Avalanche vs. Snowball vs. Intelligent Avalanche

An argument that persists because both sides are right

Personal finance has argued for decades about the "right" way to pay off debts, with the fervor of a theological dispute and roughly the same resolution rate. The argument persists for an interesting reason: both sides are right — about different things. This chapter lays the three contenders side by side, prices what each one optimizes, and lands on the only genuinely defensible conclusion: the best plan is the one that scores acceptably on all three dimensions — not perfectly on one.

The mathematician's answer

Avalanche — highest APR first — is the mathematician's answer, and within its domain it is simply correct: it provably minimizes total interest paid. Every dollar aimed at the most expensive debt neutralizes more interest than that dollar could anywhere else. If humans executed financial plans the way spreadsheets do, the chapter would end here.

The psychologist's answer

Snowball — smallest balance first, regardless of rate — is the psychologist's answer, and its logic is behavioral rather than arithmetic: killing an entire debt quickly delivers a win you can feel. An account closed is a visible, countable victory in a way that "slightly less interest accrued this month" never is — and that feeling compounds. People who feel progress keep going; a plan you actually follow beats a perfect plan you abandon in month four.

And this isn't folk wisdom — the motivational effect has real empirical support. A 2012 Journal of Marketing Research study by Kellogg School of Management professors David Gal and Blakeley McShane found that consumers who closed a greater share of their debt accounts — not necessarily the largest dollar amounts — were more likely to eventually eliminate all their debt.¹ The "small victories" logic behind the snowball isn't a consolation prize for the mathematically weak; it's a documented driver of the outcome that actually matters, which is finishing. Morgan Housel's framing belongs here: the goal isn't to be coldly rational, it's to be reasonable² — to pick the strategy the real you, with your real patience and your real Tuesday nights, will sustain to the end.

The third lens

Intelligent Avalanche (Intelligent Avalanche) adds the dimension both classics ignore: credit mechanics. Neither Avalanche nor Snowball knows that scoring reads utilization per card, that a near-maxed card is its own flag, that utilization damage has no memory, or that a mortgage application in eight months changes the price of everything. Utilization triage first, then rate order, deviations priced in dollars, weighted by upcoming credit needs — the full treatment is the previous chapter; its role here is as the third lens: cost, behavior, credit. Three different questions — What's cheapest? What will I sustain? What does my score need? — and any payoff order is implicitly answering all three, whether or not its author knows it.

Pricing the lenses against each other

The three lenses stop being abstract when you price their disagreements.

Snowball's extra interest cost versus Avalanche, on typical multi-card debt, often runs a few hundred dollars over the whole payoff — real money, but modest, and worth stating in income context: a few hundred dollars spread over a two-year payoff is a small premium for a plan that gets finished, at any income. Against it: quitting the plan in month four costs the entire remaining benefit of having a plan — the most expensive outcome on the board, and the one the pure-math crowd systematically underprices because it never happens to a spreadsheet. And from the third lens: walking into a mortgage application with a maxed card can cost more than both combined — thousands per year in rate, for holding the "mathematically optimal" order at the wrong moment.

Every lens has a price tag; only one of them — the interest — is printed on a statement. The other two are paid in abandoned plans and repriced loans, which is why they get ignored, and why they shouldn't be.

Choosing deliberately

The conclusion is not a winner; it's a method. Compute the interest cost of each ordering with your actual balances and rates — Plenee does this side by side, so "how much does the motivating path cost me?" gets a dollar answer instead of a debate. Check the credit lens: any near-maxed cards, any applications coming? Then be honest about the behavioral lens, which only you can score: will the pure-math grind hold your motivation for eighteen months, or do you need the early kill? A hybrid is entirely legitimate — snowball one small account for the win, then run Intelligent Avalanche on the rest — and its cost, like everything here, is computable in advance.

The "wrong" method chosen consciously beats the "right" one abandoned — because the chosen one ends with the debt gone, and the abandoned one ends with a spreadsheet nobody opens anymore.