On Thursday, June 7, 2012 11:46:27 PM UTC+1, Michal Bielinski wrote:
> Once upon a time, Darren Grey wrote thus:
> > Well, I'm guessing this'll spark a number of disagreements, but I'm
> > curious as to what other people have to think or contribute to this.
> > I've written this article today and would like to polish it a bit
> > more before putting it on RogueBasin. But reflecting on this is as
> > much for my own benefit as it is intended to be a guide for others.
> Very well. I will say up front I disagree with much of what you wrote.
> Pretty please do something about this wretched formatting. It extends
> the time I have to devote to writing reply because I have to fix it
Sorry sorry :( You don't disagree with all that much I must say.
Plus open debate is all well and good :)
> > For some this article is irrelevant - they only care about existing
> > Roguelike players touching their games.
> What a generalisation. Are there *any* developers out there like this?
Krice ;) Plus there is an attitude amongst some of "If you don't
bother to learn the controls you don't deserve to play my game". I've
seen this in specific response to this article. It's something I
disagree with, but every dev can have their own perogative.
But more than that you have to consider that if you're not going out
of your way to consider accessibility and all you do is follow the
genre's conventions then essentially you are just making games for
existing roguelike players.
> > All fine and well, but I’d like to see more people enjoying all the
> > great games we have around, and in particular I have a goal of making
> > the first truly accessible ASCII game. Er, wish me luck :/
> Define truly accessible. I fear this might require the game not to be
> a roguelike. Interface is not everything to accessibility.
"Pick up and play" is what I define as truly accessible. Mechanics will
take some learning, but the controls will be automatic and intuitive.
> > One problem though – the interface. I’ve watched hardcore gamers try
> > out Broken Bottle (which is considered to have a good interface for
> > a Roguelike) and they’ve not been able to figure out how to move, or
> > how to leave the first level.
> I presume hardcore gamers means playing games a lot. Not necessarily
> playing difficult games. Otherwise where would you find such people?
> I have introduced PRIME to a bunch of university students. Some are
> hardcore gamers but most are not. Walking was trivial for all of them.
> Leaving first level was not so easy though. Only a portion of them
> managed this. Still, PRIME does not tell you how to use stairs at all.
Interesting... My hardcore includes other indie devs, and people who
play the likes of Dark Souls gladly. I've not witnessed anyone without
roguelike experience for whom walking was trivial. Plenty managed
basic 4-way just fine, and some picked up 8-way after a while, but it
was never trivial.
And stairs... yeah, Jesus, no one gets stairs :(
> > All the assumptions we have about the interface are completely null
> > to them. It’s immensely painful watching someone try to play your
> > game and being so unfamiliar with the controls that they move at a
> > snail’s pace, and have such trouble with even movement that they
> > fail to see the rest of the depth of play.
> Agreed with this. On the other hand my experience is different. Maybe
> it is caused by my audience? I show roguelikes to students mostly.
> After all Rogue also originated in a university. Most troubles people
> had was with searching (when that was only exit from a room) and
> leaving the level. They quickly got to experiencing the game more.
What sort of students? I don't know why your experience has been
better. Oh, and was it on a normal computer or a laptop?
> > 4-way movement
> > Ooh, I can hear the traditionalists grinding their teeth already…
> I am no traditionalist. I simply dislike pointless constraints.
If you view it as a constraint... And importantly, how much of a
constraint? Does it really make that much difference?
A lot of people moaned about Dredmor only having 4-way movement.
It was... well, stupid. What would diagonal movement add to the game?
The game was designed around 4-way movement and worked just fine like
that. The people arguing were seeing it as a constraint rather than
accepting it as part of the game design.
I'm not a big fan of 4-way, but I think it's important not to dismiss
it out of hand. It's an alternative design that produces different
gameplay and design consideration, nothing more or less. Make the
game content compelling and the game options more interesting than
choosing between north and north-west movement.
Perhaps one option for making 8-way clear is to have a red outline
or shading on all available moves, a bit like how various tactical
games use when moving units. Would have to be something you could
turn off, of course...
