Full notes
Full Teamfight Manager 2 update
Read the full published notes in a cleaner layout. The original post stays linked below.
Repeated intro
Hello, this is Team Samoyed! We plan to share our development progress for Teamfight Manager 2 through development logs every 2-3 weeks. We look forward to your continued interest and support!
What changed
- Maps
- Gameplay
- Fixes
- UI and audio
Match Engine Improvements
Work on improving the match engine is proceeding continuously alongside other development tasks. In this update, we focused most heavily on verifying and improving how much influence stat differences have in the current match engine state.
At the current stage, when a team consisting entirely of players with stats of 100 plays against a team consisting entirely of players with stats of 1, the win rates are approximately as follows:
When both teams use the exact same champion composition and tactics, completely excluding factors other than stats, the win rate is about 93 : 7.
When both teams draft using the same level of Ban/Pick algorithm, the win rate is about 84 : 16.
In the game, an average stat of 90+ represents a world-class player, while an average of 10 or below represents the lowest-tier player in a minor 2nd league. Considering this, the win rate differences mentioned above are realistically still insufficient. We believe that if a team of world-class players faces a team of bottom-tier minor leaguers with similar team compositions, it should be difficult for the weaker team to win even once or twice out of 100 games.
In this regard, the balancing of stat influence still requires further improvement and research. However, implementing this is quite difficult because we need to show a clear difference in ability solely through player AI, while ensuring that the unskilled team does not perform nonsensical or "broken" plays. We will continue to strive to improve the impact of stat differences.
Beyond the actual win rate difference, it is important that when watching the match, the side with higher stats actually feels like they are playing better. The two images above show a mid-laning phase trade between the same Magic Knight vs. Sniper matchup. One shows the Magic Knight with 1 stats and the Sniper with 100, and the other shows the reverse. While the difference isn't extremely drastic yet, you can see that differences in positioning and trade timing based on stat levels lead to quite different developments.
In terms of win rate, in the matchups above, the Magic Knight takes the lane lead when she has 100 stats, and the Sniper takes the lead when she has 100 stats. We are working to ensure that these forms of stats influence the internal game structure and that players can visually verify this impact.
Tactics Improvements
The overall framework of tactics remains the same, but there have been some detailed adjustments.
When split push tactics are set, you can now designate specific positions to push the side lanes.
Ban/Pick AI Improvements
We have slightly improved each team's Ban/Pick AI so that each champion's position is more clearly fixed. Now, champions will mostly be used in only 1 or 2 specific positions. We improved this because the previous AI logic used by the computer had issues with not correctly calculating a champion's positional suitability.
New Game Options
We have added features to set whether player stats are revealed and to adjust the Ban/Pick difficulty when starting a new game.
Hide Attributes
An option to hide each player's stats has been added. If you choose to hide attributes, as shown above, some stats for players with insufficient information will not be visible or will be displayed as a range. You can gradually secure information to accurately grasp a player's stats by dispatching scouts to investigate or by playing matches against teams containing that player.
Ban/Pick Difficulty Settings
An option to adjust the Ban/Pick capability of other teams has been added. In computer-vs-computer tests, matches between "Ban/Pick Easy" and "Ban/Pick Normal," as well as "Normal" vs. "Hard," resulted in a win rate of about 4:6. Matches between "Easy" and "Hard" resulted in a win rate of about 2:8 solely due to the difference in Ban/Pick capability.
Match Replay
A feature to view replays of past league matches has been added. You can check replays for all officially played matches from the full schedule list.
Practice Match
A feature to request and play practice matches has been added. You can choose a day when neither team has an official match to hold a practice match, which is played as a single set.
Transfer Market
Additionally, we have implemented the basic functions of the transfer market related to the negotiation process between teams and players. Based on this, we are currently refining the UI/UX and logic for the transfer market in actual gameplay.
That is all for this development log. Thank you for reading!
Source
Changelog.gg summarizes and formats this update. How we read updates.