Humans do not play at constant strength. You have good games and bad games. You get focused after a nice combination. You tilt after a blunder. You play worse when the clock shows 5 seconds.
Most chess engines do not care about any of this. They evaluate every position at full strength, regardless of context. The output is technically correct and practically useless if you are trying to understand how humans play.
We modeled human psychological states directly into our engine. Here is how it works.
Session Mood
Every game session starts with a random mood value. This is not a gimmick. It reflects the reality that you walk into a chess session with a certain mental state. Some days you are sharp. Some days you are distracted.
The mood value falls into four ranges:
Focused
mood < 0.85The engine plays conservatively. It tends to pick the strongest candidate move more consistently. This simulates a player who is locked in and making fewer mistakes.
Normal
0.85 to 1.1Standard behavior. The engine picks the top move most of the time but occasionally considers alternatives.
Distracted
1.1 to 1.35Slightly more random. The engine sometimes picks the second-best move when the top two are close in probability. This mirrors a player whose attention drifts.
Tilted
above 1.35More impulsive. The engine is more likely to pick suboptimal moves. This simulates the player who just lost three games in a row and is no longer thinking clearly.
In-Game Mood Shift
Mood is not static. It changes based on what happens during the game.
If your opponent blunders, your mood improves. You feel more focused. The engine reflects this by tightening its move selection.
If you blunder, your mood worsens. The tilt effect kicks in. Move selection becomes more erratic. This is not a punishment. It is an accurate model of how humans actually behave after making a mistake.
If the position improves, mood stabilizes. If it deteriorates, mood declines. The engine tracks this continuously.
Time Pressure Modeling
The clock matters. A lot.
When you have more than 30 seconds, the engine behaves normally. Between 8 and 30 seconds, it gets slightly more impulsive. Below 8 seconds, it enters panic mode and favors fast, obvious moves over deeper calculations.
This is not arbitrary. Research on human chess behavior shows that time pressure fundamentally changes how players think. Under severe time pressure, players abandon calculation and rely on pattern recognition. The engine does the same.
How Mood Affects Move Selection
When the engine evaluates a position, the neural network produces probability distributions for candidate moves. Mood modifies how these probabilities are used.
Say the network outputs three moves: e4 at 35 percent, Nf3 at 25 percent, d4 at 20 percent.
This is weighted random selection based on the network's probability output. The mood acts as a multiplier on the randomness. The result is variation. The engine does not always play the same move in the same position, which is exactly how humans behave.
Why This Matters for Analysis
If you are analyzing a game and want to understand why a player made a certain move, context matters. A move that looks bad at first glance might make sense if you know the player was in time trouble, or had just blundered a piece, or was tilted from a previous loss.
Traditional engines give you one answer: the strongest move. They do not explain why a human would choose something else. Titan Chess models the psychological factors that lead to those choices.
This does not mean the engine plays badly on purpose. It means the engine plays like a human at a specific strength level, in a specific mental state, under specific time constraints. That is a more useful model for anyone trying to understand real chess games.
The Tradeoff
Mood modeling makes the engine less deterministic. Running the same position twice might produce different suggestions. This is intentional. Humans are not deterministic.
If you need consistent, repeatable output, use Stockfish. If you want an engine that behaves like the opponent you actually face across the board, the mood system is the difference.
See the Mood System in Action
Try Titan Chess and watch how the engine adapts to game state, time pressure, and position changes.
Read Next
Why We Built an Opening Book with 585,000 Positions
Most chess engines do not have an opening book. They calculate from move one. Here is why we built one.
Why Human-Like Chess Engines Are Better for Learning Than Stockfish
Stockfish finds the strongest move. Titan Chess finds the move you can actually learn from.