Game Development Based on Experience/1.4.3

Typically, the goal of game development is to get a high review score (preferably 9.5+). However, during the game development your actions do not influence your review score directly, but rather, your game score.
 * 1) In Game Dev Tycoon there are a lot of factors that go into creating a high quality game. This article aims to provide the information needed to create good quality games.

The Game score  is a hidden value you can only see by Modding (do not confuse it with the review score you see), is basically the sum of your Design + Tech, divided by a modifier based on game size (this compensates for the fact that bigger games take longer to develop and accumulate more points) and then multiplied by several quality modifiers. All those modifiers are usually capped at a value of around 1, and have lowest values of around 0.6. Therefore, failing even one quality check can lower your game score significantly. This game score does not translate into review score directly, but rather is compared to your previous high score (read next paragraph for details) to calculate your review score.

You can influence the sum of Design + Tech you receive mainly outside game development, by researching more features (allows your employees to contribute more points while developing the field that feature belongs to) and training your employees / hiring additional employees (increases point contribution total), while during game development, you can mostly influence all those quality modifiers of your game.

Good quality ≠ Good reviews!
First thing everybody needs to understand: Good quality game DOES NOT MEAN good reviews

In this game, you compete exclusively against your own previous high score (except in the very beginning of the game, where you compete against a pre-set top score value, until you can beat it and set your first high score). Your game score is compared to your high score (with an added increment of about 10%-20%), and that is your final review score (before it gets randomized a bit and you get to see it). Therefore, in order to get a good review score, you need to perform slightly better than your previous score.

This is very important to understand:


 * To get a good review score once you can just keep on playing and eventually, no matter what you do, you will get a 9.5+ game (if you don't lose before) !' 


 * In order to keep getting good review scores every time, you need to be consistent, not good. You need to perform just like you performed last time, but a bit better.

Why would you care to make high quality games then? Because it is easier to make a top-quality game (game with all quality checks maxed) than it is to make a non-top-quality game of the same level of (non)quality.

For example, in order to get best quality based on allocation of sliders, you only need to never apply less than 20% to certain topics (~3-6 per genre), never apply more than 20% to certain topics (~0-3 per genre) and at least twice apply more than 40% to certain topics (~3-6 per genre). However, in order to make a worst quality game, you need to invert all of that - always apply less than 20% to certain topics (~3-6 per genre), always apply more than 20% to certain topics (~0-3 per genre).

Therefore, if you go for best quality, you have 60% freedom (meaning 60% of 100% of the slider is okay, remaining is not) in 2 sliders, 80% freedom in ~1-4 sliders, 20% freedom in ~0-3 sliders and 100% freedom in remaining ~0-6 sliders, and if you go for worst quality, you have 20% freedom in 3-6 sliders, 80% freedom in 0-3 sliders and 100% freedom in remaining ~0-6 sliders. If you compare that, you see that you have ~500%-800% (out of total 900%) freedom in the first case, and 360%-660% in the second case. Therefore, if you adhere to "best quality" ruleset, you have more creative freedom than if you adhere to "worst quality" ruleset.

This is even more radical for, say, genre/topic combo. There are way more Great Combos (quality modifier of 1) out there than Strange ones with lowest possible quality modifier of 0.6. Therefore, if you focus on making only worst strange combos, you will be very limited in what genre/topic combo you can choose (not to mention you will hardly have any multi genre games to do). However, if you focus on making only great combos, you have much more freedom.

For example, Dungeon, Airplane, Fantasy, and many other topics can produce only 1 worst Strange Combo but 4 Great Combos. Overall, there are very few topics that give more strange combos than great combos (Game Dev and Superheroes with 3 strange vs 1 great, Romance, Startups, Hospital and Surgery with 2:1), while most of them produce way more great combos than strange combos.

Therefore, this page will not tell you how to make games that get good review scores. It will tell you how to make top quality games, which will in turn help you be consistent in your score, which will in turn help you get good reviews. If you develop games of consistent quality (great combo, tech/design balance upheld, no bugs etc.) your game score will be almost exclusively based on the sum of Design and Tech points you see, and thus you will have an easier time judging how well you did (before you see the reviews).

In the end it's all up to you, free your mind, think of what you want to make/train or research next, and always keep your cool. The World is waiting for only one thing, and that thing, is you.

For more information on how the quality of a game is assessed, and how it is reviewed, see Review Algorithm.

