The Decision Swear Jar — catch your biases in real time

You can't see your own biases. That's what makes them biases. But you can hear them — because biases have a vocabulary. Specific phrases, used at specific moments, that reliably indicate which distortion is active.

The Decision Swear Jar is a collection of these trigger phrases — the verbal signatures of cognitive biases caught in the act. When you hear yourself saying one of them, you've just received a real-time bias alert. Not from a retrospective review. Not from an external observer. From your own words.

How it works

A physical swear jar collects money every time you use a banned word. The awareness of the cost changes the behavior. You don't stop wanting to swear — you start noticing when you're about to.

The Decision Swear Jar works the same way, but with cognitive biases instead of profanity. When you catch yourself using a trigger phrase, it's a signal — not to punish yourself, but to pause and check which bias has entered the decision.

The phrases aren't inherently wrong. Sometimes they describe a legitimate situation. But they're statistically correlated with biased reasoning, which makes them useful as early warning signals.

You can't feel a bias from the inside. But you can hear one — if you know what to listen for.

The trigger phrases

Perception biases

"This is obvious." Possible bias: Confirmation bias. When something feels obvious, it's usually because you're only seeing the confirming evidence. The contradicting evidence — the part that would make it less obvious — has been filtered out.

"Everyone knows this." Possible bias: Bandwagon effect. The fact that others agree doesn't make something true. It makes it popular. These are different things.

"Look at that candle / that number / that move." Possible bias: Anchoring or Volume Blindness. Reacting to a single data point without checking the context around it. The candle without the volume. The number without the trend. The move without the structure.

"This is just like last time." Possible bias: Narrative fallacy. Pattern matching to a previous event without verifying that the structural conditions are actually similar. Surface resemblance is not structural equivalence.

Decision biases

"I can't sell now — I'm already down too much." Possible bias: Sunk cost fallacy. The amount you've already lost has no bearing on the forward-looking value of the position. But it feels like a reason to stay.

"I just need to get back to breakeven." Possible bias: Loss aversion + anchoring to entry price. The entry price is irrelevant to the current decision. But the brain treats it as a reference point that must be recovered before exit is acceptable.

"It'll come back." Possible bias: Hope masquerading as analysis. Without specific evidence for why it will come back — a structural thesis, not a wish — this is loss aversion providing comfort.

"It's too late to exit." Possible bias: Status quo bias. It's almost never too late to exit. The cost of exiting might be high — but the cost of staying might be higher. "Too late" is a narrative, not a calculation.

"I'll just do a small one." Possible bias: Foot-in-the-door escalation. The small commitment creates psychological ownership, which makes the next commitment (larger) feel like a continuation rather than a new decision. It's easier to approve what you've already started.

Evaluation biases

"I knew this would happen." Possible bias: Hindsight bias. No, you didn't know. You might have considered it as one of several possibilities. But the certainty you feel now is a product of knowing the outcome, not of your pre-outcome analysis.

"My instinct was right all along." Possible bias: Self-serving attribution. If the outcome was good, you credit your judgment. If it was bad, you credit external factors. The pattern is invisible from the inside because it feels like honest assessment.

"This always happens to me." Possible bias: Selective memory + availability bias. "Always" is almost never accurate. The negative events feel more frequent because they're more emotionally memorable. The positive events are forgotten because they don't generate the same emotional charge.

Self-assessment biases

"I've done my research." Possible bias: Illusion of control or preparedness. Research creates the feeling of readiness, but the quality of the research — and whether it was genuinely open-minded — is what matters. Hours spent confirming a thesis isn't research. It's rationalization with a bibliography.

"This is a sure thing." Possible bias: Overconfidence. Nothing is a sure thing in uncertain environments. The phrase itself is the alarm — because calibrated confidence never uses the word "sure."

"I'm not emotional about this." Possible bias: Ego suppression. The most emphatic denial of emotion is often the strongest indicator of its presence. If you weren't emotional, you wouldn't need to declare it.

"Just this once." Possible bias: Special case reasoning / Identity bias. Exceptions that are "just this once" have a pattern of recurring. Each one erodes the rule it exempts until the rule no longer functions.

Using the Swear Jar

Step 1: Learn the phrases. Not to memorize them — to recognize them. Read through the list. Notice which ones you've used recently. Those are your high-frequency bias signatures.

Step 2: Practice real-time detection. For one week, simply notice when you use a trigger phrase. Don't change anything — just notice. The awareness alone changes the dynamic because the phrase now carries a flag.

Step 3: Pause on detection. When you catch a trigger phrase, pause. Not for a full analysis — just for the question: "Which bias might this phrase indicate? Is it active right now?" A 10-second check is enough. If the bias is plausible, extend the pause to a full evaluation.

Step 4: Log the triggers. Keep a simple record of which phrases appear and in what context. Over time, the log reveals your personal bias profile — which biases you're most susceptible to, and under what conditions they activate.

The log is the most valuable part. Not because any individual trigger matters — but because the patterns across triggers reveal your systematic tendencies. If "I just need to get back to breakeven" appears twelve times in your log, that's not twelve separate events. That's one bias operating chronically.

The organizational Swear Jar

The Swear Jar scales to teams. In meetings and decision discussions, trigger phrases serve as real-time bias alerts for the group.

The implementation is simple: the team agrees on a list of trigger phrases. When someone uses one, it's flagged — not punitively, but as a check. "That sounded like an anchoring phrase. Are we anchored here?"

This creates a shared language for bias detection. Instead of one person having to articulate "I think we're exhibiting confirmation bias," anyone can simply say "Swear Jar — 'this is obvious.'" The shorthand makes bias flagging faster, less personal, and more culturally acceptable.

The team version works because biases are easier to detect in others than in yourself. Your own biases feel like clear thinking. Other people's biases look like biases. The Swear Jar harnesses this asymmetry — everyone watches for everyone else's trigger phrases, creating a distributed bias detection system.

The words you use when deciding are the fingerprints of the biases shaping the decision. Learn to read the fingerprints, and you can catch the bias before it reaches the conclusion.