s what you are supposed to do.

      1. Jim Partridge

        I used the grav gun to destroy them… I would brush my teeth with the grav gun if I could though. lol

  3. 7. Definitely agreed. There’s one puzzle in HL1 that frustrated the crap out of me. If I recall correctly, it was a room with two levers in it or something. There were also some bolts of electricity here and there. No idea what the levers did or what the goal of the puzzle was. It’s just silly.

    6. Disagree, mostly. I think it’s enough that the player uses his brain to figure out how to solve the puzzle. I don’t really care that much about whether the puzzle requires any more skill after the player has put his smarts to use.

    Also thanks for mentioning my map. 🙂 I don’t really treat the wire going from the generator to the combine force field as a puzzle by itself. I use those wires simply to tell the player that some element of a puzzle is connected to some other element of it.

    Oh, and about the bar that you have to put in the hole in the wall. It’s really easy and not fiddly at all when you do it with E and not with the gravity gun.

    5. Yep. Not much to say about this. Agreeeeeeeeee

    4. Haha, I do this a lot. I can think of a lot of puzzles in my maps that kill the player upon failure. I think it’s fine as long as there’s an autosave right before the puzzle.

    3. Yep

    2. Yep

    1. I’ve sometimes violated this rule. For instance the see-saw puzzle with cinderblocks, the orange gate wheel thing you talked about, and various others. I don’t think it’s necessarily always a bad idea to copy people’s puzzle ideas, but new puzzles are generally a better choice. The problem, of course, is that new puzzles are really hard to come up with.

    Also, the puzzle I am probably proudest of in my own maps is in Redox. It hasn’t got much credit in the reviews of that map, though. It’s the one with a deep hole and two buttons. One button causes the hole to be temporarily plugged with this gate thing. (but the gate thing is too low for the player to use it to jump to the other side) And the other button causes a big box to fall from the ceiling. The player immediately knows that the goal of the puzzle is to get to the other side of the big hole in the ground. The player doesn’t immediately know what the buttons do, but he can try out the buttons and find out. There’s no cost to trying the buttons, they can be re-used however many times the player feels like pressing them. Now the player knows the goal of the puzzle and the tools with which he’s supposed to achieve that goal. Now he just has to think a little bit to solve it. Bam.

    1. Jim Partridge

      on your example.. yep. I’d agree that’s a perfect puzzle!

      1. Jim Partridge

        But I’d disagree that new puzzles are hard to come up with. A little bit of thought and a bit of lateral thinking is all it takes.

        1. When you’ve already released 12 maps that all have a bunch of puzzles in them, it does get difficult to think of new ones.

    2. Kyo

      I don’t think it’s a bad idea to take an existing puzzle idea and use it either. Because often times, you’ll find in the practice of using it that it can slot into the level in a different way from its source, or that there’s some layer of complication you can add to it that complements both your level and the puzzle.

      That’s why I wouldn’t be so quick to veto a puzzle just because it’s been done before. The point is, you may not understand how it’s going to work until you start building it and the idea dawns on you.

  4. andyb

    I dont think its fair to use portal in the example Phillip.Its a different game altogether from HL.Fall in the water you you die.Thats what the signs tell you so it should be common sense to take extra care on that particular part.Same with laser grids and anything else deadly like turrets.You dont just go waltzing in and expect to live.Most mappers have lots of saves and for myself I use quick save like there is no tomorrow on some portal maps.Harder the more I use it.The more I use it also gives me an idea of what to rate a map at.

  5. Derbler

    After watching his Cube playthrough, I think William may agree with many of your points.

Leave a Reply

Comment Formatting Guide

Well formatted comments are much easier to read. Please copy and paste the HTML Tags to use in your comment

  • HEADER: <div class="fix"></div><div class="sbe3">TEXT HERE</div>
  • BOLD: <strong>TEXT HERE</strong>
  • ITALIC: <em>TEXT HERE</em>
  • SPOILER: <span class="spoiler">TEXT HERE</span>