Bayesian Thinking

The Queue at the Bijli Counter

Kiran had been standing in the electricity bill queue for twenty-five minutes, phone at eleven percent, when the man two spots ahead started talking loudly enough for the whole line to hear.

"Buy it today itself, I'm telling you. My cousin works there. Next week this stock will not be at this price, remember I said this."

He said the name of the company like it was a secret he was doing everyone a favour by sharing. An old man nodded like he already knew. A woman in a green saree pulled out her phone to note it down. Kiran did too, thumb moving before the brain had caught up — typed the name into the notes app, added three exclamation marks, and felt something settle into place that hadn't been there two minutes ago. A certainty. This one sounded like a sure thing.

By the time Kiran reached the counter and paid the bill, the stock felt less like a rumour and more like a fact already proven.

Two days later, scrolling before falling asleep, a small news item mentioned the same company had signed a new client — nothing dramatic, just a line in a business roundup. Kiran's thumb paused on it longer than the headline deserved.

The day after that, at work, a colleague mentioned the company's quarterly numbers had come out — steady, unremarkable, the kind of update that doesn't make anyone excited or worried.

And then on Friday, the same colleague, refilling her coffee, said almost as an afterthought, "Arre, did you see, their factory had some one-day shutdown. Production issue I think."

Kiran sat with that for a moment. The stranger at the counter had sounded so sure. The client news had felt like proof he was right. The quarterly numbers had felt like nothing. And now this — a shutdown, one day, nobody seemed panicked about it, but it was there.

Kiran opened the notes app again, looked at the line with three exclamation marks, and didn't know what to do with it anymore.

• Explain why a first impression about a financial situation should be treated as a starting estimate rather than a final verdict.
• Distinguish between a small, evidence-sized belief update and an overreaction that swings a belief all the way on one data point.
• Given a fresh piece of financial evidence, decide how much (not just whether) to adjust a current belief.

Right now, somewhere in your financial life, you're carrying a belief you formed too fast — a stock someone praised, a budget number you guessed at, a friend's tip you took at face value. That belief matters because it's quietly steering what you do next, and if you never let it move, you're making decisions based on a snapshot instead of the full picture as it develops.

Here's the truth: your belief about any financial situation should shift a little every time real evidence shows up — not stay frozen at your first impression, and not swing wildly because of one new fact either. This is called belief updating, and it's really just the "new fact vs. feeling" filter you already know, applied again and again instead of just once. Every time something genuinely new arrives — a news mention, a quarterly result, a colleague's comment — you check it against what you currently believe and ask: does this deserve to move my view, and if so, how much?

The real skill isn't whether to change your mind — it's by how much.

That second question is the part most people skip. They treat updating a belief as a yes/no switch — either I ignore this, or I flip my whole opinion. But the real skill isn't just deciding whether to change your mind. It's deciding by how much. A rumour deserves a small nudge. A confirmed result deserves a bigger one. Weak evidence, weak movement. Strong evidence, stronger movement. Matching the size of your update to the strength of what you just saw is the whole game.

Think about why this is hard. Your first impression sticks in your head with more weight than it deserves — that's called anchoring. Once you've formed a strong first opinion, later evidence has to fight against that anchor just to be noticed at all. This is exactly why people either dismiss new information (because the anchor is too heavy to move) or overreact to it (because once the anchor finally breaks, it breaks all the way).

Neither of those is the disciplined move. The disciplined move is smaller and quieter: notice the evidence, judge its strength, adjust your belief by a fair amount, and keep watching. Do this enough times and your belief stays close to reality as reality actually unfolds — instead of lagging behind it or overshooting it every time something new appears.

By Wednesday, Kiran had almost forgotten about the note with three exclamation marks — until the small news item about the new client showed up while scrolling before bed. Nothing dramatic, just one line in a roundup. Kiran read it twice, felt a small lift, and thought: okay, that's a point in favour, not proof. The belief moved up — a little. Not "this is definitely happening," just "slightly more likely than I thought two days ago."

Thursday brought the quarterly numbers a colleague mentioned over chai — steady, unremarkable, nothing that changed the story either way. Kiran noticed the urge to read meaning into it anyway, the way you squint at a neutral text message looking for a hidden tone. But there wasn't one. So the belief stayed almost exactly where it was. No news isn't good news or bad news — it's just no news.

