2. Content
• Balanced game
• Dominant strategies
• Randomness
• Fair game
• Difficulty
• Positive Feedback
3. Balanced Game
• Meaningful choices
• Randomness don't make players skills
unnecessary
• Game is fair from players perspective
• Chance to catch up
• No dead ends
• Consistent difficulty levels
5. Examples of Dominant Strategies
Tank rush in Command &
Conquer: Red Alert
Akuma’s fireball attack in Super Street Fighter II Turbo
6. Balancing Dominant Strategies
• Strategies have direct and hidden "cost”
• Rock-Paper-Scissors - RPS mechanism
• Every unit has unique capabilities
7. Randomness
• Game should be based on skills, not on luck
• Use chance with the small risk and reward
• Allow players to decide how much to risk
8. Fair Game
• Equal chance to win
• Symmetry - same resources and options
• Asymmetry - different resources and options
• Provide protection to newcomers
• Challenges are suitable for difficulty level
• Lot of warnings before loosing the game
• No dead ends
• Provide information needed for making decisions
• Information should be in the game
• Avoid challenges not typical to genre
11. Difficulty Types
• Absolute - skill required and stress
• Relative - difficulty of the challenge is relative to
the power provided to meet the challenge
• Perceived - how hard the player finds the
challenge to be
– In-game experience
– Outside factors
15. Adjusting Positive Feedback
• Increase absolute difficulty level
• Don't provide to much power
• Use effect of chance to reduce the size of reward
• Add negative feedback - cost of the achievement
16. Avoid Stagnation and Trivialities
• Provide enough information what player has to
do next
• Don't bother player with small things
17. Guidelines for Game Tuning
• Separate code from data
• Modify only one parameter at a time
• Make big adjustments
• Keep records
• Use pseudo-random numbers
Design a game that is not too easy and not too hard
Balanced and fair
Meaningful choices
- player can choose between different strategies
- none of them is sure win
- dominant strategies are missing
Randomness don't make players skills unnecessary
- Better player should win
- bad luck is OK?
Game is fair from players perspective
Chance to catch up
- In PvP games
- Failure in the beginning of game
- Reasonable opportunities
- Depends on game duration
No dead ends
- dead locks
- no chance to win
Consistent difficulty levels
- difficulty remains within reasonable range
- no big jumps and drops
- don't change suddenly
Strategy - plan for playing the game
E.g. Aggressive, Defensive
Dominant strategy (formal game theory) - best outcome
E.g. preventing opponents to get points
Allow smaller number of different actions
Tank rush in Command & Conquer: Red Alert
- producing large force of tanks in the beginning go the game.
- Unprepared opponents are helpless
Akuma’s fireball attack in Super Street Fighter II Turbo
Unstoppable offensive plays in Madden NFL
Strategies have direct and hidden "cost”
e.g. wizard is better than elf. elf is better than human … why to be human?
e.g. one car is better in straights other in curves
- hidden cost - player is not warned directly
- entirely hidden costs are unfair
Also used in positive feedback - player gets more power or better items when game proceeds
Rock-Paper-Scissors - RPS mechanism
Balanced
One strategy s better in certain conditions
E.g. Virtual Fighter 3 (Strategy Games): Attacking beats throwing, throwing beats blocking, blocking beats attacking
Every unit has unique capabilities
Each unit (that player can control) in the game should be totally different from others
Different characteristics (dimensions). Not only more power.
e.g. different weapons
e.g. chess pieces (player must use them together, no dominant strategy)
Chance - Use it wisely
With small and repeating challenges
Allow players to choose most suitable actions
?Is Poker based on chance or skills?
Different in PvP and PvE games
Equal chance to win (PvP)
Symmetry (PvP only) - not so interesting
Asymmetry (PvE always)
e.g. Fox and Geese
- players may disagree with the fairness (does all have equal chance to win?)
- AI asymmetry - for making stronger opponents and difficulty levels
- Making (asymmetric) resources from (equal) raw materials - e.g. StarCraft
Provide protection to newcomers (Continuously running game) - e.g. multiplayer online games
Challenges – spikes (tipud) are unfair
Loosing - learn by dying is unfair
Dead ends - e.g. not grabbing important object in one level and now in the second level …
Information
- factual knowledge
- except trivia
- ? serious games ?
Avoid challenges not typical to genre
- e.g formal logic puzzles in flight simulator
- Psychologist Mihaly Csikszentmihalyi
- abilities balance difficulty
- enjoyable state of peak productivity
- balance between anxious and boredom
- balance game = manage difficulty of challenges to keep players in flow state
- problem - how to compare difficulty?
Absolute
- e.g. enemy that can be killed with one hit.
- e.g. enemy that can be killed with 2 hits, hits back (time pressure) and moves (more skills needed)
Relative
- e.g. player strength, health
- e.g. car parameters
- e.g. damage to enemy + resistance to damage
Perceived
In-game experience
- how to use UI
- how the core mechanics work
Outside factors
- Previous experience
- Native talent
- Opponent's skills
if game gets easier - unbalanced
e.g. easy, normal, hard
Alternative method - dynamic difficulty adjustment (DDA)
- game detects the player's level of skills
- game makes adjustments
Achievement causes changes in the game (adds more power).
Next achievement is easier to get.
Core mechanics provides rewards to success
Avoids dead ends
e.g. Monopoly 2000 for finishing round
Positive feedback brings game to end - not too soon
…
Cost of the achievement (e.g. 8 ball pool)
Victory conditions are not related with feedback cycle - e.g. chess
Secret agreements against leader - forming alliances. e.g. Diplomacy
Player don't know what to do next
- not enough information
- e.g. need to find the key when all trolls are killed already
Don't bother player with small things
- e.g. where to store money (in construction games)
- e.g. Alpha Centauri - you can do micro management or leave it to computer
big adjustments, not small ones - then reduce it iteratively – change is visible
records - don't make twice
pseudo-random numbers – can track the bugs and effect