More/Fewer Options = Greater Diversity?

Which leads to greater character diversity: more options or fewer? There are several moving parts there: the number of choices, the number of options for each choice, the importance of those choices, and the ability to change them. I should note that I will be considering the majority or the people who care; there will always be people in the tails of the bell curve who are trying to make the worst possible choices because of contrariness or a character concept.

The importance factor seems the easiest: the more important a choice is, the less likely there is to be diversity. If something really really matters, there will be one best way to do it or a few best combinations; even if you give me more options for that choice, anything less than the best means gimping the character. Contrarily, if the decision has no practical importance, the distribution will be random, and more options will lead to even greater diversity.

You can see this one easily in evolutionary biology. Do you have attached or detached earlobes? It really does affect your survival, so our species is random there. Can you roll your tongue? Innie or outie?

Take character appearance as an example. This will have few effects on your ability to play, and while some looks are more popular (Hey wow, another CoV villain in black with a skull motif. Think of that one yourself?), someone will like just about anything. In City of Heroes, you have a practically infinite number of options when creating a character, if you include all the possible combinations of body parts, colors, and sliders. You see more diversity there than in YAFMMORPG, where you have some options on a few parts.

Of course, in a heavily itemized YAFMMORPG, those initial appearance choices do not matter, both in the sense that they have no effect and that you will not see them. Your character will be covered in gear, which is where real diversity of appearance comes from. Now appearance is very important because it is a side effect of how effective you are. Now that it is important, there is one best set of gear for your class, so everyone is building towards that. Given enough time, everyone will look the same, and diversity in character appearance comes from who has spent less time farming.

Moving along, all things being equal, more options for each choice should lead to more diversity. Even if there is a central tendency, having a wider range leads to a broader spread. Having some options that are clearly better or worse will decrease the effect of this. Adding more useless crap does not increase diversity much.

Does having more choices increase diversity? This seems like a live option to me, particularly based on whether those choices are from the same pool or not, the range of options in those pools, and the other factors surrounding choices. Let’s get some examples to explore.

You choose one race from x races and one class from y classes. These are what I am calling separate pools: you pick one race and one class. If you can multi-class, you are still making one pick from the race pool, then one or more from the class pool. Then you have your skills, traits, spells, etc., which usually have many picks from many options, possibly from separate pools. If you have mutually exclusive pools, that is effectively another level of pooling, a sub-class.

You get one choice for race, unless your game has multi-racial characters or a system with more variables (kind of like a skills system for race instead of/in addition to class). Does this create the most diversity, or would having more choices? This will probably depend on the number of choices and their importance.

Trivially, being able to pick more than one race would lead to more options. If there are five races for each side in WoW, there are ten bi-racial choices, twenty if order matters. Having tri-racial characters, however, does not give you more options unless order matters: there are still ten combinations. Letting you pick four brings us back to having five combinations, and all five means just one. You increase that a bit if you allow both single-race and multi-racial characters. The same thing applies to classes: Guild Wars gets more diversity from its class options by making you pick two, and order matters there so they get even more.

What I am really interested in here is the question of when you have multiple picks from a pool with multiple picks. This could be skills in Asheron’s Call, powers in City of Heroes, talents in World of Warcraft, or traits in The Lord of the Rings Online™: Shadows of Angmar™. For any of them, order does not matter (much), importance is high, and the number of picks varies widely so we have multiples cases to consider.

First, remember the issue before with combining a small number of races: as the number of picks approaches 0 or all of the possible choices, the number of different options shrinks. If you pick all but one, you have the same number of options as if you pick only one. (If you want to try the math at home, see combinatorics.) The same applies here. If you can take almost everything, you do not get much diversity. Skills produce no diversity in WoW because you can buy all of them, and you expect everyone to do so.

This would seem like a trivial math exercise: with the number of options you have available, figure out which number of picks gives you the most options, done. The Lord of the Rings Online™: Shadows of Angmar™ has 19 virtue traits, so there are 92,378 options if you let people pick 9 or 10.

But wait, are any of them completely useless? Useless to your class? We can eliminate those. Are any of them obviously the best or effectively required for your class? Assume everyone takes those. Now you are no longer picking 9 from 19; if there are four “required” and four “worthless,” you are picking 5 from 11. That is still 462 options: not shabby, but far fewer than the 92,378 it looked like. Anyone want to bet on whether folks will do the math to get a perfect ordering of the traits, so that you really do have a required list for all of them with few to no choices?

You can make it so that there is no “best,” just “best at this.” Unfortunately, one sort of this will probably vastly out-number the others, so you are back to one best. Take EVE Online: your stats help you train faster, and different stats affect different skills. I want to have high Charisma for my trader, right? No, unless you play for a short time with a highly constrained list of skills, it is the least important stat. There are just not enough skills that demand it. Take City of Heroes: you can have the best protection against cold attacks in the game, but you could go days without seeing an enemy that uses cold attacks unless you specifically hunt Nerva demons (or Crey over the range where Cryo Tanks appear).