> > Years of d-pad or arrow key use has made them think purely of the
> > cardinal directions. Also it’s hard to get 8-way movement to work
> > nicely on a laptop, whatever crazy shift+direction scheme you might
> > invent (and don’t even *think* about forcing vi-keys on players).
> > WASD and arrow keys are all outside gamers can cope with.
> Among my audience there was a vim user. Guess how we solved problem
> of his laptop lacking numeric keyboard? That was a nice surprise for
> him. He was amongst the worst score-wise though.
Well, you clearly had an unnatural group then ;)
> > An alternative is hex, with QWE/ASD for the 6 directions. I’ll be
> > trying this with Rogue Rage, but I’m not sure yet how well it’ll
> > work. But from a technical point of view hex is the ultimate for
> > all games ;)
> Bah. Go 8-way movement with maximum metric. You eliminate all and
> every sight/movement artifact you can think of with this.
You absolutely do not! How often have there been stupid threads on
here about making diagonal movement cost 1.4 turns? And the Square
LOS considerations to boot. 8-way doesn't make geometric sense.
But it's a game, so that's fine. But ah, hex! How gloriously
elegant all systems are in hex! I spit on your inferior 8-way :P
> > Mouse control
> > Easily and quickly move to another part of the map (with appropriate
> > interruptions for damage or spotting a threat),
> A partial auto-explore? Existence of a "need" for such mechanics shows
> a design flaw. You should make levels interesting and dangerous enough
> to explore manually. Also, you should not make your players travel
> across explored levels. Ironman is amongst easiest and most successful
> ways to accomplish this.
It's not meant as a partial auto-explore, but an easy way to move
several spaces with a single command, letting the computer calculate
the optimal route inbetween. And even apart from that you can have
clicking turn by turn to give an alternative input mechanism. It
helps with the laptop problem a little bit (though not much since
trackpads also suck).
> > and right-click for
> > contextual menus to quickly choose available interactions with
> > enemies and items.
> Nah, it would obscure the screen too much (if one would like to display
> really all the possibilities a roguelike game hides) or the font would
> be too small.
I don't get what you mean by obscure the screen? I mean context menus
like in ToME, where right-clicking on an enemies pops up a list of
available actions. Consider in a game being able to go into your
inventory, right click on an arrow pile, choose "equip and fire" and it
then equipping them and setting you straight into targetting mode to
click on an enemy.
If the list gets too long make it scrollable. Makes the default
action appear at the top (attack enemy, eat food, read scroll, etc).
Maybe left-click or double click would result in using that default
There's a lot that can be done with mouse commands.
> > Mouse control makes a game hugely more accessible without hindering
> > the game’s complexity.
> If you had omitted the "hugely" word I would be inclined to agree. As it
> stands your statement appears to be false.
Everyone knows how to use the mouse. There's very little interface
teaching involved. There are also easy standards to be picked up from
other games with mouse usage. The fact that few roguelikes use mouse
is in a way an advantage - there are less traditions from the genre to
be followed when implementing them.
Of course mouse useage should ideally not be required. There are many
advantages to keyboard only control that it would be a shame to lose.
> > ToME4 is a complex game that can be comfortably played with either
> > mouse or keyboard, or a combination.
> Oh, it can be played with keyboard? So how the mouse is actually even
> needed? I'll quit playing straw man here. You simply give too much value
> to mouse input. It is nice and all but in no way hugely improves any game.
Playing with just the mouse in ToME is a dream. There are all sorts of
great customisations you can implement, like binding your most used spell
or ability to left click, so clicking on an enemy triggers that when
available (but defaults to attack or move towards otherwise). It speeds
up a lot of play. There are also mouse gestures configurable for favoured
commands. Plus you can play the whole game with one hand, not even
needing to move that hand much. Optimal play for me is a mixture of both
keyboard and mouse (numpad is nicer for movement) but the interface is
vastly improved by mouse control.
Same for Dredmor. Aiming ranged attacks is so much easier with the mouse.
