Debt snowball vs avalanche: the numbers on three real debts

Avalanche wins on interest, but on ordinary debts it wins by less than people think, and the extra you add each month moves the result far more than the method does. That is the short answer to debt snowball vs avalanche, and the tables below show it on three debts, then on two harder cases, then on 30,000 random plans. Every figure was computed, none estimated, with the same code that runs the debt avalanche calculator and the debt payoff calculator, which is the debt engine of the Tavala app's debt payoff planner.

The examples are in US dollars with US credit card rates. For scale: the Federal Reserve's G.19 release of 7 October 2026 puts the average rate on credit card accounts that paid interest at 22.15% in the second quarter of 2026 (20.94% across all accounts).

The two methods

Both methods pay every debt its minimum every month. They differ only in where the money left over goes.

• Snowball sends it to the smallest balance. When that debt is gone, its minimum joins the extra and goes to the next smallest.
• Avalanche sends it to the highest interest rate. When that debt is gone, its minimum rolls on to the next highest rate.

The monthly total never drops: a cleared debt's minimum keeps being paid, to the next debt in line. That rollover is why a plan finishes sooner than each debt's own schedule suggests.

How every number here was computed

The site's calculators run the Tavala app's debt engine, ported line for line from the app's Dart code to TypeScript. It counts money in integer cents and rates in basis points (21.9% is 2190), so nothing drifts by a fraction of a cent over four years of months. The rounding is Dart's: half away from zero, once per debt per month.

Each month, in this order:

1. Interest is added to every debt: balance x APR / 12, rounded to the cent.
2. Every debt is paid its minimum.
3. What is left of the month's budget (the extra plus the minimums of debts already cleared) goes to the priority debt, then the next one if the first is paid off. The priority is decided on the balances after this month's interest.
4. A debt that reaches zero records its month.

Minimums are fixed amounts, one rate per debt for the whole plan, every payment on time, no new spending on the cards, no fees. Here is month 1 of the first example below, with $150 extra, to show the arithmetic:

The credit card's interest is $2,140 x 21.9% / 12 = $39.055, rounded to $39.06. Avalanche puts the $150 on the credit card, the highest rate. Snowball puts it on the store card, the smallest balance.

Three ordinary debts

A store card of $900 at 12% with a $35 minimum, a credit card of $2,140 at 21.9% with a $65 minimum, and a car loan of $7,900 at 6.4% with a $190 minimum. That is $10,940 owed and $290 a month in minimums. These are the example debts the site's calculators open with.

Interest is the total over the whole plan. Three things stand out.

With no extra, the method does not matter here. The store card clears on its own minimum in month 30 under both. Its $35 then goes to the credit card under both, because the card is both the higher rate and, by then, the smaller of the two balances left. Same plan, to the cent.

With an extra, the gap is about $40 to $90. That is 5.6% to 6.3% of avalanche's interest. The car loan is the biggest debt and the last one either way, so both methods finish within a month of each other.

Snowball's first win is real. At $150 extra, the store card is gone in month 6 instead of month 14 (avalanche clears the credit card in month 12, then the store card). At $300, month 3 instead of month 9.

When the gap is large

The gap grows when the biggest balance also carries the highest rate, and snowball spends months on small debts that cost nothing. Here: a credit card of $9,500 at 26.99% with a $240 minimum, beside a $600 medical bill and a $1,200 furniture plan, both interest-free with $50 minimums, and a $4,800 student loan at 5.5% with a $55 minimum. That is $16,100 owed and $395 a month in minimums. The card's rate is above the Fed's 22.15% average.

Here the method matters even with no extra at all. When the medical bill clears in month 12, avalanche moves its $50 to the card, and snowball moves it to the furniture plan, at 0%. Every dollar snowball sends to an interest-free debt is a dollar left on a card charging 26.99%. At $200 extra, snowball pays $2,280.49 more and finishes four months later. Its interest is 56% higher than avalanche's.

When snowball's first win comes much sooner

Snowball's case is strongest when the smallest debt is small and the highest-rate debt is big, so avalanche goes a long time without closing anything. A main credit card of $6,800 at 24.99% with a $170 minimum, a store card of $450 at 19.99% with a $25 minimum, and a personal loan of $3,200 at 11.5% with a $105 minimum: $10,450 owed, $300 a month in minimums.