Even if you do succeed in making a “best at,” you still have different classes. Being the toughest mage in the world means that you are now two-ply toilet paper. Go you. Being the warrior with the most mana is not going to do much for you, either. You are going with best at your role, so the diversity question falls back to how many roles your class design allows. Wow, all that work with classes, skills, and talents, and it all comes down to whether a Fury Warrior out-DPSes a Rogue.

Like many of my posts, this is getting long, so I am going to tie it off with the ability to change choices. Easier changing effectively increases the number of picks, because you change it for the situation at hand. You are no longer picking 9 of 19; you are picking 19 of 19, but only 9 at any given time. Gear is probably an easier way to think of this: you can wear your healing suit or your damage suit, your fire protection or your acid protection. You can switch gear in a moment, though, so unless this fight calls for you to be the primary healer and primary DPS against enemies dealing heavy fire damage and heavy acid damage, there is still going to be one best pick for you. Click a button to change it when the acid-spitting bugs replace the fire-breathing dragons. Now your range of options is limited only by pack space.

How is that working out in whatever game you are playing? What needs to be fixed?

: Zubon

9 thoughts on “More/Fewer Options = Greater Diversity?”

  1. I’m currently playing WoW and have heavily played CoH. There’s one more thing you need to take into account, which is gear. In CoH, that’s enhancements, but it’s still gear.

    Gear can significantly change your choices. As an example, I’ve got a level 41 druid. Up until recently, I was playing him as a straight feral druid, all the way from 10 to 41. Recently, I decided to try out a balance spec. Feral gear is mostly +strength/+stamina/+agility gear, at least before +attack power/+crit/+defense gear becomes regularly available. However, for my balance gear, I need +int/+stamina/+spirit/+spell damage gear. The server I’m on (Sentinels) is relatively young, so finding balance-friendly leather gear on the Auction House for my level took a couple of nights. I still haven’t replaced my belt, shoulder piece, rings or necklace, so that’s still feral. In addition, all of my balance gear is green, whereas a lot of my old feral stuff was blue.

    As it stands, I’m an “okay” balance druid. I’ve got a great talent spec, but not nearly enough mana to fuel it, so I sit and drink a lot. In addition, I don’t have as much spell crit%, since that plays partially off of intellect, so my damage isn’t as high as it could be. However, once I do get all geared up in my Balance gear, with the changing out of one or two pieces of gear (swapping out +spell damage gear for +healing), I’d be a pretty good restoration druid, even without a talent swap, simply because Balance and Restoration work mostly off of the same stats, albeit with usually a little difference in emphasis.

    A similar thing would happen with my protection specced paladin if I turned him into a retribution specced one, or, to a lesser extent, my fury warrior into a protection warrior. Different specs require different gear in most games, and having mismatched gear and spec severely hampers effectiveness.

    I guess what I’m basically saying is that there’s at least one more axis (gear) in most games to add to the race/class/talents axes.

  2. “Does having more choices increase diversity?”

    Yes, as long as the choices matter for the players. If they don’t, players will inevitably gravitate towards the select few choices or group of choices that, amongst the general mediocrity, end up mattering the most. As a dev, you can dream a balanced system on paper, but paper is cheap. What matters is what your players do with that system in game. It doesn’t matter if you give them 20 choices of which only five are commonly viewed as ‘good’ or ‘worth the slot’, or 7 choices of which only five are ‘good’ or ‘worth the slot’. Your players will inevitably end up choosing the five good ones, the rest can go to hell. It doesn’t matter if your players discard 2 or 15 choices, because those choices are the discarded ones. What matters is that everyone chooses the same 5.

    This is something that just recently popped up in the LOTRO boards with the recent patch, reducing the number of equippable virtues from 8 to 5. One camp of players complain about the apparent reduction in choices, everyone ending up being cookie-cutter, etc… while another camp is basically saying it doesn’t matter, because only a fraction of those virtues (reduced number of them or not) are really worth it and everyone will equip those anyway. Cookie-cutter, big or small, however you slice it.

    It’s the same with gear, it’s only more visible. It really doesn’t matter if at L60 you have 100 shoulder pieces to choose from the itemization, if only 10 of them are worth a damn. We’re focusing on the wrong number – we should be looking at the 10, not the 100. If your itemization only offers 10 good ones, then that’s all players will equip, no matter if the selection is from 100 or 100.000.

    Like Bitter said above, I think this is tied to efficiency and how players perceive efficiency. Players (as a mass) are generally dumb –in the sense that they will make mistakes– but not stupid –in the sense that once a choice is proven right, they will go with that.