Introduction
First of all, things to always avoid: As stated above, game quality is calculated as a combination of different modifiers that all are applied as multipliers to your sum of Design + Tech. Setting aside those outlined above as things to avoid, you have the following modifiers influencing your game quality: Therefore, in order to make a guaranteed perfect game (aside from trending, which is random), over the whole development process you have to
 * Developing two games in a row with exactly the same Topic/Genre or Topic/Genre/Second Genre
 * Don‘t make a expansion or sequel before 40 weeks.
 * Developing a sequel using the same engine. (Does not apply to expansions)
 * Developing a large game without using 2D Graphics V4 or higher / 3D Graphics V3 or higher
 * Developing an AAA game without using 3D Graphics V5 or higher
 * Developing an AAA game without assigning at least 3 specialists to respective fields that are considered important for the genre.
 * Tech/Design balance
 * Slider (time percentage) allocation
 * Game genre/topic combination
 * Game genre/platform combination
 * Trending
 * Bugs


 * Get a correct final balance between Design and Tech points


 * Conform to all slider allocation rules


 * Choose a "great combination" of genre and topic


 * Choose a platform that fits your genre (or both of your genres in case of multi-genre)

The process of creating a game can be divided into two major parts: Pre-production and Development.
 * Have no bugs.

Pre-production phase
During pre-production is when the major decisions for the game are made.

Choices available during pre-production:
 * Game Title. The game's name.
 * Size. This affects the amount of time it takes to develop the game it also alters the cost of development. Small games are almost impossible to reach an average score of 10. Therefore, if you wish to receive a perfect score it would be ideal to produce larger games.
 * Topic.
 * Genre.
 * Platform.
 * Engine.

Multi-Genre combinations
The only way to get a "Great Combo" for multi-genre games is to use two genres that when used on their own with the chosen topic result in a "Great Combo". This means that topics like Surgery can not get "Great Combo" on multi-genre games. -source

Topic/Audience combinations
You will occasionally get messages about matches or mismatches of target audience and platform you have chosen. Your choice of audience primarily affects the number of sales your game will generate. The higher the sales modifier of the selected Topic is, the more sales you will get with that target audience. -source

These combinations are covered in the table above.

Platform/Genre combinations
Each genre has its own popularity on each platform. Your choice of genre, relative to platform, slightly affects the review scores. The higher the sales modifier of the selected platform, the more sales you will get with that genre.

When you develop a Custom Console, the genre and audience combinations are generated based on your list of previously released games; they aren't predetermined like the standard consoles. They are therefore different between one playthrough and another, and even potentially different between one custom console and another within the same playthrough.

For genres, your two most frequently used genres will be +++ (Great), your third most frequently used genre will be ++ (Good), fourth and fifth will be + (Okay), and sixth/least frequently used &minus;&minus; (Bad).

For audiences, your most frequently used audience will be +++ (Great), second most frequently used ++ (Good), and least frequently used + (Okay).

Platform/Audience combinations
You will occasionally get messages about matches or mismatches of target audience and platform you have chosen. Your choice of audience primarily affects the number of sales your game will generate. The higher the sales modifier of the selected platform is, the more sales you will get with that target audience. -source

These combinations are covered in the table above.

Development phase
Design sliders allow the player to control the time allocation for the various aspects of the project. The higher a slider is set to with respect to others, the more allocated time it will receive during development, which can be previewed at the bottom bar of the game development screen.

Important: Development has a large influence on your review score.

This means that, for example, if you assign 50% development time of Stage 2 to Dialogues, 30% to Level Design and 20% to AI, then during 50% of the time spent on the phase, 90% of bubbles generated by your employees will be design bubbles and 10% tech bubbles, then during 30% of the time spent, 40% of the bubbles will be design and 60% will be tech, and during last 20% of the time spent, 20% of the bubbles will be design and 80% will be tech.

Game quality and correct Tech/Design balance (exact value you must aim for is based on the chosen genre) is very important during rating calculation.

Single Genre focus
When reading the tables below, remember, it is the bottom composite bar consisting of three parts (located under the sliders) that is important, not how much percentage you assign to each slider individually. All those 40%'s and 20%'s refer to the relative size of the field's part on the bottom bar.

Multi-Genre Focus
The point of multi-genres is to eliminate a genre's requirements, giving you more flexibility with what features to cram in. For example, Strategy/Adventure and Strategy/RPG has only 3 "+" fields and everything else is not bound by rules, while if you'd be making Strategy or RPG, you'd have 6 "+" fields and 1 "-" field to worry about, therefore, multi-genre gives you more creative freedom when you combine genres properly.
 * 1) For multi-genre combinations, the values are weighted with the first chosen genre being twice as important as the second one. Refer to the Review Algorithm for more information on how the following values were calculated.

