*Shadowbane*and

*DC Universe Online*and is now a lead designer at KingsIsle Entertainment. Her background and focus is a traditional RPG viewpoint, and she builds from there to go through the process of designing and building stats, character attributes, experience rewards, the leveling curve, and itemization. While her talk is aimed at those in the industry, it offers a unique perspective into what goes into those stats behind our gear and why it takes us so long to level up. Read on for highlights from the presentation.

Schubert began her talk by touching on the RPG promise, which says that things in games make sense and that you earn what you deserve. The rules in games, unlike in real life, are clearly defined, and a player's time is well-spent. But in order to do this, developers need predictable and clear math. If there's one guiding principle behind what goes into RPG math, it's the mantra, "How does it make you feel?"

She explained that the numbers should match up with expectations and that they should be consistent throughout the leveling process because they define the pace of the game. She first looked at character stats, considering the traditional MMO model of health, DPS, and core variables. Using Excel, she demonstrated how to set up a consistent curve of exponential increases to DPS, reminding everyone that even though it significantly goes up with each level, it still needs to "feel" right to the player.

As the talk went on, Schubert discussed several variables that come into play when computing the MMO math. She pointed to the math behind "time to kill" and said it's important because it defines the pace of the game. After suggesting developers keep the math consistent, she used

*World of Warcraft*as an example where it wasn't necessarily the case. She described a stair-stepping of player power in

*WoW*, where player DPS doesn't increase on par with mob level, which results in mob health increasing while DPS remains the same. Naturally, the way players overcame that is by getting a better weapon. The player might feel more powerful in the end, but without the weapon, that would not have been the case.

Schubert explained that itemization means different things at different times: It can mean the stats on the items or it can also refer to the items themselves, as in, "When will that dungeon be itemized?" Several variables need to be taken into consideration when computing how powerful weapons and gear should be, but she stressed that every item needs to go back to those core DPS and health curves that are initially created. She also warned against adding too many different stats because that can get complicated fast, and then as you add more items and gear sets, things can become unbalanced very quickly.

Next, Schubert touched on the math behind the experience curve. In short, she said the two main approaches are to go with a flat curve or an exponential curve for experience per level. With a flat curve, players have much more incentive to go back and do older content, but if you go with that approach, it's important to cap experience from significantly higher mobs so that players can't get multiple levels in just one kill.

Interestingly enough, she finished her talk by turning to the MMO endgame, and as she put it, "endgame" is where it's time to put the spreadsheets away and start all over because endgames tend to feature different player behavior based on different variables. Instead of questing and grinding to level, players at the endgame are running dailies and focusing on raiding, so the math needs to match up with that.

A lot of these terms probably feel familiar to players, but Schubert showed how easy it is to fall into a trap of letting the numbers get out of control without careful organization and solid formulas when setting up the spreadsheet. In addition, she explained that it's common for more than one developer to end up working on the same sets of stats over time, and if the fundamental math behind the game is solid, it means fewer headaches for both players and developers in the long run.