Twelve portfolios, 2 to 6 cards each, $5,000 to $60,000, average balance $25,183.33. Every one gets $300 a month on top of its minimums, minimums are held fixed for the whole run, and no new charges are added. These are constructed portfolios, not user submissions — every balance, rate and minimum is printed below so you can re-run them. Numbers from this site's public payoff engine.
Twelve portfolios, one $300-a-month habit
Most comparisons of the snowball and the avalanche stop at a sentence: the avalanche saves money, the snowball keeps you going. That sentence is true and it is not a number. So we built twelve portfolios out of the card mixes people actually carry — two cards, four cards, six store cards, one big balance with three small ones — and ran every one of them twice: once paying smallest balance first, once paying highest rate first.
The inputs are identical inside each pair. Same balances, same rates, same minimums, same $300 extra, same order of operations. The only thing that changes is where the extra $300 goes each month. The table is the whole result.
| Portfolio | Cards | Balance | Snowball | Avalanche | Gap |
|---|---|---|---|---|---|
| Two cards, $5k | 2 | $5,000 | 15 mo · $745.84 | 15 mo · $664.66 | $81.18 · 0 mo |
| Three cards, $11k | 3 | $11,000 | 30 mo · $3,372.97 | 28 mo · $2,722.57 | $650.40 · 2 mo |
| Four cards, $20k | 4 | $20,000 | 47 mo · $9,990.95 | 42 mo · $7,431.46 | $2,559.49 · 5 mo |
| Five cards, $50k | 5 | $50,000 | 70 mo · $31,572.74 | 56 mo · $24,381.58 | $7,191.16 · 14 mo |
| Five cards, $60k | 5 | $60,000 | 74 mo · $39,765.88 | 58 mo · $30,943.21 | $8,822.67 · 16 mo |
| Six store cards, $7.4k | 6 | $7,400 | 21 mo · $1,465.55 | 21 mo · $1,341.86 | $123.69 · 0 mo |
| Two cards, both near 30% | 2 | $13,000 | 30 mo · $5,188.49 | 29 mo · $4,699.05 | $489.44 · 1 mo |
| Five cards, $36k | 5 | $36,000 | 54 mo · $14,754.99 | 48 mo · $11,839.48 | $2,915.51 · 6 mo |
| Three cards, wide spread | 3 | $25,000 | 53 mo · $15,626.75 | 44 mo · $11,070.21 | $4,556.54 · 9 mo |
| Four equal balances, $24k | 4 | $24,000 | 44 mo · $8,036.56 | 44 mo · $8,036.56 | $0.00 · 0 mo |
| One big + three small | 4 | $33,500 | 49 mo · $17,739.91 | 45 mo · $16,502.78 | $1,237.13 · 4 mo |
| Six cards, $17.3k | 6 | $17,300 | 43 mo · $6,918.11 | 39 mo · $5,470.42 | $1,447.69 · 4 mo |
The average is $2,506.24, the middle is $1,342.41
Add the twelve gaps and you get $30,074.90. Set that against the $155,178.74 of interest the snowball paid across the twelve and the method choice is worth 19.4% of everything you hand the card companies. Divided twelve ways it is $2,506.24 each.
That average is not the number to plan with, because two portfolios do most of the pulling. The $50,000 and $60,000 runs account for $16,013.83 of the $30,074.90 — more than half the total gap sits in two of the twelve. Sorted smallest to largest, the middle of the twelve is $1,342.41, and five portfolios sit under $700: $0.00, $81.18, $123.69, $489.44 and $650.40.
The months tell a different story again. Nine of the twelve finish 1 to 16 months sooner under the avalanche, but three finish in the same month — the $5,000 pair, the six store cards and the equal-balance portfolio. On those, the snowball costs dollars and not a day. That is the case where the argument about methods is real but small.
