Ok, so that’s the best picture I could come up with.
In the previous power post, I went over the various paths to acquire power, as a player. This post will go over the developer’s mandate to control that path in order to provide game and player balance.
In a traditional start/end game, like Mass Effect, distributing the power is usually limited by two factors – time (experience) and content. The longer you play, the more items you acquire, the better they are and the better your skill set becomes. Some games give a carrot-on-a-stick effect to start the game, like God of War, where you have extraordinary power then lose it shortly after. Still, as the game goes on, your experience and time played provides more power. A developer can incrementally increase the power of enemies and map it along with your increases. A game like Final Fantasy where you can “grind” levels to make content easier also provides a hardcore mode for players that want to complete the game at level 1.
The second part is tiered content. Using Final Fantasy 13 as a great example of tiered content, the developers limit your power acquisition to chunks of gameplay. No matter how much you grind in section 2, the developers have given you a minimum level of power to start and a maximum amount to end with. The variance is also usually pretty small too. Any game that locks out zone X until you complete Y is artificially limiting your power. That’s why games like Fallout and Skyrim can make some sections really easy and others impossible as an open game does not assume your power level – it gradually increases difficulty based on your power. This requires more coding by the developer but less planning since they never have to guess at your power level.
In games with no end content, such as MMOs, balance is difficult. A straight boost to a power stat (power, critical chance, critical multiplier, speed) means that an item from the lower levels can and will compete with a higher level one, unless you inflate it to higher levels. WoW had trinkets that provided 2% critical chance in the lower levels. For a higher level trinket to provide a better stat boost, it had to be 4% at the minimum. Push this across 10 item slots and you have players with incredibly boosted stats. Add in an expansion and now the items need to be 6% or more. Power growth is uncontrolled.
Developers learned fairly quickly that you need to convert something into that base stat – hence ratings. Ratings work under a formula that is based on your character’s level so that an item with 10 rating would give 2% crit at level 20 but only 0.1% at level 50. As you increase levels, you also decrease the conversion ratio so that at any given level, a character should only have X amount of a base power stat.
Moving on a bit to more complicated matters – theorycrafting. In games with multiple playstyles – ranged, melee, casting, tanking – certain stats are more interesting than others. Some classes will prefer to stack critical rating, others speed boosts and to that end, will stack a single stat that provides a linear gain. If I know that 100 crit rating gives me 1% crit, then I will stack 10,000 rating as it’s my most attractive stat. From a balancing perspective, this means that the items that you design will have weights based on classes and some items will simply never be used.
Diminishing returns help with this issue somewhat as the more of a stat you stack, the less valuable it becomes. If you’re familiar with a logarithmic scale, you can see that at a given point, a stat rating becomes less attractive than another. Let’s say critical chance is your best stat, 2:1 compared to speed. Once your critical rating stat gives half the rating it did at 0, then you can start putting points into speed rating. This way, you still have the increased power from critical but the loss of power from the diminished returns is off-set by the increase in power from speed with has yet to have diminishing returns.
From a balance perspective, it’s also easier to balance diminishing returns as you’re effectively capping the possible value for a stat source. If critical rating is the only way to get critical chance, you could make it diminish completely at 30%. If you have 1 or 2 more sources, diminish those as well so that the total possible in-game is 50%. You know that the majority of people will be in the 35% range while the ultra players will be higher, 40-45%. That’s much easier to balance than 10% vs. 80%.
Diablo 2’s magic find stat worked like this. For basic items, it provided a linear return and for the best unique gear, it was a diminishing return. So while you might always be gaining a better chance to find magic items, once you reached about 400 Magic Find, there was little benefit on stacking more of it if you were hunting uniques.
On the flipside, if you have hard caps on some stats, where once you reach a given point there is no more benefit, you’re essentially limiting your gear selection for players. SWTOR for example had an original issue where Accuracy and Speed were completely worthless stats since you only needed a small amount to cap, yet 75% of all the high end gear had these two stats present. With the game’s ability to swap stat packages around, high end players farmed a small set of gear (IA gloves mostly) for the Crit/Surge boosts, which were 3-8 times more valuable for every class. This is poor design. Changes were made in a patch to reduce the return on some stats and the next big patch, 1.2, should change the way stats are distributed across equipment. Still, the stat system is flawed at its core in that game.
Developers have an intricate balance to play if they have multiple paths to power. The more you have, the more you have to make sure that all of them provide benefit to the player in a balanced fashion. If one stat simply shines for everyone, expect game breaking situations (armor penetration in WoW is a good example). Ratings and diminishing returns provide the mechanics that devs can use to balance stats between each other, between classes and across all levels of content.