The downside to using a multi-genre with less than 2 desired fields in a stage is that your game quality will not be increased as much as a multi-ged fields. -source

Employee allocation
Starting with medium sized games and up, you can assign certain employees to certain fields. You must assign an employee to each of the nine fields, and only one can be assigned to each field.

Based on the size of the game, employee will get "used up" in the project at different rate. For medium projects, the rate is 1, for large it is 5/3 and for AAA it is 2. This means that for each percent of time allocated to a certain field, employee assigned to it will receive that many usage percent. Therefore, medium game will require a total of 3 people to complete, large will require 5 and AAA will require 6, without overloading your employees. The total amount of employee's effort used is displayed as a progress bar under employee's name on the slider allocation screen. If you do not let any employee get a serious overload, you will be awarded with a 30% experience bonus for Good Management (this does not affect review scores). Overloading an employee results in less points towards Design and Technology.

No matter who is assigned to what field, every employee keeps contributing design and tech points to the project. However, it seems that depending on who you assign, overall points contributed are slightly varied. Assigning a person that fits better to the field will make you end up with more design/tech points overall. You can refer to the table in the beginning of the article to find out who is best fit for what field.

Case study:

Late in the game (year 40) while developing a medium game assigning employees correctly (every employee has enough design and tech points to fully cover the requirements of the field) gave net result of 2536 total points (average between 2 tries), and assigning employees incorrectly (every employee has not enough points to cover at least one requirement - for example, a 300/800 employee assigned to a 810/90 field) gave net result of 2404 total points (average between 2 tries). This is only a 5% increase but maybe if the difference in stats was more drastic (like, assigning a 720/180 employee to a 180/720 field) there would be a bigger difference in net result.

Since you only compete against your own high score, having a balanced all-around team (every member with very close design and tech stats) would be much more viable because it would be easier to make consistent games. A downside would be that you would have to wait a little bit longer before you can open your R&D lab.

Development phase for dummies
If you are absolutely clueless about what to do after reading the previous paragraph and want a simpler solution, just set the sliders themselves to the percentage values provided in the table below:

Read this before using the table below!

Sliders are not the main way of getting good scores, it is a mix between many factors. Some of them include previous reviews, your staff, what year it is, the genre to the platform and the main one is your tech/design ratio. I'm not saying this table is worthless (it kinda is), but if you follow this table and your tech/design score is way off, then you will get horrible reviews every now and then, because your game quality will be unstable - sometimes you will hit the nail on the head with this table, sometimes you'll be far from top quality, and instability is what leads to bad reviews. Please check out some other pages and read up on the game before using this table or just have fun and make games how you would want to play them. The table way above with + and - is a much better reference if you know what you are doing.

Say you are Tech 300, Design 300, Speed 300, and Research 300

Your 2nd employee is Tech 500, Design 100, Speed 150, and Research 200

3rd Employee is Tech 400, Design 200, Speed 200, and Research 100

Your tech/design ratio will be off. Your tech would be way too high for most genres. Rpg, Adventure...etc would be getting bad quality (due to you getting wrong tech/design balance), while Simulation and Strategy would be getting good quality (due to getting correct balance). If your staff would have an excess of design over tech, it would be vice versa. Therefore, if you'd to follow the table below, you would have good quality games of some genre, but bad quality of other, meaning you would keep getting very bad reviews every now and then for your inconsistency in game score.

While this will work for you at the beginning of the game, when you have one employee and a few features, later in the game, when you have multiple employees and multiple features, this table will no longer provide you with the highest possible game quality but more of a rough guide.

Specialization Training
To specialize in a specific slider you will need to be level 7 and meet required levels of design and technology. This can also serve as a guide for which employees to use for each slider (the Design/Tech requirements for specialization are proportional to the Design/Tech components of each Dev focus). Great games need employees focused on either design or technology, as well as employees with a good balance of both. Save your well balanced employees for Stage #3. Specialization costs 200 research points and 5M credits per person.

More visually explicit version:

Additional Versions
Downloadable PDF version: PDF Version 1.2

Game Dev Tycoon - Game Development Chart - Template: Spreadsheet - Spreadsheet version, includes a Combination Grid for tracking combinations across all Topics and Genre