So is moving items between inventory, dungeon and equipment window. And
tooltips help a lot with learning details of items, skills and enemies,
without having to go to a separate screen all the time or memorise a lot
> > Tooltips also help convey detailed information quickly without forcing
> > the player to read through manuals or help files – most players never
> > read help files or manuals.
> Their choice I'd say. I have read whole NetHack guidebook before even
> running the game for the very first time. Did you know they tell you
> about E-word there? Still, on rgr.nethack it is still considered somewhat
> spoily proving your point. Everything one places in manual and readme
> should be found in game. Documentation just exposes information in more
> friendly form.
I read the whole ADOM manual before ever playing the game. It was one of
the things that drew me in. I'm not sure I'd do the same for many games
these days though. Why should I read through the manual if I don't know
if I'll enjoy the game yet? Good documentation is still important of
course, but don't rely on all players making use of it. Modern gamers
especially are used to forced tutorials (ugh!) and hand-holding in new
games. They are conditioned not to read such materials, or to even expect
them to exist.
> > Auto stairs use
> > A stairs should be treated like a wall square you can see through.
> > Enemies can’t walk on it, items can’t drop on it, and once you bump
> > into it you change level.
> The no-walk and no-item-drop are additional properties that ordinary
> walls do not have. So it is much more than a transparent wall.
Well, depends on the game, but true. I wasn't suggesting it be an
ordinary wall of course, just to rethink the idea that it should be
like a floor tile. Why is a hole in the ground or a rising stairwell
treated like as just a common flat space with an added escape element?
It makes just as much sense (if not more) for it to impossible to
walk over. It doesn't negatively impact on the game rules, though of
course due consideration in design must be made.
Put another way, if Rogue had done it like this then every roguelike
since would have still copied it.
> > without removing any tactical depth. Means stairs need space around
> > them (can’t be in corridors) or need to be in-set in the walls (see
> > Dungeons of Dredmor, where stairs use is made obvious to the player),
> > and player should be placed adjacent to the stair tile when put in
> > a new level. Makes a big difference to new players when they don’t even
> > have to think about using the stairs, it’s just natural. Moving into
> > stairs to use them should be as natural as bumping into doors to open
> > them or moving into monsters to attack them.
> Acceptable solution but requires a substantial game redesign for most
> existing roguelikes.
Really? Make sure stairs can be pathed around is the main
consideration, and that's not hard. Player placement on level change
is another, but many roguelikes already have this with monster following.
It's a pretty minor redesign from my perspective.
> > Tell a player to press < and he’ll press the , key, and then wonder
> > why it didn’t work.
> The don't know how to use the keyboard properly? I don't buy it.
Maybe I should change to title to "How to Make Roguelikes Playable for
Idiots" ;) Honestly, I have seen this! Multiple times!
> > In a reduced command set there’s no reason to use upper cases or
> > keys on the shift line.
> Reduced set of commands that are necessary to play and win is fine.
> Reduced set of all commands by itself just impoverishes the input
> channel of your game. For what? Hide all advanced commands but make
> them available. Nobody says OpenOffice writer should remove its
> shortcut keys. Or are you?
Hmm, valid point - advanced commands for experienced users can be
very handy. I guess I should say all core commands instead. And help
files of commands should split the core commands and advanced commands
to make it clear to new users that you can play with just the base
set (unlike ADOM's list of commands which is almost unusable).
> > Besides, on international keyboards these can be in different places,
> > so might screw up your input in other regions. Stick to simple lower
> > case letters or symbols easily accessible. And don’t over-rely on
> > mnemonics – those who aren’t native English may struggle to understand
> > q for quaff. Better to have commands grouped together in one part of
> > the keyboard (but at a distance from the movement keys).
> Remappable keys take care of most of that. PRIME does a small experiment.
> It offers several key maps based on other commonly played roguelike games.
> Sadly I have not collected feedback from players how helpful (or not)
> it is so cannot comment on whether it is a good thing to do.
