Probabilities in Plain Language

Jenni and Greg are in a planning meeting. Jenni says, “It’s very likely we’ll hit our Q4 sales target.” Greg nods, thinking “very likely” means about 90% chance, as in almost certain. Carol chimes in, “I’d say it’s possible, but not certain.” Greg hears “possible” and thinks she means maybe 30 - 40% chance, which surprises him because he interpreted Jenni’s confidence as near surety. In truth, Jenni might have meant “very likely” as 70% chance in her mind, Carol’s “possible” might be around 50%. They could actually be closer in estimate than it appears, or maybe not - but because they spoke in imprecise terms, the team walks away with false consensus or unresolved disagreement. This scenario (or a variant) happens everywhere: people use words like rare, maybe, likely, probably, confident without a shared understanding of the magnitude of those words. The result? Misaligned expectations, faulty risk assessments, and sometimes nasty surprises. The antidote is to start speaking about uncertainty in concrete, plain language probabilities. That means using frequencies or percentages, ranges or buckets that everyone understands the same way. It sounds analytical, but it’s actually deeply human: we’re giving our innate sense of chance a common dialect so we all hear the same thing.

Translate uncertainties into frequencies or ranges. A helpful technique is to express a probability as a frequency, like “1 in 4 chance” or “3 out of 5 times this would happen.” This can be easier to grasp than abstract percentages. For instance, instead of saying “There’s a 20% probability of delay,” you might say “About 1 in 5 chance we’ll experience a delay.” People can imagine 1 in 5 more concretely (perhaps picturing 5 similar projects, one of which ran late). It connects the probability to potential realities. Similarly, giving a range rather than a point estimate communicates uncertainty clearly: “I’d say there’s a 20% to 40% chance of closing that big client this quarter.” That conveys both modest likelihood and the fact you’re not certain to within a single number. When others hear this, they can calibrate their decisions accordingly (maybe count on that revenue only cautiously). Using ranges also protects you from the false precision trap where saying “34% chance” might give an impression of certainty that isn’t there. It’s often wiser to say “roughly one - third chance” or “between 30% and 40%” unless you truly have high precision data. The plain language approach might feel awkward at first (we’re not used to casually peppering speech with numbers), but after some practice it becomes concise and effective. For example, consider risk in everyday language: telling your friend “I’m about 80% sure I left my keys at the office” is far clearer than “I think I probably left them at the office.” The latter leaves your friend guessing how strong that “probably” is - 80% sure or just slightly leaning that way? By stating it numerically or with a fraction (“4 out of 5 chance”), you leave less room for misinterpretation. In a team, if everyone starts couching their confidence this way, you’ll quickly notice decisions become more data - driven. It’s harder to ignore a “1 in 3 chance of failure” than a “maybe this might fail.” The former begs follow - up: “One in 3? Why so high? What can we do to mitigate that?” That’s exactly the kind of productive conversation you want.

Standardize a few probability buckets. To make communication smooth, consider agreeing on a small set of verbal probability categories and what numeric range they correspond to. For example, a team might adopt five buckets:

Rare: ~0 - 10% chance (e.g. “rare” means very unlikely, under one in ten).

Unlikely: ~10 - 30% (say one to three in ten chance).

Possible: ~30 - 60% (nearly a coin flip, the middle zone).

Likely: ~60 - 90% (more yes than not, but not guaranteed).

Almost certain: ~90 - 99+% (only a slight chance it doesn’t happen).

