 |
WYSIWYG lootOctober 27th, 2006 |
It’s been a while since I did a straight-up design topic, and both Sara Jensen (at her new blog!) and Brian Green jumped in to reply to Ryan Shwayder’s original post on the subject, so why not perpetuate it?
Basically, the issue is this: when you kill some dude standing around in pink tights, a floppy hat, and elfin chain mail, do you get the pink tights, floppy hat, and chain mail? Or do you get something else, if anything?
The arguments against are pretty straightforward:
- You’re going to generate a lot of items if you have WYSIWYG loot.
- This also means a surplus of items nobody wants.
- It complicates the loot tables.
- It also crashes the value of this type of item on the game market.
I tend to approach this from a simulationist point of view. The whole notion of “loot tables” bugs me, to be honest, because I am not really a fan of the whole item-centric approach. When I approach creating a system to populate a town with NPCs, I see two classes of NPC:
- Special ones, with personalities and quests and some role to play.
- Spear-carriers. In fact, most NPCs everywhere are spear-carriers. Most everything you kill is a spear-carrier.
Special NPCs
Special NPCs have tended to get custom models or special outfits. This is because you want them to stand out — they are special, serve as entry points for quests or as bosses to signal player mastery. And in fact, special outfits with WYSIWYG loot has a long tradition from the MUD days, because then if you kill the guy, you can keep the custom item that the guy had as proof. It serves basically the exact same role as special items dropped from loot tables: the role of trophy. It’s something everyone recognizes: “oh, you have the feathered hat that Falstaff wears! That means you killed Falstaff. And that means you’re a bad ass.” Bottom line, WYSIWYG loot is just about always desirable on the bosses for a very “gamey” reason: the only reason to have special named items is as trophies anyway. So the more signaling you give the trophy, the better.
If you have gone the custom model route, having niceties like attach points and equipment slots on your custom model is actually extra work. So you may hack it by putting the Flambe Sword in their hand in the model and then making a separate one to spawn in loot. This also means that you probably don’t make the feathered hat. This is a shame, because the feathered hat has more and greater potential for play than the Flambe Sword does; you’re going to outlevel the Flame Sword in another two levels anyway, whereas the feathered hat can serve as an object of conversation throughout your entire play career. By tying the trophy too tightly to game mechanics, you actually limit its utility to the player in the long term. (This is why WoW devolves into a chase for the next trophy).
In games where even spear-carriers are promoted to special status, by features such as more varied NPC conversation systems, capabilities to give rumors, persistence (such as persistent randomly-named but killable shopkeepers, like UO had), or even procedural quest generation, then you end up wanting to make even your spear-carriers have these characteristics. If Falstaff happens to be rolled up, has something memorable attached to him, and hangs around in the bad part of town for six weeks until he is killed and never returns, that can be of real value. His hat can still be a trophy.
Spear-carriers
You can have spear-carriers all look the same — and in fact, this is the default case, really. They exist in order to make the area feel populated and in order to provide endless fodder (I remember when we were trying to tune the spawn populations of the area outside Mos Eisley in SWG; John Smedley kept asking for more critters and NPCs in the “newbie yards,” until you could peer around the outskirts of town and see dozens of hopping things. His goal was, of course, to provide plenty to kill right off the bat. But it did look rather incongruous).
It saves work to have spear-carriers and special NPCs share the same underlying body models, and be distinguished by clothing and other forms of parametric customization. (You can still make a custom model for when you really need one, of course). Your special NPCs get a table of equipment that includes their special items and whatever other generic items you need in order to make them look right. Your spear-carriers can get generic outfits assembled from whatever clothing you have available for the body type.
The advantage to his is that your spear carriers look varied. The disadvantage is that your spear carriers look varied. Players cue off of special-looking NPCs, so when you see the pink-and-purple fat dude, you think that he means something, when he was just rolled up randomly (in the bad part of town, where he doesn’t fit in). These days, when we cue players with giant glowing icons over NPC heads, this is perhaps less of an issue from a player direction point of view, though still from an aesthetic point of view.
Either way, though, once you are developing in this fashion, WYSIWYG loot happens by default; you have to implement special-case code to not have it.
There are significant tradeoffs to developing this way. Take kobolds in WoW. Sure, you could end up with immensely more varied kobolds, but who cares? It’s not like the kobolds form that large a part of the overall experience. Any time spent on making kobolds customizable NPCs with attach points and morphs and whatnot is time that could have gone into making different monsters altogether. On the flip side, if you spend a lot of time with the kobolds, it’s extremely apparent that they are cookie cutter. A little algorithmic variation would go a long way towards making the process of killing 45 of them less tedious.
Basically, though, if your spear-carriers are essentially consumables, then it makes little sense to do more than the most cursory algorithmic variation. If they are intended to persist to any degree, then we have a very different situation.
The other tradeoff is technical; the algorithmic variation, particularly if it’s visual variety, carries a heavy penalty in terms of client-level performance. It chews video memory and it requires a lot more packets to be sent down. This can result in client-side lag.
