Optimisation doesn’t belong in my MMOs

A long time ago, in the esoteric pen and paper RPG world of GURPS (don’t worry if you have never heard of it), somebody wrote a gaming supplement called GURPS Vehicles. It contained rules for players to design and ‘build’ any kind of vehicle they could imagine, from any genre, and then use it in their GURPS game. These rules were physics based and very detailed, and you really needed a spreadsheet to work it out properly.

It was written by a very talented game designer called David Pulver, but I think even he was surprised at the very specific fanbase who adored his vehicles book, given that there really wasn’t much for roleplayers in it at all. Communities grew up based around sharing their vehicle stats and descriptions with each other, without actually playing the game itself.

And many of the regular GURPS players shunned the book, on the grounds that it was way too complicated and – frankly – unnecessary in a RPG where the GM could always handwave any interesting vehicles as necessary. It was almost as if the frenzy of design and optimisation was a separate game in itself.

Why optimisation is not the enemy

There has always been a healthy player base for games or puzzles based around optimisation, where you were able to sit down and carefully design your character/ vehicle/ simulation and then drop it into a simulated world and see what happened. Then tune it a bit for better performance, maybe even make it fight against other people’s simulations to see whose was best. This is just one step away from all the cross-over fanfics you ever imagined – so what happens exactly when Doctor Who fights Dracula? Is Thunderbird 1 faster than the X-Men’s fighter jet? etc etc

Mechanical optimisation is part and parcel of a very simulationist method of playing games and resolving any conflicts.

The alternative, more narrative/ dramatic method, is to decide in advance which of the characters/ vehicles etc should win and then tell the story appropriately. If the stats don’t work as intended, then you ignore or handwave them as necessary.

So – one of these methods lends itself far better to computer RPGs (hints: it’s the simulation model.) A computer is very well able to model a fight as a set of dice rolls with mathematically modelled entities. It’s not so well able to wing a story.

However, optimisation doesn’t necessarily make for a great gaming experience, because most of the optimisation is usually done as a perquisite to the game. So for example, with GURPS Vehicles, you sat down for a few hours with your spreadsheet and designed your amazing creations, and only after that could you play with them. Plus you couldn’t easily tweak them in play without another spreadsheet session. It’s better, I think, to look at optimisation as a separate game in itself, and a good optimisation game will not be easily boiled down to a few ‘correct’ solutions.

Another challenge with a good optimisation game is that you may not know 100% in advance what challenges your model will have to face. In a fighting game, you may not know what sort of opponents you will face. In a racing game, you may not know much about the terrain in advance. Because if you did, you’d just optimise for that environment and you’d be back to the few correct solutions again.

So in a computer game, I’d assert that optimisation is more fun as a minigame when you’re facing randomised challenges. That’s what makes the hybrid designs more interesting and the specialist designs more of a risk.

Why optimisation is the enemy

I commented on Nils’ post yesterday that I thought optimisation is one of the big enemies for players in MMOs these days.

This is because they’re not really well designed optimisation games in the first place. Optimisation becomes a tedious step of looking up builds/ gearouts online and copying them semi-blindly. Now, precisely who finds this fun? Is there really any player who derives any kind of fun from copying a spec from a webpage? (Aside from the relief of not having to worry about it so that they can get on with the parts of the game which they like.)

I don’t think so. The fun in optimisation is in designing your own character, trying it out, and then tweaking to make it work better. It’s not in being told “fire mages suck this patch, noob.”  Or “if you don’t have 5 resto shammies in the raid, you might as well stay home!” And yet MMO design – particularly WoW endgame design – has become so minmaxed that players (and raid leaders) who don’t use the optimal loadouts are at a disadvantage, and are seen as a disadvantage to any group they are in. A side effect of this is the ferocious emphasis on balancing the various specs. When tools are available that can rate the performance of any given spec to within 0.001% AND content is tuned for minmaxers then what can you expect? So MMO devs have managed to create an optimisation game that isn’t very fun for the majority of players. Well done, guys.

Right now, far from having any fun with optimisation, if there was a button in the game that said ‘optimise my character’ that would tweak talent trees, inform the player of the optimal dps rotation, and assign some optimal gear for the current raid then most players would HAPPILY press it.

It used to be that part of the fun of optimising your MMO character was actually collecting your gear, which might have come from a variety of different sources. That too has been largely optimised out – a combination of gear lists, token loot, known instance and crafted loot and readily available information from websites makes even this feel more like a chore. How much fun is it really to gear up in WoW these days? It’s a combination of running random heroics (with relevant tabards for rep also) and whatever you can buy from the auction house.

