Add Adaptive Perk Selection for AI in Practice/Custom Games

Hello Overwatch 2 Dev Team and Community,

The new Perk System in Season 15 has added incredible strategic depth to matches, allowing players to adapt dynamically to battle conditions. However, AI bots in Practice or Custom Games currently do not use Perks, which creates a few issues:

  1. Mismatched Experience: AI battles feel too simplistic compared to real matches, making it harder for new players to learn Perk strategies.
  2. Limited Training Value: Without Perk usage, AI opponents lack realism, reducing their effectiveness as practice tools.
  3. Untapped Potential: The Perk system’s impact is diluted if AI modes don’t engage with this core feature.

Proposed Solution:
Enable AI bots to automatically select Perks based on in-game scenarios, such as:

  • Role-Based Logic: Tanks prioritize survivability, DPS focus on damage boosts, Supports choose healing/utility perks.
  • Battle Context:
    • Defensive perks when losing (e.g., shields, damage reduction).
    • Aggressive perks when winning (e.g., cooldown reduction, damage amplification).
  • Counterplay: Adapt perks to counter enemy team comps (e.g., anti-heal perks vs. high-sustain teams).

Implementation Ideas:

  • Weighted Priorities: Assign perk selection weights based on real-time data (e.g., low team health = +30% healing perk priority).
  • Difficulty Scaling: Easy AI = random perks; Hard AI = advanced tactical logic.
  • ML Training: Use player data to simulate human-like perk choices.

Benefits:

  • Improve AI realism for better practice and solo/co-op gameplay.
  • Help newcomers transition smoothly to competitive play.
  • Stress-test perk balance through AI-driven edge cases.

If you want smarter, more engaging AI opponents, upvote or share this post! Let’s work together to make Overwatch 2’s PvE modes as thrilling as PvP!

2 Likes