Making Predictions With Outer Banks: Going To, Might and Probably
Prediction grammar gets messy when a textbook turns words like might and probably into fake percentages. Real conversation is more useful than that: speakers choose a form based on what evidence they have, what kind of future statement they are making, and how strongly they want to commit.
Use a Prediction Dial, not a calculator. Move from open possibility → likely judgment → strong confidence. But first ask one crucial question: Is this actually a prediction? Going to can also express a plan, and might can soften a suggestion.
Open possibility: maybe, could, might.
Likely: probably.
Expected / confident: going to when context supports an expected result.
Very strong: definitely, absolutely, gotta be, I guarantee.
Important: these are qualitative positions, not fixed numerical probabilities.
First separate plans from predictions
The form going to—often gonna in informal speech—does more than one job. If you miss that, prediction practice turns into a grammar guessing game.
Going to for a plan
John, we're gonna come out there tomorrow
This is a future action the speaker intends to carry out. The useful question is not “How certain is the prediction?” because this is primarily a plan.
Original practice: “I’m going to call her after lunch.” That states an intention. You are not forecasting your own behavior from mysterious evidence; you are saying what you plan to do.
Going to for an expected result
- [JJ] I got this. It's gonna work!
Here the speaker presents success as an expected future result. Same grammar family, different job: this time it functions as a confident prediction.
No one's gonna know.
This is another prediction: present circumstances lead the speaker to expect a future state. A good learner habit is to ask, “Plan or expected result?” before labeling every going to sentence a prediction.
Might: future uncertainty, but not prediction-only
Might is useful because it lets you keep a future outcome open. It does not tell the listener that the outcome is expected; it tells them it is genuinely possible.
I might have to vanish for a bit."
Here might have to means an action may become necessary. The future is unsettled: necessity is possible, not certain.
Original practice — same situation, different commitment:
“I’m going to leave early.” → plan/intention.
“I might have to leave early.” → leaving may become necessary.
“I’ll probably have to leave early.” → the speaker judges that early departure is likely.
But might can soften a suggestion
- ...you might help.
This is why “might = prediction” is a bad rule. In context, might makes a suggestion more tentative. The modal is doing social work, not simply forecasting the future.
When you see might, ask: Is the speaker describing a possible outcome, possible necessity, or softly suggesting an action?
Probably: likely, but still defeasible
Probably is your workhorse when you think something is likely but still want room to be wrong.
we're probably gonna have to reschedule.
This combines a likelihood judgment with a future necessity. The speaker is not merely saying rescheduling is possible; they are leaning toward it as the expected outcome.
- ...it probably belongs to someone else.
Same stance, different time frame: a likely inference about the present. That matters because probably is not only “future grammar.” It marks the speaker’s confidence in a proposition.
Build the low end of the dial: maybe and could
- Maybe we'll get a finder's fee.
Maybe leaves the future open. It introduces one possible outcome without saying that outcome is especially likely.
[JJ] That could possibly be it.
Could plus possibly makes the commitment especially cautious. The speaker is considering an option, not staking their reputation on it.
- They could come looking for it.
Here could keeps a future risk on the table. It is useful warning language precisely because it avoids pretending the event is certain.
Inference language: show the listener how you got there
Predictions are only one part of uncertainty. In real conversation, you constantly make inferences about what is happening now or what already happened.
Guys... I think there's a boat down there.
I think signals a judgment while leaving room for correction. It is much safer than stating an uncertain observation as fact.
- Maybe it hit the jetty or something.
The combination of maybe and or something makes this a deliberately loose hypothesis. The speaker is proposing an explanation, not claiming proof.
- [John B] I'm not really sure why, though.
Sometimes the most accurate confidence marker is simply admitting uncertainty. This is excellent English, especially when the alternative is fake confidence.
he's supposedly my legal guardian.
Supposedly creates distance from a reported claim. It tells the listener: “This is what is said or believed; I am not fully owning the claim as certain fact.”
Turn the dial up: definitely, absolutely, and strong inference
- Definitely over 50.
Definitely signals strong confidence. Use it when you genuinely want to commit much more strongly than maybe or probably.
- He's absolutely okay.
Absolutely is another strong intensifier. The language sounds reassuring because the speaker removes almost all visible uncertainty from the claim.
- [John B] Gotta be money, right?
Gotta be here works like an informal strong inference: the speaker treats one explanation as the most plausible one. The final confirmation-seeking tag also leaves a tiny social door open for someone else to disagree.
The dangerous end of the dial: “I guarantee”
And I guarantee
I guarantee is not just “probably, but stronger.” It is a major commitment. If your evidence is weak, this wording can make you sound confidently wrong rather than confidently fluent.
Original repairs:
Too strong: “I guarantee they’ll agree.”
Better: “I think they’ll probably agree.”
Safer with weak evidence: “They might agree, but I’m not sure.”
When you actually need precision, say the probability explicitly
a one-in-three chance of survival.
If a situation genuinely requires numerical precision, use explicit probability language rather than pretending that one modal verb secretly equals one exact number. Phrases such as a one-in-three chance communicate a quantitative estimate directly.
The Evidence Meter: choose your prediction in three questions
What do I know? Separate observed facts from guesses, reports, and assumptions.
What am I inferring? Name the part that goes beyond the evidence.
How strongly do I want to commit? Choose possibility, likelihood, expectation, or strong certainty.
One situation, five different predictions
Situation: Dark clouds are moving in, but the forecast is uncertain.
Original practice examples:
| Dial position | Possible wording | What it signals |
|---|---|---|
| Open possibility | “It might rain later.” | Possible; low commitment. |
| Open possibility | “Maybe it’ll rain later.” | One possible outcome. |
| Likely | “It’ll probably rain later.” | You lean toward rain. |
| Expected result | “It’s going to rain.” | You present rain as the expected result, usually because context supports it. |
| Strong confidence | “It’s definitely going to rain.” | You commit strongly. |
Now change the evidence. If the sky clears, your grammar should change too. Good prediction language is responsive to evidence, not memorized in isolation.
Three contrast drills
Drill 1: Plan or prediction?
Original pair: “I’m going to meet Sam at six.” = plan. “That shelf is going to fall.” = prediction/expected result. Ask what context supports before labeling the form.
Drill 2: Might or probably?
Original pair: “I might be late.” = lateness is possible. “I’ll probably be late.” = the speaker judges lateness as likely.
Drill 3: Prediction or suggestion?
Original pair: “It might rain.” = possibility. “You might want to take an umbrella.” = tentative suggestion. Same modal, different job.
Practice the dial while watching
With FunFluen, pause just before a character states a guess or future outcome. Make your own prediction first, then reveal the subtitle and compare the character’s confidence with yours. For speaking practice, replay the same situation three times and deliberately change your commitment: might → probably → a stronger evidence-based prediction.
The point is not to find one “correct” prediction word. It is to make your grammar match your evidence. Once you can separate plan, possibility, likelihood, prediction, and strong certainty, going to, might, and probably stop being a grammar-table headache and become tools for saying exactly how sure you are.
Explore more language-learning guides in Learn English.