12 July 2010

Drop systems

Traditional

In MMORPGs, monsters have a list of items they can drop when they are killed. A player can get 0, 1 or more items from the same monster, but a dice is rolled for EACH of the dropable items.

WoW: To get Zod's Repeating Longbow, one knows Lady Deathwhisper is the only one monster to be killed. The game rolls a 1000-sided dice, and if 136 or below comes out, the bow is dropped. However, one can also drop the Nibelung from Lady Deathwhisper at 15.6%. Hence, approximately 2.1% of the time, the dragon will drop both the Longbow and the Nibelung, and maybe other things as well.

RO: To get an Ice Pick, one has to kill Eremes, LoD or RSX, but they drop it at respectively 15%, 0.1% and 0.05%. But Eremes also drops Glittering Jacket at 90% or Exorciser at 35%. So 4.7% of the time, Eremes will drop at least these 3 items.

Procedural

Roguelikes generate a lot of their content procedurally. Unlike MMORPGs, roguelike drops are most of the time randomly-generated: monsters do not drop specific items at certain percentage, they drop classes of items.

Diablo (and Diablo II): the drop system consists of Treasure Classes (sets of items) and a drop process. When a player kills a monster, a first random number is used to determine which TC will be dropped. For Andariel, some of these TC are armo18 (1.02%), weap69 (0.017%) or nothing (28.35%). All monsters have the sum of their TC = 100%. In other words, 28.35% of the time Andariel will drop nothing, and the rest of the time, it will drop one item. Once the TC has been chosen, another random number comes out for "the odds". "The odds" is the chance a particular item is chosen from a TC. For the weap69 TC, the Eldritch Orb has 1/15 chance to be selected. In the end, a player has around 0.001% chance to drop exactly the Eldritch Orb from Andariel. But Baal also drops the weap69 TC at 0.54%, so Baal can drop Eldritch Orb at 0.54% * 1/15 = 0.036%. Finally, various elements such as the item property (fire, ice, ...), name prefixes/suffixes or durability are generated.

Scheduled

Team Fortress 2: the Item Drop System used to give (at 25% chance) a weapon to the player every 25 minutes of game. To avoid unlucky streaks that players complained about, Valve changed it: every time an item drops, the time of the next drop is determined.

10 July 2010

[Literature] Characterizing and Understanding Game Reviews

In Characterizing and Understanding Game Reviews, Zagal et al. give the most salient features and qualities that game reviews have. See the table below for the feature list. They analyzed 120 reviews from 2006 on IGN and Gamespot.

Theme Description
Description What you need to do to play this game as well as its features, modes, and characteristics.
Personal Experience Emotions felt due to the game (during or after play. Also includes technical problems experienced.
Reader Advice Recommendations, strategies for success and enjoyment of game as well as discussion of the skills or abilities necessary to play this game.
Design Suggestions Discussion of features that are missing or lacking or suggestions for future improvement of game.
Media Context Contextualization of game with respect to non-game media properties from film, books, TV shows, comic books, and so on.
Game Context Contextualization of the game with respect to other games, game genres and their conventions as well as the history of games in general.
Technology Affordances and role of hardware on which game runs. Includes discussion of the controllers used or other capabilities.
Design Hypotheses Design Goals that developers/designers had for the game
Industry Discussion of state, issues, or trends of the games industry as a whole.

Zagal et al. also identified other interesting facts about game reviews. For instance, game journalists assumed game developers read their reviews because they sometimes were directly addressing the creators of the game. Maybe reviewers realize they arguably played many more games than most game developers and may thus know more about the medium. Some reviews also commented on company business models.
Reviews also help preserve videogame history because they embed the historical context during which the game was published.

However, reviews had certain flaws. First, discussions pertaining to the methods and means through which game reviews are conducted were missing from reviews. Second, Zagal argued that students taking videogame-related classes might have difficulties expressing ideas about gameplay or articulating their experience with games because most of what students read about games are videogame reviews, and that they are thus generally lacking in models of what in-depth analysis or critique about games look like. Third, reviews commonly assume that the reader is familiar with other videogames and their genre conventions, but they were not providing details as to what those conventions refer to or mean. This could make game reviews inaccessible to the most casual readers. Should game reviews be targeted to fans only? Or could they actually be helping the inexperienced readers in providing references to other video games? Fourth, Dang argued reviewers focus too much on the (lack of) innovation of a game compared to other games (Dang, A. (2006). "The 5 Problems with Videogame Journalism." Retrieved Dec 11, 2008, from http://firingsquad.com/features/problems_with_video_game_journalism/). Zagal thinks the innovation bias is rather a feature of the medium of videogames. Movie sequels don't "improve" on the original. Games do, for the most part.