Friday, over coffee, the same colleague mentioned the one-day production shutdown, almost as an afterthought. This was the moment Kiran's stomach did something — the old instinct to swing hard, to delete the note, to decide the stranger at the counter had been wrong all along. But Kiran paused instead. One-day shutdown, no panic from anyone else, no repeat mentions. Mildly concerning, not disqualifying. The belief nudged down — a little, not all the way to "bad company."

By Friday evening, Kiran's view of the stock was neither the excited certainty from the queue nor a dismissal — it was calmer, steadier, and closer to what the week had actually shown. Three pieces of evidence, three small honest movements, in three different directions — landing somewhere far more trustworthy than the first exclamation-marked guess ever was.

Point: The size of a belief update should match the strength of the evidence — weak or mixed signals deserve small nudges, not a frozen belief or a dramatic reversal.

• Feeling bad or embarrassed that a first guess turned out to be off. — Being 'wrong' is treated like a failing grade in school and in life, so people assume a first estimate was supposed to be correct from the start. Fix: Treat the first guess as a starting estimate, not a test answer. The only real mistake is refusing to adjust it once real evidence shows up.
• Treating any new piece of information as a reason to completely flip a belief. — Dramatic reversals stand out in stories and news, so 'changing your mind' gets confused with an all-or-nothing swing. Fix: Ask how strong the new evidence actually is, then move your belief by a matching amount — a small nudge for weak evidence, a bigger one for strong evidence, rarely a full flip.
• Waiting until you're completely certain before adjusting what you believe. — Certainty feels responsible, like the careful choice, while acting on partial evidence feels risky. Fix: Accept that full certainty rarely arrives on schedule. Make small, honest updates on the partial evidence you actually have, and keep watching.

• Your belief about a financial situation should move a little every time real evidence arrives — not stay frozen, not swing wildly.
• A first impression is a starting estimate, not a verdict — it's expected to be imperfect.
• The size of your update should match the strength of the evidence: weak signals deserve small nudges, strong ones deserve bigger moves.
• Anchoring makes your first opinion feel heavier than it should — that's exactly why it needs deliberate updating.
• Small, honest updates made often beat one dramatic reversal made too late.

Q1. According to belief updating, what should a first impression about a stock or financial tip be treated as?

• A final verdict that shouldn't change
• A starting estimate that can move as evidence arrives
• Something to ignore completely until certainty arrives
• A mistake that should be forgotten immediately Answer: A starting estimate that can move as evidence arrives — A first impression is only a starting point. It's meant to shift as real evidence comes in, not stay fixed or be dismissed.

Q2. Priya hears a strong stock tip from a friend and instantly believes the stock will double. A week later, one small neutral news update appears about the company. Priya immediately decides the whole tip must have been wrong and drops the idea entirely. What mistake is Priya making?

• Anchoring too heavily on the friend's original tip
• Overreacting by swinging her belief all the way on one piece of evidence
• Waiting for complete certainty before deciding anything
• Correctly matching her update to the strength of the evidence Answer: Overreacting by swinging her belief all the way on one piece of evidence — A single neutral or weak piece of evidence should only nudge a belief slightly, not cause a complete reversal. Swinging all the way on one data point is overreaction, not a wise update.

Q3. A belief should only ever be updated once we have complete, total certainty about a situation. Answer: False — Waiting for full certainty usually means updating too late to be useful. The wiser habit is making small, honest updates as partial evidence arrives, then continuing to watch.

Q4. Arjun believes a company is doing well. He then reads a short, unconfirmed rumour on social media suggesting the company might face a minor delay. How much should this rumour shift Arjun's belief?

• Not at all, since he already formed his opinion first
• A small nudge downward, since it's weak, unconfirmed evidence
• A complete reversal to believing the company is doing badly
• He should wait for total certainty before reacting at all Answer: A small nudge downward, since it's weak, unconfirmed evidence — Weak or unconfirmed evidence, like a rumour, deserves only a small adjustment. The size of the update should match the strength of the evidence, not trigger a full swing or be ignored.

Q5. Arjun has believed for months a stock is undervalued. This week an anonymous, unsourced social media post claims a huge deal is coming. He becomes fully certain and puts in his entire bonus. Does an anonymous post deserve a belief update this large? Reveal: Weak: yes, it confirms what he believed, full certainty is fine. Strong: the size of a belief update should match the strength of the evidence — a sourceless rumor is weak evidence and deserves a small nudge at most; confirming what you wanted to believe doesn't make weak evidence strong.