You don’t have to use those exact words; choose ones that feel intuitive in your culture. The key is to get rough alignment. Maybe you even explicitly discuss: “When we say ‘likely’, let’s all mean around two - thirds to three - quarters chance or more.” This pre - agreement can prevent misunderstandings like Jenni and Greg’s earlier. If Greg knows “very likely” on Jenni’s lips means say ~80%, and Carol saying “possible” meant ~50%, he immediately sees the gap. Better yet, Carol might have said “I’d put it at about fifty - fifty,” which leaves little doubt. Some organizations (like the intelligence community or safety engineering teams) actually publish their probability lexicon, e.g. “We use the term ‘highly likely’ to mean ~80% or above, ‘unlikely’ to mean ~20% or less,” because they’ve learned how critical it is. You can do this informally by just discussing with your team or including a footnote in strategy docs. The effect is subtle but powerful: meetings become more precise without losing accessibility. People can still speak plainly (“It’s unlikely we’ll finish by Friday”), but everyone hears a number in their head (~20% or whatever you agreed) instead of each person imagining their own. Additionally, standard buckets make it easier to track predictions and outcomes. If you consistently label something as “likely” and it fails to happen, maybe you need to recalibrate what you call likely. Over time, having a shared scale improves collective intuition. It’s like tuning instruments to the same reference pitch: we’ll still hit wrong notes sometimes, but at least we’re trying to play in harmony. One more tip: encourage attaching a time frame to probabilities. “Likely to hit target” should be “Likely to hit target this quarter” or “by year - end,” otherwise you get differences in assumption about time horizon. Probabilities are always tied to a period or condition, so state it: “I’m 70% confident we’ll resolve the bug by Monday.”

Avoid false precision and overconfidence. When you start quantifying uncertainty, there’s a temptation to push for very narrow ranges or very specific percentages to sound confident. Resist that. If data is sparse, saying “between 40 - 60% chance” is fine - yes, it’s a wide range, but it honestly reflects uncertainty. Don’t tighten it to “around 50%” just to appear sure if you’re not. Similarly, avoid spurious decimals: “We estimate a 83.4% chance success” is probably overkill and misleadingly precise (unless you truly have a precise model outputting that, and even then, you’d likely round to 83%). Rounding to nearest 5% or using broad terms (“roughly two - thirds chance”) often better reflects the fuzzy nature of reality. Another pitfall: once you assign probabilities, some might take them as unchangeable truths. But probabilities should update with new information. If halfway through a project, things are going better than expected, you might raise the probability of on - time completion from 60% to 80%. That’s a good practice called calibration: repeatedly revising your probabilities as you learn. It shows you’re responsive to evidence. A red flag is if probabilities never change - could mean people set them and forget them, or gave a number without underlying reasoning to update. Encourage a culture where it’s okay to say, “Earlier I said 30% chance, but given this new data, I now think it’s more like 50%. Here’s why.” That’s not flip - flopping; that’s good analysis. Also, if you make lots of forecasts, consider tracking their outcomes in a simple way to see if your probabilities were well - calibrated. For example, of all things you said were “80% likely,” did roughly 8 out of 10 happen? If only 5 out of 10 did, you know you’re overconfident at that level. Likewise, if 95% of your “80% likely” predictions happen, you were under - confident (too conservative - 80% should fail 1 in 5 times). Over time you can adjust your language to better match reality. Not everyone has the time or need to formally score themselves, but even casual reflection helps. The goal is to communicate risk accurately so that decisions weigh upside and downside appropriately.

Ask for probabilities in discussions. Let’s say you’re in a meeting and someone says, “I’m concerned this approach could backfire.” Instead of either panicking or dismissing it, a good response is, “How concerned? Would you say it’s a 1 in 10 chance, or more like 1 in 2?” This pushes for clarity. Perhaps they reply, “I think maybe 20% chance of backfire.” Now you have something to work with: not negligible, but not an even bet either. You might then discuss how to mitigate that 20% scenario or decide if the risk is acceptable. By normalizing these questions, you prevent vague fears or hopes from dominating without examination. Similarly, if someone says, “I’m confident we can land this client,” you might ask, “Confident as in 90%, or more like just above 50/50?” They might say “I’d give it 70%.” That still means a 30% chance of losing them, so perhaps you prepare a backup prospect or don’t count revenue until it’s signed. The quality of decisions improves because you’re planning for realistic scenarios, not just best or worst case implied by fuzzy words. Encourage team members to challenge each other gently this way too. If you find someone always says “likely” and you suspect they mean barely above even, ask them to quantify. At first folks might find it odd (“how can I put a number on it?!”), but with encouragement they’ll at least bound it: “Well, not a sure thing, but definitely better than 50/50.” That’s progress. You can offer your interpretation to check alignment: “So maybe around 70% chance? Does that sound like what you mean by definitely better than 50/50?” They might agree or adjust. This calibrates your understanding of that person’s language for next time, even if they don’t explicitly use numbers themselves. In essence, you’re building a team intuition pump: converting qualitative statements into quantitative sense, which then feeds back into better qualitative judgment.