I think it's a good thing, just be mindful that 90% of players will
likely stick with the default. Another possibility is duplicate keys -
having both g and p for get/pick up for instance. I don't like it much
as a designer, but it can be handy when there are multiple standards
around. The classic use of this is having both vi-keys and numpad as
available move inputs.
> > Use standard controls from computer RPGs
> > I for inventory is the big obvious one. C for character is also
> > oft-used. Generally the less you have to use the better, but when you
> > want to use keys think of what are the standards outside the genre as
> > well. H for help is probably more standard than ? for instance.
> Sure it is. But that sacrifices H! I would wager F1 is even more widely
> used so bind help to this and keep H for something that deserves it.
Like vi keys? :) Yeah, F1 may be better. I'm not sure what the
standard is these days in mainsteam games.
> > W for Wear will confuse many roguelikers,
> Perhaps you wanted to say non-roguelikers? Very few RLs have W bound to
> something else than wield or wear.
No, I meant roguelikers. With a background in ADOM I am always
flummuxed by games with Nethack-esque inventory systems.
> > Pop-up text hints
> > Both for interface and for gameplay, though you might want an option
> > to turn it off. Especially at the start of the game you should have
> > some text briefly showing the commands. It could be as easy as arrows
> > around the character showing the move key for each direction as a
> > quick lesson/reminder,
> Oh, something like Abura Tan had and newest DoomRL does? Several other
> roguelike also implemented this but somehow it never caught on. I think
> tutorial is much better.
Tutorial is more needed for a bigger game, I think. But the optional or
one-off hints are better for smaller games. Works especially well in
7DRLs where you're not sure what defaults the designer is working from.
> > But during gameplay this can be useful too, especially when
> > introducing new features or content. If you don’t have auto-stairs
> > use then walking over the stairs might provide a message “Stairs down
> > to Dungeon level 3, press space to descend”.
> Yes, that is very nice. Here an example from game could be useful.
> I seem to recall messages like this being printed in some game but
> cannot exactly tell where it was.
I'm sure I've seen some 7DRLs with this... but none specifically
spring to mind. I find the beauty of a system like this is that you
don't even notice or remember it much - it's just a natural background
part of the game that doesn't take away from the immersion.
You have a good point that examples would make this whole article
> > Similar for walking over items (how to pick up), or when picking up
> > items (how to see inventory or equip). Seasoned players might not even
> > notice these messages when they’re used to seeing them, but new players
> > will really appreciate this help.
> Be very wary with the part about not noticing. Hints in DoomRL almost
> made me quit in frustration. In my opinion information how to turn hints
> off should be *the* first hint. Make that doable from the game without
> closing it and editing some config file.
Yes, good point. Partly it's down to how verbose and intrusive the
messages are. Pop-up boxes with an "OK" confirmation are the worst thing
possible :( But saying in the log "Your clip is empty (R to reload)" is
> > With ASCII it’s hard to tell what new enemies are, or how powerful
> > they might be.
> This is good. Keep it that way! Tiles are horrible in this regard
> because not only they make it hard to tell creature's power rating
> but may also confuse you! "This looks like those chickens I killed by
> dozens already. Lets whack it! <The cockatrice hisses at you!> ...".
Haha, yes. Plus tiles are generally fugly. Some considerations for
- Make weaker enemies or ones with no special abilities lower case
- Make very special enemies symbols (& for intance)
- Restrain use of the colour palette, with particularly weird colours (magenta, cyan) used for rare and intersting creatures
- Gives bosses or superpowerful enemies a background colour
> > So how about first time an enemy type shows a piece of text appears
> > to say “Crikey, a raging bulbosaurus!” or something similar. Bosses
> > might have more dramatic description to denote their importance.
> DCSS introduces all monsters by name when you meet them. What you say
> was being done in DeadCold and was being annoying from the second
> character you played. I recommend against going much more beyond monster
It should be stored across games so you never see the same notification
Of course tooltips make such a system irrelevant.
> > So let’s use the biggest key on the board for it. Stairs? Space to
> > change level. Item on ground? Space to pick up. Trap? Space to disarm.