At $100 extra, snowball closes the store card in month 4. Avalanche closes nothing until month 22, and then only because the store card's own $25 minimum has paid it off. By month 18, snowball has closed two of the three accounts and avalanche none. The price of that is $495.42 over three years, and one extra month.

Whether 18 months of earlier wins is worth about $500 is not a math question. It depends on whether you would still be paying in month 15 without one.

30,000 random plans

Three hand-picked cases prove little, so the engine also ran 10,000 random sets of three debts at each of three extras. Each debt's balance was drawn evenly between $300 and $10,000 and its rate between 0% and 29.99%. Its fixed minimum was 2% of the balance plus the first month's interest, at least $25, so every debt could be paid down. The random numbers came from a seeded generator, so the run repeats exactly.

In 30,000 plans, snowball never cost less interest than avalanche. It tied about one time in six, and every tie was a plan where the smallest debt also had the highest rate, so both methods chose the same order. The median gap, ties included, was $129 to $188, or 4.4% to 11.3% of avalanche's interest. One plan in ten had a gap above $464 to $822, depending on the extra. Snowball cleared its first debt sooner in 50% to 61% of the plans and later in 2% to 12%; in the rest, both cleared their first debt in the same month.

So "avalanche almost always" is right, and "by a lot" usually is not. The gap is a few hundred dollars over a few years for most mixes of debt, and a few thousand when a large card sits beside small cheap debts.

The number that matters more: the extra

Against the method, compare what the extra itself is worth. On the first set of debts:

Snowball with $50 extra pays $1,668.87 in interest, which is $596.71 less than avalanche with nothing extra. The worse method with a small extra beats the better method without one.

On the second set, where the method matters most, the order still holds, but the margin narrows. Snowball with $100 extra pays $8,368.79, which is $1,042.94 less than avalanche with no extra. Avalanche with $100 extra ($5,677.39) beats snowball with $200 ($6,359.64) by $682.25. In that mix the method is worth more than an extra $100 a month, so getting the order right matters as much as finding more money.

What the research says about the early win

The case for snowball is behavioral, and there is evidence for it. David Gal and Blakeley McShane studied nearly 6,000 clients of a debt settlement firm (Journal of Marketing Research, 2012, summarized by Kellogg Insight). Clients who closed more accounts were more likely to finish the program than clients who had paid off a similar dollar amount across fewer accounts. The authors note the limits: account counts and dollars are hard to separate, the data comes from one firm's clients, and starting small makes most sense when the rate differences between small and large debts are modest. That last condition is exactly the difference between the first table and the second.

The same pull can cost money. Amar, Ariely, Ayal, Cryder and Rick (Journal of Marketing Research, 2011) found in four experiments that people paid off small debts first even when the larger debts had higher rates, a habit they called debt account aversion. Pointing people at the interest each debt was building helped them cut their total debt faster.

Ramsey Solutions recommends snowball and says why: "Consistency and momentum beat other methods that are based on interest rates" (Ramsey). The tables above let you price that momentum for your own debts instead of taking it on faith.

Choosing between them

This is not advice, and these are not your debts. The method that works is the one you keep paying, and the tables suggest a way to read your own numbers:

• Find the extra first. On most mixes of debt it is worth several times the choice of method.
• Look at the gap in dollars. Put your debts into the debt avalanche calculator, which shows both methods side by side with the interest saved. If the gap is under a couple of hundred dollars, the early win may be cheap.
• Watch for the large-card case. A big balance at a high rate beside small, cheap debts is where snowball gets expensive.
• Consider an order of your own. The debt payoff calculator also takes "Your order", for a reason the rates do not see, and shows what that order costs against the other two.

In the Tavala app

The app's payoff planner defaults to Highest rate first, which it describes as "Highest interest rate first (avalanche) - least total interest." Smallest first is the other pill under What if, and a comparison line says how much interest one saves over the other on your debts. The planner, the payoff order and both strategies are free with any number of debts; the extra-payment, lump-sum and consolidation what-ifs need Tavala Plus once you have two or more. The app has no custom order: only those two strategies. The site's debt payoff calculator has one. See plan to pay off your debts and track a loan or a credit line. If you are looking for the extra, it usually turns up in a few weeks of tracking what you spend.

What the model leaves out

Real card minimums usually shrink as the balance falls; these are fixed, which is how the planner takes a debt's monthly payment. Promotional rates, deferred interest, fees, late payments and new spending are not in it. A variable rate changes everything after the month it moves. The numbers show the shape of the trade between the two methods, not your exact payoff date.