Express forecasts as likelihoods of hitting thresholds. Another plain - language tactic: instead of saying, “We forecast sales of $1M with a standard deviation of $200k” (which might not be plain language to many), you could say, “We’re about 75% confident we’ll hit at least $900k, and 25% confident we might exceed $1.1M.” Framing in terms of hitting benchmarks or ranges can be more digestible. It answers practical questions directly: what’s the chance we at least meet goal X? What’s the chance we blow past it? It also inherently uses ranges (“at least this much” implies not capping the high end, giving a sense of spread). Executives and stakeholders often think in terms of targets (“will we hit the target or not?”). So you can respond with probabilities like, “About 2 in 3 chance we’ll meet the target with current trajectory.” That is more useful than a single number forecast which can be misread as guaranteed. By doing this, you also highlight uncertainty around critical thresholds, which prompts risk management or contingency plans. If a manager hears “one in three chance we’ll miss the target,” they might allocate some extra budget as a buffer or start Plan B discussions. Without that phrasing, if you just said “I think we’ll meet the target,” they might not realize you see a 33% risk of failing at this moment. Sometimes, teams internally use these probabilities but report only a binary confident/not confident upward, which loses nuance. Try including some notion of odds in reports: “Confidence level: High (~80% chance of success)” or “Risk of delay: Moderate (~30% chance project extends past Q2)”. This, again, aligns expectations and reduces the shock if unfavorable outcomes occur (people were at least aware of the 30% possibility).

Shortcut routine estimates with probability phrases. Every day we estimate things: “Can you finish this by Friday?” Instead of “Probably, yes,” consider saying, “I’m about 90% sure I can.” Or if uncertain, “I’d give it 50 - 50; there’s a significant risk of spillover into Monday.” This helps your manager or colleague plan accordingly. If you tell them only “probably,” they might assume 90% and not create a backup plan, whereas you meant more like 60%. By practicing this in small things, you build a habit that then carries into larger decisions. It also builds trust - people know you’re not just saying what they want to hear; you’re giving a considered assessment. And when you do say “I’m 95% sure,” they can take that to the bank (as much as anything can be in dynamic work). On the flip side, if someone keeps saying they’re sure and then it doesn’t happen, you can point to the miscalibration: “Let’s put numbers on your surety next time so we can pinpoint where the uncertainty lies.” Maybe they were ignoring a risk which was 20% but happened, showing that one in five events do occur eventually.

Practice by rephrasing recent statements. Take a few statements from the last meeting or email thread that used squishy terms (likely, unlikely, maybe, should, confident, etc.). For each, rewrite it as a probability or range. For example:

“We should likely finish the feature by the end of March.” - > “We estimate about a 70% chance of finishing the feature by end of March.”

“There’s a chance the client might need more time.” - > “There’s perhaps a 1 in 4 chance the client will need more time.”

“I’m confident the new campaign will boost sign - ups.” - > “I’m about 80% confident the new campaign will boost sign - ups (so about 20% it doesn’t have a noticeable effect).”

Share one or two of these reformulations with colleagues and see if it matches their understanding. They might say, “Oh wow, you think there’s as high as a 20% chance the feature slips to April? What can we do to reduce that?” - which is exactly the productive conversation numerical clarity can spark. Or they might say, “70% chance by March? That seems low to me - I was thinking more like 90%.” Aha, now you’ve uncovered a difference in assumption and can discuss where it comes from. Perhaps you know of some hidden complexity they weren’t aware of. This alignment is gold. As you get comfortable, you don’t have to mechanically convert everything to a number - just the act of thinking this way improves your internal calibration. You may find you automatically avoid words like “sure thing” for anything that isn’t 99%+, or you clarify “unlikely” by adding “(maybe 10 - 20% chance)”. The goal isn’t to turn every conversation into a math class; it’s to achieve clarity and shared understanding so decisions and expectations are based on the same reality. With probabilities in your linguistic toolkit, you reduce confusion and make risk more manageable. Now, armed with base rates and clear probability - speak, you can tackle decisions involving risk and reward head - on. In the next chapter, we’ll combine probability with outcomes to reason about expected value and making smart bets, so you can seize opportunities without gambling the farm.