Piling up junk
The issues with accumulations of junk largely happen for two reasons:
- The stuff that is dropped is not itself randomly customized, and therefore has zero value.
- The flow of “getting something, evaluating whether you want it, and then discarding it” leaves the item in a place where it’s harder to garbage collect it. (For example, pushing players to “get all” and sort it out later means that you’re more likely to end up with junk laying around than if you force players to pick and choose what they take from the corpse, then decay the corpse and all contents afterwards).
The former case can easily be illustrated by the ways in which these things worked out in UO and SWG. The famous green cloth that Janey always pursued in UO was the result of one of these random customization spawns: a particular NPC happened to randomly get a shade of green dye that wasn’t necessarily easily available. People chased after NPCs with particular colors of clothing in SWG because they wanted it for their own customization (in fact, there’s an additional side effect there, of people “killing for sneakers” so to speak). Both of these are examples of further detail in the simulation creating value for the players in what might have been useless throwaway loot. (Obviously, the majority of what is generated is still useless to most people, and has little market value).
Now, if your userbase is chasing primarily functional value, rather than aesthetic value, then all these drops are effectively creating factories of these items. You should, as Brian points out, expect the market value of these undifferentiated items to fall through the floor. Anything that is dropped regularly by a spear carrier becomes a commodity item with little market value (just like the in-game currency does). Needless to say, a game economy without wear and tear and decay will not support this well at all, and therefore WYSIWYG drops should only go in games with item decay.
Some final thoughts
WYSIWYG loot really only fits in games with certain assumptions:
- That how equipment looks can be as important as how it performs from a mechanical point of view.
- Games which assume a robust enough economic model will be present: decay and floating item prices.
- Finally, an assumption that spear-carriers are not solely consumables, and that NPCs are part of the persisted world state.
In other words, if you want WYSIWYG loot, you need to be looking in worldy games. Gamey worlds have a host of design characteristics which encourage moving away from this particular feature, and they are good reasons. Looked at from an extremely mechanical and reductionist point of view, the classic Diku-style gamey world is about popping many identical XP bags to get loot. What drops and what the bag looks like are largely irrelevant, and only the most basic veneer of plausibility needs to exist. In a worldy game, you want every bag to be different, and you want there to be a reason to pop it, and you don’t always get the same reward; it’s about organic consistency rather than mechanical consistency.
You can follow any responses to this entry through the RSS 2.0 feed. Responses are currently closed, but you can trackback from your own site.
If you show it to any individual player, someone in the playerbase as a whole will want it, period. There’s nothing “unreasonable” to want — we often wish it were so, from the design side, but if we dangled the carrot, then it’s reasonable to reach for it.
Again I’m a little late on this (one day I’ll make posts in a timely manner), but, last Friday, there were several blog discussions stemming from this post on Nerfbat about ‘WYSIWYG’ Loot.
Over at Raph’s Place today, Raph posted some thoughts about WYSIWYG loot. Basically, the question is this: If you see a mob with a spear, leather jerkin and sandals, why are you not able to loot that spear, leather jerkin and sandals when you kill it? And, if you could, in what game could you best do this?
[...] Comments [...]
[...] So I ask you, given the history of SOE in throwing away progress, i.e. Raph Koster’s original design, CURB, and the CU, how can you possibly think logic and SOE have any relationship to one another? They don’t! They are complete strangers to one another. Wherever the Ouija board points is where Smedley and Gang take SWG. Please stop giving them credit for either caring, or thinking in advance. They don’t know how to do either. http://www.raphkoster.com/2006/10/27/wysiwyg-loot/#commentsRaph said on October 27th, 2006 at 2:40 pm:….I’ve never heard the term “bio-linking.” I assume you mean WYSIWYG loot (which is also an ersatz term that Ryan coined). But maybe you mean soulbinding. [...]
[...] Realistic loot (and inventories) Submitted by Abalieno on October 30, 2006 – 16:22. I read what Raph wrote about WYSIWYG loot and I cannot avoid to criticize some parts. [...]
[...] As those of you who follow the blogging scene have noticed, WYSIWYG Loot is the newest hot topic in our space. It all started when Ryan Shwayder wrote his “MMO Rant #2: WYSIWYGn’t Loot”, which has gained the attention of some other developers to also post their reprisals and to show why and how come this type of thinking is not being put in to our games. Brian Green, the same developer that I will take on his challenges to provide ‘new ideals’ posts his “Useless features” article. Sara Jensen then replies to the article written by Biran Green with her “Why Mobs Shouldn’t Drop Their Equipment” and finally Raph Koster then puts in his thoughts over at his site with an article titled “WYSIWYG loot”. [...]
[...] The MMO-dev blog topic of the moment is WYSIWYG loot. This one started at Nerfbat, and spread over time to Psychochild, Raph, Sara Jensen, Darniaq, and probably a dozen other places I haven’t found yet. [...]