> > Door? Space to open/close. There’s situations where this won’t work,
> Ha! The problem is those situations are far too numerous to list. If they
> are not you run into even worse problem. Take DoomRL's doors. Close
> command usually targets the only door in your vicinity. Only when there
> are more targets to choose. If that happens in a critical situation your
> muscle reflexes may be faster than your brain noticing a sudden query.
> You have closed like two thousand doors already, why now you need to
> point at it? I lost a good character that way.
Hmm, muscle memory is a valid point. I've had similar problems in ADOM
but never a death from it. And I always view it as my fault. Plus I
remember the pain of having to always choose which damn door in earlier
versions. I'll gladly take the context over the rigmarole of specifying
obvious details in every circumstance.
> > but the exceptions shouldn’t get in the way of what would be a much
> > smoother experience for most of the game.
> They should. By ignoring those you get seemingly better interface but
> in reality it becomes ridden with traps.
I disagree. But I guess that's where options come in :) ADOM has its
auto-select doors as an option you can disable. I'm not aware of any
that have it disabled though...
> > Don’t rely on the game log
> What I feared most. The game log is integral part of roguelike spirit.
> Eliminating it like Cogmind does significantly changes the feel of the
> game. I view roguelike games as something closer to the book than to
> the movie. In the log lies power to describe events in game in a way
> that invites you to imagine them. Make the log less important and you
> have less of a roguelike in result.
I didn't say dismiss, I said don't rely. Of course if you have a
*good* game log it's not as much of a problem. Part of making the log
good is getting rid of the chaff like the hit and miss notifications.
Restrict the log to important things and it becomes more noticed.
Use contextual and quick feedback for the common things that tend to
fill up the log with repetitive lines. The log should not be used to
communicate every petty detail to the player.
> > Easy Inventory system
> Ouch. All games I work on are guilty of not fulfilling this.
How very true! But roguelikes tend to be especially bad :)
> > Don’t give the player some weird custom inventory with containers
> > and different commands for each slot.
> Yeah! Take out the fun! Also, take out the ability to play at blazing
> speeds. Have you seen a veteran NetHack player striding through the
> dungeons? He rarely invokes the inventory because he remembers places
> of several of his most important tools. Since inventory does not come
> up too often suspension of disbelief is maintained better.
> ADOM does not assign constant letters to inventory items and it fares
> well. However, TB does not exile them. Picking item by letter is a very
> fast way to make a choice.
The article is about attracting new people, not catering for the old.
Of course when it comes to proper design you have to consider how you
want to balance between the two. ADOM has a nice balance in its
inventory system in my opinion. Nethack is just unplayable to me.
Some people may be used to it and enjoy it, but I have lots of other
good roguelikes I can play instead of forcing myself through the
pain of getting used to its system.
> > Don’t limit their space too much either. Players want freedom and
> > ease of use. Look at how professional games do inventories.
> Or go read some weblog about game development. Much faster and gives
> more insights. Also, does not require computer with good hardware.
Well, I said look at, not play. Youtube works :)
> > Clarity of mechanics
> > Don’t hide important stuff from the player and hope that they’ll get
> > it. Make sure your mechanics are clear and understandable. You may
> > have coded a cool system, but if people don’t know what’s going on
> > they won’t understand the subtleties and won’t stick around long
> > enough to enjoy it.
> I think NetHack remains among few offenders. Nowadays roguelikes usually
> have transparent mechanics.
Yes, but it bears reminding :) I've played many 7DRLs with obscure
mechanics, which seems the absolute worst place to have them since few
will replay the game often. The biggest offender is in item mechanics,
with weapons and armour not showing their power and special items not
saying at all what they do.
> > This is where watching someone play your game is important. Without
> > giving them advice watch how they interact with the game elements and
> > see how easily they grasp what is going on. Don’t think you can get
> > by on the Nethack style of “players will learn over many years” – you
> > are not making Nethack, nor should you be.