    I don’t know how it is now with BC, but it’s the same reason why, for example, of the Rogue talent trees (which was quite a good template, offering a lot of possibilities), there were only 2 or 3 *viable* builds around which +90% of the Rogue population always chose. Combat daggers, combat swords and a sub build with hemo for those who wanted that. That was it.

    Reason being that it didn’t matter how much flexibilization the talent tree system allows in all its possible combinations — where the rubber met the road, most of those combinations were either crap or sub-optimal.

    The lesson here is that it’s *much* better to make sure all the options are as equally viable as possible, than to have many options available but the huge, vast majority being crap.

  3. I don’t think it’s possible to make “meaningful” choices that aren’t min/maxxed to death, particularly by a US player base (I could be wrong, but this obsession with being the absolute best of the best seems to be more prevalent in the US than elsewhere). Any exclusive choice will be poked, prodded, and analyzed until someone makes a conclusion that one is better than the other, which will spread throughout the player base. The sad thing is if that choice is adjusted in the future, it’ll STILL be considered a non-choice as most of the players won’t consider new information.

    That being said, I enjoy the idea of allowing players different ways of doing the same thing. Imagine a warrior, for example, who can either be a weapon master, gaining increased chance of critical strikes or a barbarian type who simply bashes you with whatever’s handy, doing increased base damage. They both do the same overall damage, but have different feels.

  4. Silvanis: That’s true. Whatever you put out there will be min/maxed whether you like it or not. But, if you know that from the beginning, you could conceivably create a system or a set of rules that minimize the practical in-game impact of that min/maxing.

    Unfortunately, all the examples that I know of so far, are examples where this min/maxing goes rampant precisely because of this huge difference between viable and non-viable choices. It’s too evident in many cases (see: WoW’s Rogue builds). Players will catch on to what’s going on immediately.

    If there was a system or a set of rules that, while allowing some choices to be better than others, can keep the distance (in effectiveness) between good and not-so-good choices short, you’ll see more variety there.

    For example, back to the Rogue builds, if you have a build that gives you the tools to take down a mob in 10 seconds, and another one that takes down the same mob in 60 seconds, then it’s a no brainer – people will go massively with the 10 seconds one. However, if all the choices you give your players, essentially enable them to take down that same mob between 10 and 20 seconds, depending on choice, then yeah I imagine you’ll see more build variety because in that case personal preference does not get in the way of effective gameplay.

    Think of a player that quite likes to go around stunning mobs, doing a lot of CC. In many cases that avenue of play is invalid because of its flipside – normally, that since we’re giving the player so much power to stun mobs, when he has to kill them he does so very slowly. We either give him tools to stun good, or kill good. Not both. Many of these players will balance on one side the way they like to play, on the other effective gameplay and will make a choice. If you put a lot of separation in effectiveness between their choices, they will go with the most effective.

    Min/maxing is in a way the inevitable conclusion a player reaches when trying to deal with a game world that works against him. It’s the drive to make every bit of stat count, every piece of gear be the best, because in most cases that’s exactly true. In most cases every bit of stat ultimately counts. If not in one fight, then spread along 100 fights. So naturally every player choice will be weighed with that in mind. If you ask a player “Would you rather play the way you like, or be effective and good at this?”, what do you think will be the answer?

    I think collectively (in PvE environments at least), we really need to get out of this mindset of not empowering players too much. I think to a large extent the lack of choices, or the irrelevancy of a lot of choices, is a reflection of the pathological fear we seem to have of players somehow manipulating the system, and using the tools given to gain unfair advantages.

    *On paper* (please note the emphasis), there is no reason why players shouldn’t be equally effective regardless of their preferred style of play. Within the boundaries of design, of course. The reason why this concept breaks apart when it hits code and hits the server I think in many cases is due to inadequate implementations on one hand, and fear of having too powerful players around.

    As a side note, the mindset you mentioned, I’ve observed it almost equally in US, Asian and EU players (I’m not american, btw). EU guys do seem to be more generally nicer and relaxed about a lot of things though. No, don’t point me to the guy screaming for more dots on Ony. There are exceptions ;)

  5. I’ve always wanted gear based games to put in multiple looks for gear. So if i pick up up Helmet of Uberness, I then choose the way it’s going ot look from a selection provided by the developer.

    Would greatly reduce the oh everyone has the same look deal while still allowing min/max type mentality.

    Blizzard REALLY screwed the pooch on gear looks in BC. They reused the same meshes they’ve had since launch and I’ve seen the stupid helmet that looks like a wolf skull helmet(the Hunter Dungeon mesh) about 20 times on everything from cloth to plate.

  6. /e snaps at reading the word “cubicle”

    D-d-d-do you h-have my s-s-stapler… red s-s-stapler… I’ll j-j-just… burn the b-b-blog down…

Comments are closed.