[...] http://www.raphkoster.com/2006/10/27/wysiwyg-loot/#more-777WYSIWYG lootIts been a while since I did a straight-up design topic, and both Sara Jensen (at her new blog!) and Brian Green jumped in to reply to Ryan Shwayders original post on the subject, so why not perpetuate it?Basically, the issue is this: when you kill some dude standing around in pink tights, a floppy hat, and elfin chain mail, do you get the pink tights, floppy hat, and chain mail? Or do you get something else, if anything? [...]
[...] Raph has an interesting post up that made me think.In most games of all types, the enemies are faceless bundles of statistics, save for bosses. Similarly, the player upgrades by buying the next generic bundle of statistics, whether it’s a new gun or a nifty wizard cloak.I’ve played every Elder Scrolls game that’s been made. You know what the biggest rush in those games is for me? The slew of random crap you pull off of every corpse and out of every treasure chest. The idea that you’ll find a dagger with a few new capabilities, or a left bracer with a funky new graphic, or a loincloth of burning.When I play those games, I’m totally addicted for ten, fifteen hours. Then I lose all interest. Why? Because the loot doesn’t hold up. By the end of those hours, I’ve explored everything the game has to offer, at least as far as I’m concerned. The first weapon I find that does blammies when I stab someone? What an incredible rush. The next one? Not so much. The seventh one? Not at all.It devolves: the only thing that interests me is the graphics associated with the items. A fruit-filled hat is worth more than the flamey sword of Nimbulus, because the flamey sword is just an extra +5 to something… but the bananahat is unique. Plus, I can enchant it to be a fruity hat of flaming, if I really need the flame.Of course, a steady progression of “unique” is required there, too. The bananahat only holds my attention for so long before I must move on to the next hat.Actually, that’s a failure on the designer’s part. The problem is that there isn’t really enough feedback to prolong my joy in my bananahat. If everyone commented on my bananahat and changed their interactions with me in some interesting way, the bananahat would become extremely interesting to me. Also, generally speaking, there isn’t much in-world feedback.You might be able to see yourself, but it’s a rear view and the costumes aren’t generally very interesting from the back. Notice that the new “custom-avatar chats” always show your character from the front, even when they’re full 3D? Yeah, fronts are better in terms of feedback.Worse, the costumes themselves leave only a small mark on the screen, especially in Elder Scrolls games. World of Warcraft got this right: the costumes are extremely loud and large, totally dominating your character’s appearance. Of course, there’s the problem that you have fewer pieces to play with, and that’s a big drawback…Moreover, there’s only so much joy you can get from permutations on the same stock. No matter how many hats I wear, they all go on top of the same head, with the same art style and the same model. The base gets boring, even if the hats don’t, and that drags the hats down. Don’t get the hats down!This is true even in games like SecondLife. It doesn’t matter that there are 50,000 different kinds of “hats” and more coming out every day. The stock beneath is the same, so they stop being interesting after a while. Thus the thriving business in morphing your avatar: you can’t really wear clothes, but in changing the baseline you have changed your whole… um… baseline.Okay, as per my recent unfortunate habit, I’ve started to ramble. What I’m saying is:Manufactured or unique is the wrong question to ask. Randomly generating 500,000 different kinds of sword will only broaden the game so much. In the beginning, it’ll be awesome, but by midgame, you’ll be just as bored of the random swords as you would be of 100 carefully scripted, balanced swords. You’ll know the parameters. Random generation is really a “wide” solution rather than a “deep” solution, and unless you plan on absurdly restricted access to randomly generated things, it’s not going to add play depth for anyone other than newbs.Subtracting out the gameplay elements actually deepens the play, because now the system follows supply and demand. Nobody cares that there’s only three blue swords of cystic fibrosis, because they’re worse than the ten thousand red swords of blammifying. But if all swords are equal, the rarity of those blue swords makes them incredibly valuable. The same idea applies for hats.The feedback you get on your non-combat-related equipment is pretty strong in a MMORPG, although exceedingly weak in a one-player game. This means that you don’t require as much depth in a MMORPG, because feedback will create more depth. In a one-player game, you’ll need to go further. Much further.For example, being able to dress a whole roster of characters in whatever fashions you prefer. Again: linking these things to play bonuses is basically a bad idea, because it dramatically limits the player’s options.Another idea is to be able to change your avatar, either piece by piece or in whole. You could pull a Shiny trick from Messiah: let the player inhabit whatever randomly generated NPC they can lure into a dark corner alone. NPCs can have some immediate gameplay results (such as being better warriors, or having access to certain places), but in the long run have fundamentally interchangeable capabilities. NPCs should look dramatically varied – it might be best to use animal-people, since they look very different from each other. Elves vs dwarves is about the minimum.This would allow the player to grab an avatar, equip it, and run around. If he or she wants, he or she can jump into a new NPC – one that looks very different and people react to in very different ways.This allows them to change the baseline and all the stuff on top. That’s cool. I think that would be a fun game, either one-player or massively multiplayer. Imagine the economy that would spring up in body sales. Some NPCs are extremely hard to get because they are always surrounded by people, and those call in the highest prices.Obviously, there would need to be some, I dunno, GAME involved at some point. But, pshaw, that’s the easy part.Labels: balance, game design, long-term play, loot [...]