> I have been lucky to have someone do Let's Play videos of ZapM which is
> PRIME's parent. Also, the people I have gotten to play it are even better
> because I could ask "why did you do that?" and receive answer immediately.
Yeah, these things help a lot. Watching over someone's shoulder is best
I find. I had Broken Bottle on show at an Indie Games Expo last weekend,
which was especially revealing since it had people in who didn't exactly
choose to play the game (someone in a Let's Play is bound to keep on at
a game for a few minutes rather than quit in frustration, after all).
> > No cheap deaths
> > By cheap I mean impossible to prepare for. This is especially
> > frustrating early game. Challenge is good and fun, but unavoidable
> > impossible obstacles are not.
> This paragraph is completely out of place in this document. Not only
> this applies equally well to roguelikers and non-roguelikers but the
> former will be able to point out unfair deaths much more accurately
> than the latter.
True. I should have specificed "early game deaths" more. A cheap
game death can put off a new player permanently. Roguelikers are a
bit more used to this and willing to put in more persistance.
> > Traps are the worst for this. Don’t do traps unless they’re
> > interesting.
> Saldy traps are difficult to do right. Either they are harmless and
> just annoy the player or too powerful and result in "there was nothing
> I could have done to prevent this" moments.
Yep! Only game with compelling traps is Dredmor in my experience,
since you can use them as a resource against enemies.
> > Quick restarting
> > Make it easy for the player to start again, such as a “Restart Same
> > Character” option on the death screen that repeats the character
> > creation with the same options as last time. “Restart Random” is
> > a good option to have too. Generally your creation menu should be
> > slick and fast anyway, otherwise people will get bored of it on the
> > 100th playthrough.
> This also means no quitting after death obviously.
Heh, that's one way to keep a player hooked.
> > Small levels
> > Can help quite a bit, though isn’t necessary. Don’t waste the
> > players time exploring big empty spaces. Have everything densely
> > packed in small levels. Makes it easier on display too, reducing
> > the needs for a minimap.
> I would say large levels densely packed with content and features are
> much more interesting tactically while being very fun to explore.
> The key here in my opinion is not small levels. It is densely packed
> levels. Size is mostly a non-issue.
Both, I'd say. In respect of new players I think small levels lets
them get to new floors quicker, giving a sense of progress. It also
doesn't take anything away from the game if done right.
> > 4-way movement helps with this a bit – it makes the areas effectively
> > twice as big for play moves, letting you pack monsters and items
> > in a tighter screen space.
> True only in melee oriented game. In ranged combat heavy games it just
> makes feel you slower. Not very fun.
Oh? I've never felt that with Powder.
> > Permadeath optional
> > Okay, I don’t really like this, but it works well for Dungeons of
> > Dredmor. In it permadeath is on by default, but you have the option
> > of turning it off. Mentally it sets the player thinking that this
> > is the right way for the game to be played, rather than feeling like
> > it’s a crazy design flaw. And they can be proud that they didn’t
> > untick that box, letting them embrace the mechanic rather than be
> > annoyed by it.
> Decker did it too by having ironman box to tick. Not sure if it was
> on by default. Optional permadeath has all the sad consequences of
> save-scumming and can ruin the game for players. I deem it as my
> obligation as a game developer to prevent that because I know how
> my game was designed. It should be made possible but not in a way
> that I make anything to hint it is designed like that. Developers
> ought to tell how is the intended way of playing the game and then
> allow players to play any way they please.
> In Dungeons of Dredmor permadeath feels not like the way the game should
> be played. It feels like equally valid way of playing the game which
> is endorsed by the developers. This is my gripe with making permadeath
> a check box.
Yet most players play on permadeath. The mechanic is embraced. And part
of that embrace is making it clear to the player that it's their choice,
so when they die and have to restart they can't blame the game. Since
it's their choice they are more mentally ready to accept it.
Still, other games like Binding of Isaac get by with straight permadeath,
so perhaps there's more acceptance for the mechanic than I give outside
gamers credit for.
Thanks for your input, lots of good points :)