Tobold's Blog
Thursday, July 17, 2008
It's the numbers, stupid!

Sorry for borrowing slogans from Bill Clinton, but sometimes I think that we are overly fixated on feature lists of MMORPGs, and don't look enough at the numerical details. In this post I am going to argue that numbers are extremely important in the design of a MMORPG, and can completely change how a game is played.

But first a real-life example: In the air that you breathe there is 20.9% of oxygen. If for some reason the oxygen concentration drops below 19.5%, this officially counts as Immediately Dangerous to Life and Health. As you can see, a small numerical change in oxygen concentration (less than 2% absolute, 10% relative) can have rather serious consequences.

MMORPGs are ruled by numbers. Every sword blow, every spell is a calculation. Your level, and your experience points, are counted in numbers. Your whole character is a collection of numbers in the form of stats. And what these numbers are, and how they evolve, is to a large extent decided by game developers. A big part of game development consists of balancing numbers, and the decisions taken there have a huge influence on how the players will play the game.

I recently realized one interesting example when I made a gnome deathknight in the Wrath of the Lich King demonstration version at the WWI 2008 in Paris. I made a female gnome deathknight with pink hair, just because the idea was so silly. But in reality the gnome deathknight only *looks* silly. While in many other games a small character like a gnome or halfling would do considerably less melee damage than a big character like a tauren or ogre, in World of Warcraft there isn't any noticeable difference. The gnome deathknight is as efficient a killing machine as the tauren deathknight, despite the obvious difference in height and muscles. That is a design decision by Blizzard, and it leads to people playing and race/class combination without having to worry about the effect on stats.

Another example is the group xp bonus in World of Warcraft. There isn't any for a 2-person group, so if a mob gives 100 xp if you kill it solo, it gives only 50 xp if you kill it in a 2-person group. As in most situations a 2-person group does NOT kill twice as fast as a single player, you end up getting less xp per hour if grouped with a friend. There is a 40% xp bonus for 5-man groups, so the same 100 xp solo mob now gives 28 xp if killed by a full group. In that case you would need to kill 4 times as fast as a single player to make more xp per hour in a group than solo, which again is hardly possible; the spawn rate alone would prevent you from doing so. It is easy to see that the group xp bonus is completely arbitrary. But now imagine we would increase the bonus in a way that generally you would get *more* xp per hour if you killed monsters in a group than if you did it solo. It would turn World of Warcraft into a game where between level 1 and 70 a lot more grouping would happen than now. It wouldn't change the fact that you can solo all the way up to 70, but it would increase the incentive to group with strangers and make new friends, thereby adding the missing "massively multiplayer" part to the leveling part of World of Warcraft. Change the numbers, and you change the game.

My last example is combat. Having played several different MMORPGs this year, I noticed that one typical fight in each game takes a different amount of time, and a different number of keystrokes. For example in Age of Conan combat is very quick; at level 1 some casters can kill a level 1 mob with a single spell. World of Warcraft has a medium speed combat. And other games are even slower. Again that is an important design decision which is completely numerical. If combat is faster, it becomes less tactical. By the time my rogue in WoW has three combo points on a mob and could use the "expose armor" finishing strike, the mob is dead already, and the ability doesn't make sense for this short combat. When he is in a group, fighting a longer fight against some boss mob, the ability suddenly becomes interesting. Again the game would be very much different if there was a numerical change to damage output and health of characters and mobs, making fights longer or shorter.

Of course there are many more examples, I can't list them all. But I hope I was able to make my point that numerical parameters in games are at least as important, if not more, than the typical bullet point list of features you can see in any game announcement. If you see an announced Wrath of the Lich King feature like "New tradeskill: inscription", giving you the ability to modify your spells and abilities to do more damage, or crit more, or have added effects like knockback, that sounds very cool. But how important this new inscription profession is going to be depends solely on the numbers: How much do the inscriptions cost? How much more damage or crit or added effect are they giving? By changing the numbers inscriptions could range from totally useless to something everyone must have. There could be interesting tactical choices, or it could be a boring must-grind-for-best-mats affair where one inscription is strictly better than another. The bullet point announcement really tells us nothing. The real effect is in the numbers.
I guess any game, simulated or abstracted reality that runs on a computer per definition relies on numerical input and feedback. So, yeah, it's important:) If you look at game development over the years the trend is to hide the inner workings and number crunching from the user. Now for me as a number- and detailfreak thats disappointing (but i can see why it is happening).

It becomes even worse when you have several sets of numbers at your disposal but their influence is trivial (AoC comes to mind) or even non existant (AoC again). It all comes down to getting the inner mechanics of the game just right (and they are working with numbers).

Accurate & timely feedback, meaningful attributes & skillsets which have concrete influence on the proceedings and respond logically to user input. At the same time, a fine balance between challenge and frustration is needed. All in all, this is not easy.
Of course numbers are important, but they need to be put in a context and there is a need to understand the game mechanics before being able to understand the numbers themselves.

In your WoW group xp example it would make no sense to just present "no xp bonus for 2 man team, 40% xp bonus for 5 man team".

