From point estimates to probabilities: chance-of-success goal planning

Every goal calculator produces the same seductive lie: a single number. Save $650/month at 7% and you'll have $412,000 in 20 years. The arithmetic is correct and the certainty is fake — 7% is an average across wildly different possible futures, and your one actual future will not be average. Professional planners stopped issuing point estimates decades ago in favor of chance-of-success framing: 'this plan succeeds in roughly 85% of plausible market histories.' You don't need their software to think this way, and once you do, a whole class of planning mistakes — both the reckless and the over-cautious kind — becomes visible.

What Monte Carlo actually does, in one paragraph

A Monte Carlo simulation takes your plan — contributions, horizon, allocation — and runs it through thousands of randomized market histories built from realistic return and volatility assumptions. In some runs the crash comes early, in some late, in some never. The output isn't a prediction; it's a census: in what fraction of plausible futures does this plan hit the target? That fraction is the success probability. The insight worth keeping is that your plan's outcome is a distribution, not a number, and the plan's quality is about how much of that distribution lands somewhere acceptable.

The hand-run version: three futures instead of a thousand

You can capture most of the benefit with three scenarios instead of ten thousand. For any stock-heavy long-term goal, run your plan at roughly 3% (a genuinely bad multi-decade sequence), your planning number (call it 6%), and 9% (a generous run). The spread between the outcomes is the honest width of your future. If the bad-case number still clears the goal, your plan is near-certain. If even the good case falls short, no market will rescue the contribution rate. Most plans live between those poles — which is exactly what a success probability expresses.

The table's last row deserves a second look, because it's where point-estimate thinking does its worst damage. In a single-number world, risk appears free: crank the assumed return from 6% to 8%, and the calculator cheerfully reports a smaller required contribution. The distribution view reveals what the calculator hid — you didn't reduce the plan's cost, you moved part of it into the bad tail, where it will be paid (if it's paid) at the worst possible moment and by the person least able to afford it: future-you, at the deadline. Probability framing doesn't forbid risk; it just insists risk show up on the invoice.

Why 99% is usually the wrong target

Here's the counterintuitive half. Pushing success odds from 85% to 99% is brutally expensive — the last percentage points require funding against ever-rarer disaster sequences, meaning dramatically higher contributions or a much later date. And the 'failure' being insured against is rarely a cliff: for most goals, the 15th-percentile outcome isn't ruin, it's arriving 18 months late or 12% short — outcomes you'd absorb with a delay, a cheaper variant, or a small loan against a mostly-funded goal. Paying thousands per year of extra contributions to avoid a survivable inconvenience is over-insurance. Flexible goals deserve 75-85% funding confidence; only hard-deadline, hard-dollar liabilities (tuition due in August) justify near-certainty — and those are better served by de-risked assets than by heroic overfunding.

1. Classify the goal: flexible (house, sabbatical, car) or rigid (tuition, a contracted balloon payment).
2. For flexible goals, plan to about 80% confidence: fund it so your planning-number scenario clears with a modest cushion, and pre-decide the fallback (delay, downsize) for the bad tail.
3. For rigid goals, buy certainty with allocation, not overfunding: glide to cash early so the bad tail can't reach the deadline.
4. Recheck annually: a goal that's drifted to 95%+ odds is overfunded — redirect the surplus contribution to a goal running behind.

The bottom line

A goal plan is a bet on a distribution, and a point estimate hides the whole distribution behind its average. Run three futures instead of one, read the spread as your honest odds, and buy more probability with the cheap levers — contributions, time, target size — rather than the deceptive one, risk. Aim near 80% for flexible goals, near certainty only for rigid deadlines, and correct with guardrails along the way. You'll never know which future you're in until it arrives; you can always know that most of them work.