On top of this has been tacked a fairly fun raiding model where how people actually play their characters during a fight becomes more important. This to me is the more fun side of gaming optimisation, where you have to react a bit more quickly and flexibly to things going on in game, and optimise your strategy/tactics, not your build.

Optimising as a part of gameplay is still fun. But optimising talent trees is not a fun part of current, tightly tuned MMOs. Sure, you can play with those talents to your heart’s content and have some fun messing around, but the choices on offer are not very real. And it can be a heartbreaker. What exactly happens in a game such as Rift if my favourite soul is not the best dps spec? Right now – nothing, my guild won’t care and it won’t stop us downing monsties (thanks Hawley). How about in 6 months time when everyone is minmaxed to the hilt and other people can check your specs?

I still think there is a lot of fun to be had from tweaking characters and character progression, but the most fun gameplay is that which happens as part of the actual session, not outside the game itself. And my ideal MMOs will be far more about how you actually play than how you spreadsheet.

The complexity, it burns! Do you use gear lists?

We know that players love collecting gear because devs tell us so. It isn’t just so that we can customise how our characters look, but  so that we can tweak their stats as well. Even City of Heroes which always shunned the gear upgrade option has  stat-items that you can slot in to improve your skills.

So it’s a very common in-game decision point to pick up a new piece of gear and ask ‘Is this better than what I have now?’

In fact, it’s one of the classic MMO questions. Games are designed so that you have to ask this a lot (i.e. every time you get a drop, there is a decision making process). This is in the same way that the classic player question in a RPG is ‘what do I see?’ Or the classic GM question is ‘what do you want to do next?’

So how do you go about answering that question?

  • Maybe check how the gear looks if the game gives a preview option.
  • Maybe you have a rough idea which stats are best for your class so you can quickly compare.
  • Maybe you ask a friend (or your guild).
  • Maybe you did your homework in advance and already have a shopping list of items you want.
  • Maybe you have an addon which helps crunch the numbers.
  • Maybe you look it up on a website where someone else has calculated long lists of ‘best in slot’ gear.

If you find yourself drifting towards the lower half of that list, it means you needed to do some work outside the game to make that in-game decision. I personally think that gear is more fun when I feel competent to make a snap choice rather than go in with a shopping list approach. But I know for some collector types, they like to tick things off on lists.

In some ways, gear sets (ie. tier sets or matching sets) provide a game-generated shopping list. They don’t even need to be great in their own right to hit the ‘gotta have them all’ collector instinct. But in general, designers like to leave the gear complexity as a toy for players to play with.

But the one thing you do know is that to keep the decision making interesting, there need to be some bad options. Some poorly itemised gear. Some drops which are absolutely worse than what you currently are using. And if you grab one of those and sell your old gear, you’ll never get it back. It’s not necessarily a critically bad decision, there will always be more gear in future, but it may be an irrevocable one.

The complexity curve

Back in my pen and paper days, I did some writing for GURPS. It’s a universal RPG system that covers just about every genre imaginable with a variety of rules and background supplements. And they had one rulebook which was an odd outlier in rule complexity. GURPS Vehicles gave really complex rules for generating just about any type of vehicle you needed in a game – it had lots of physics based rules to make sure that you didn’t generate anything wildly implausible (for some value of wildly implausible). You really needed a spreadsheet to work it properly.

And we always expressed surprise that they were still selling it. None of the other rulebooks came close in complexity. It was just a strange case. But in fact, at the time, that rulebook was one of their strongest and most consistent sellers. There was a core of players who adored the complexity and were happy to sit around creating vehicles, whether or not they actually played the rest of the game. Some of them posted up their designs for other people to use or took requests from other games. They liked it as its own minigame.

So you get this odd complexity curve where things become more and more complex and then … people decide it’s too much hassle and go look it up instead, and only hardcore complexity fiends do the number crunching.

I know that I’ve been using gear lists for awhile in WoW. We don’t receive that many drops and I need to spend my hard-earned DKP to win one so I just want to get the best value that I can. And not feel like a lemon for missing a great item because I didn’t get how the stats worked.

But I feel lazy when I use gear lists. It’s taken the decision making part of the game away from me. Sure, I could do it myself but I’m not up for making complex spreadsheets (especially when I know that other people have done them better – see: landsoul’s dps warrior spreadsheet for example). I wanted to play an MMO, not a spreadsheet. Is that really so unreasonable?

Gear is fun. Upgrades are fun. But crazy complexity is only fun for a small minority. Do you use external addons or gear lists to help decide if a new item is better or worse than what you already have?