By experience you indicate for example that the team members likely do not have many skills that affect many enemies at the same time and/or many enemies at the same time is not encountered usually. And that the assumption is here that the team would fight enemies of the same level as solo.

But combined with that knowledge you can of course deduce that WoW does not encourage teaming in terms of xp.

If the 5 man team would fight enemies a few levels higher, would that give better output xp-wise? It would most likely be more fun though assuming enemy numbers are low, since it would not trivilaize the fights.

Comparing the numbers with another game (Tabula Rasa), group xp bonus for 2 man team is 84%, for 5 man team 240%. Numbers themselves are better, but in itself it does not say if the game is better for team experience. Again other factors as enemy numbers, skillsets, levels and in Tabula Rasa's case also kill streak multipliers have to be weighed in.

And neither case says which one is more fun.
If you look at game development over the years the trend is to hide the inner workings and number crunching from the user

In this case City of Heroes/Villains have gone in the opposite direction - they used to hide a lot of the numbers behind, but are now exposing it for those who want to see them.

For those that care about numbers their quality of life has improved in the game. For those that do not care, they can still go on as usual, they do not have to see the numbers unless they want to.
This comment has been removed by the author.
You're completely right about the numbers. The best players are always the ones that understand the numbers going on behind the scenes. The same can be said for the real world I guess.

Perhaps they should make game design as a major of a Mathematics degree. At least it would improve the quality of games being released...
The main problems with relying on numbers to describe game features to people is that, to understand their effect, the person has to theorycraft, which is iffy even for people with a good understanding of the game.

Also, numbers can be changed more easily than game mechanics, so it will be easier to say "group xp bonus is up to 400% for 5 players' than to say "group experience bonus is now included in the game".
It may not be obvious to most people, but having two people group up has its benefits. I dual-box two mages and although I doubt I kill twice as many mobs I do notice:
- I use the most mana efficient spells most of the time. As a result I get much less downtime.
- Since single mobs die twice as fast I take much less damage or have to do something like Frost Nova (which costs you time and mana). Again resuling in less downtime.
- I get twice as much CC, so I run/die less when I get adds.
- When I run out of single mob pulls (TBC likes to do this to you in places) I can take the two or 3 mob pulls easier. And still survive when an add jumps me.

Overall I'd call the experience smoother. And for me that's more fun.
This is an interesting discussion, having recently returned to Guild Wars.

The numbers game is HUGE within the mechanics of PvP for GW.
But, of even more interest is the "Inscriptions" (yes, they have them already, gee makes you wonder how Blizzard thought it that were introduced in Nightfall last year.
These allow you to add certain "written" sayings to various items and you get an intersting play on the math.
With things like "Energy (mana) +4 when health is below 50%" or "+12% damage while in a stance" really makes for some interesting builds.
And isn't that what the numbers game is about...creating that "perfect" build.

What is even worse is in PvE for GW you have Heroes to control (AI based henchmen, but that you control RTS style. They have their own actions, but you can define actions deeper if you wish).
Talk about the min/max game there. You want to have the perfect counter point to your skills and inscriptions, etc.

Add to this the already awesome capability to play with your specs at no cost in GW and your different skill points can be applied as you see fit whenever you enter a town..well, we have the perfect numbers game.

Seeing that WoW will be adding this under pinnings to the meta game could prove to be quite fun.
I for one am a number cruncher both in RL (Statistics generation and Web statistics for my job) and in games (How do I play with those attributes and skills..)
Any other way to play with numbers? I am in.
But, Blizzard also continues their trend of being the "David Bowie" of MMO's (emulate and then innovate.) Do wonders ever cease.
Just a quick thing to point out...
You're borrowing the title from James Carville, not Bill Clinton.

Yes, I know Carville was working for Clinton at the time, but give credit where it is due.
You right about the grouping encouragement issue. City of heroes does encourage you to group through exp so everyone does whenever they can find someone decent.

p.s. plz come on over to my blog and see if you have any ideas to share with me and Anton for the game were making.
I would reverse my list. Numbers only matter after the feature. Its the whole chicken vs. egg debate. Do the numbers come before the feature? Or does the feature produce the numbers?

To me it makes sense that the feature comes first. Then we can compare numbers if that feature exists in both games being compared. You actually allude a bit to that as this post progresses.
One thing I liked about WoW was that you had several distinct fight lengths.

Solo, 5 man, 5 man bosses, raid trash and boss fights.

To min/max you actually used different gear sets for those different fights (as a healer fight length is a crucial parameter to determine gear).

I think you have a very strong point that its actually the numbers that determine gameplay. A very well constructed game (according to me) has several equally effective ways to play, whereas most games the numbers dictate a single optimal approach (e.g. stack 1 stat, spam 1 ability).

However you also need some ideas about what those different play styles should look like.

I've been playing around with a Heroes of MM online clone.

At a given level the different classes are not quite balanced, however the balance shifts as you level up, so while elves are considered very powerful at level 5, they are not the best at low levels or past level 7 as others gain abilities.

This matters in a game where leveling up can take a week not a few hours. Again a numeric choice to make leveling slow.
Post a Comment

Links to this post:

Create a Link

<< Home
Newer›  ‹Older

  Powered by Blogger   Free Page Rank Tool