You may notice that the "Median Outcome" figure on the simulation chart shows one number, but the Year-by-Year Details view's final balance shows a different one. That's not an error - it's two different ways of summarizing the same Monte Carlo run, each useful for a different question. This article explains both and how to read each.

A 30-second refresher on Monte Carlo

The planner runs 2,000 simulated lifetimes. Each one uses your plan inputs (income, expenses, taxes, debts, account balances) but with a different random sequence of investment returns drawn from your return assumptions. After 2,000 runs, the planner has 2,000 separate trajectories for your portfolio - each one a plausible "what if returns went like THIS instead?" story.

To turn those 2,000 trajectories into something readable, the planner has to summarize them. There are two completely different ways to do it, and the planner uses both - one on the Monte Carlo Simulator chart and outcome tiles, the other in the Year-by-Year Details view and all year-by-year visualizations.

Way 1: The cross-iteration median (chart and outcome tiles)

At each age in the plan, line up all 2,000 portfolio balances for that age and take the median (the 50th percentile, the one with half above and half below). Do that for every age, and you get a smooth curve. The blue line on the Monte Carlo chart is this curve. The "Median Outcome" tile shows where it ends. The "Worst" and "Best" tiles are the 20th and 80th percentile balances at the final year, computed the same way. The grey shaded band shows the range that contains the middle 60% of simulations.

This is the statistically honest answer to "what's the typical outcome?". Half of all simulations end above the Median Outcome number, half below.

What this view cannot do: it cannot tell you what a year-by-year financial life would actually look like. The 50th percentile balance at age 65 came from a different simulation than the 50th percentile balance at age 66. Cobble those points together and you don't get any single coherent life - you get a curve through points from different paths, with no consistent set of returns, taxes, or withdrawals that would produce it. That's why you can't read taxes and balances from this curve and have them tie out.

Way 2: The representative sample lifetime (Year-by-Year Details and other year-by-year views)

To show a year-by-year ledger where taxes, withdrawals, balances, and RMDs all match up, the planner picks ONE specific simulation from the 2,000 and shows it in full detail. The choice is the simulation whose trajectory is closest in aggregate to the cross-iteration median curve above, so it broadly tells a "near-median" story.

But this is still one coherent lifetime. Its returns happen in a specific order. Its taxes are computed precisely against that lifetime's actual income each year. Every dollar of withdrawal comes from a specific bucket. Balances tie out from one year to the next.

The Year-by-Year Details view, the Dashboard projection chart, the Cash Flow charts, the Flow Map, and every other year-by-year visualization in the planner all show this same representative sample lifetime.

Why the two answers don't agree

The representative sample lifetime is just one of 2,000 paths. Its terminal balance is wherever that particular path's returns happened to land, not where the median of all 2,000 terminals lands. Two factors compound:

• "Closest in aggregate trajectory" isn't the same as "lands at the median terminal". A path that tracks the median trajectory tightly through most of the plan can still end above or below the median terminal balance, because the planner picks the path by summing distances across all years rather than constraining the endpoint.
• The representative path uses precise per-year tax math. Once the path is chosen, the planner re-runs it with precise year-by-year tax and withdrawal calculations (more accurate than the vectorized averages used across all 2,000 runs for the chart). For any given sequence of returns, the precise calc and the approximate one produce slightly different terminal balances. The Year-by-Year Details view reflects the precise version.

In most plans, the representative sample's ending balance lands within roughly 10-30% of the cross-iteration median terminal, but in plans with significant late-life portfolio activity (large RMDs, late one-time expenses, big drawdown swings) the gap can be larger - sometimes 50% or more in either direction. The two legitimately differ - the gap isn't a bug, it's the cost of being able to show a coherent year-by-year life rather than only an aggregate statistic.

Why this design is conventional

This split - statistical aggregate for the chart and outcome tiles, one coherent path for the year-by-year detail - is the same pattern other detailed retirement-planning tools use. The reason every tool ends up here is the same constraint: no single simulated path equals the cross-iteration median, so you have to choose between showing the median number and showing internally-consistent year-by-year detail. Tools that prioritize visible coherence (taxes that actually tie out to balance changes) make the year-by-year view a single path; tools that prioritize statistical purity hide the year-by-year detail entirely. The planner prioritizes coherence: every year-by-year view shows one real simulated lifetime, which is why the detail exists at all.

What about Linear mode?

Everything above applies to the Monte Carlo and Historical Market simulation methods, which produce 2,000 trajectories that have to be summarized. Linear mode doesn't: it runs a single steady-return path, so there is no distribution to take a median of and no separate representative sample to pick. The chart, the outcome figure, and the Year-by-Year Details all describe that one path, and they agree with each other by construction. If you switch to Linear and notice the mismatch described in this article disappear, that's why. What you give up is the range and the success rate, which only exist when there are many simulations to count across. See Monte Carlo simulations for the trade-offs between the methods.

How to read each number

• Median Outcome tile, the 20th-80th percentile band on the chart, and Success Rate: these are statistical answers across all 2,000 simulations. Use them to answer "how likely is my plan to succeed?" and "what's the typical outcome range?".
• Year-by-year table, dashboard projection chart, all other detail views: these are one representative simulated lifetime. Use them to answer "what does a near-median outcome actually look like as a financial life?" - in particular, how much you'd be withdrawing each year, when RMDs start mattering, where taxes get heavy, how the buckets evolve.

If a number on the Dashboard or in the Year-by-Year Details view looks different from the Median Outcome tile, that's expected. They're two different answers to two different questions.