Same rate spread, 71 times the gap
The gap is not a function of how much you owe. It is the rate spread — your most expensive card minus your cheapest — multiplied by how long you keep a balance on the expensive one. Two portfolios that sit within a point of each other on spread show what that means: the $7,400 store-card run (11 points) clears in 21 months and the gap is $123.69; the $60,000 run (10 points) takes 74 months and the gap is $8,822.67. Nearly the same spread, 71 times the cost.
| Spread | Portfolio | Balance | Snowball months | Gap | Gap per $10k |
|---|---|---|---|---|---|
| 9 pts | Three cards, $11k | $11,000 | 30 | $650.40 | $591 |
| 9 pts | Five cards, $36k | $36,000 | 54 | $2,915.51 | $810 |
| 9 pts | Four equal balances | $24,000 | 44 | $0.00 * | $0 |
| 10 pts | Six cards, $17.3k | $17,300 | 43 | $1,447.69 | $837 |
| 10 pts | Five cards, $50k | $50,000 | 70 | $7,191.16 | $1,438 |
| 10 pts | Five cards, $60k | $60,000 | 74 | $8,822.67 | $1,470 |
| 11 pts | Six store cards | $7,400 | 21 | $123.69 | $167 |
* The $0.00 row is not a rounding artefact — it is the equal-balance portfolio, explained in the next section. It sits inside the nine-point group and is the one case where the two methods pick the same order, so the spread never gets a chance to cost anything.
Size matters only through time. A big balance that is gone in two years costs less than a smaller one that drags for six, because the extra dollars are aimed at the expensive card for fewer months. That is why "the avalanche saves more on bigger debts" is the wrong rule of thumb: it saves more on long debts with wide spreads, and those two things are not the same as big.
One of the twelve ties at exactly $0.00
The $24,000 portfolio holds four $6,000 balances at 24%, 22%, 18% and 15%, with four $150 minimums and $300 extra. Both methods finish at month 44 for $8,036.56 — the same month, the same total, and the same per-card split: the 24% card takes $1,049.04 and clears in month 16, the 15% card takes $2,170.12 and clears in month 44.
Nothing is broken. The snowball sorts by smallest balance and breaks an equal-balance tie by the higher rate; with four equal balances that produces 24%, 22%, 18%, 15% — exactly the order the avalanche sorts for. When balances and minimums are all equal, the two methods are the same algorithm wearing different names, and the gap is zero to the cent.
This is worth knowing for a practical reason: if your cards sit at similar balances and similar minimums, the method debate is not where your money is. Something else is.
In 3 of 12, the snowball costs money but not time
The $5,000 pair: 15 months both ways, $745.84 against $664.66. The six store cards: 21 months both ways, $1,465.55 against $1,341.86. The equal-balance portfolio: identical. In each case the avalanche wins on interest and the calendar does not move, because the last payment lands inside the same month either way.
If you are choosing between the two methods and your horizon is measured in months rather than dollars, that is the relevant fact: a third of these portfolios finish on the same date. The dollars are real — $81.18 is $81.18 — but nobody quits a plan because of $81.18, and the psychological case for the snowball is about finishing at all.
What this means for your plan
Three things fall out of the twelve runs. First, the gap is real and it is not small on long plans: on the $60,000 portfolio the avalanche is $8,822.67 and 16 months better. Second, it is close to irrelevant on short plans with narrow spreads: $81.18 on the $5,000 pair, $489.44 on the $13,000 pair whose two rates are only two points apart. Third, the lever you control most is not the order — it is the $300.
To put a number on that last point: adding $100 a month to the snowball on these same twelve portfolios saved $1,559.67 of interest on average, or 62% of the average method gap, without changing the order at all. The order is free to change and worth having; the extra is the thing that moves the calendar. Do both, but do not spend a week deciding the order and no minutes deciding the amount.
Reproduce these numbers
Every figure here comes from this site's public engine. Post a portfolio to /api/payoff and you get months, total interest and per-card interest back. The wide-spread portfolio is {"debts": [{"name": "A", "balance": 15000, "apr": 28, "minPayment": 375}, {"name": "B", "balance": 2000, "apr": 12, "minPayment": 50}, {"name": "C", "balance": 8000, "apr": 20, "minPayment": 200}], "extra": 300} — it returns 53 months and $15,626.75 in snowball order, and 44 months and $11,070.21 with the highest rate first.
Two conventions, the same ones every page on this site uses: APR goes in as a percentage (28, not 0.28), and minPayment is held fixed for the whole run rather than recalculated as the balance falls. Two of the twelve were spot-checked against the live API and matched to the cent before publication.
Frequently asked
Does the debt snowball or the avalanche save more money?
The avalanche, in 11 of the 12 portfolios here — by $2,506.24 on average and $1,342.41 at the median, finishing 5.1 months sooner on average. The twelfth portfolio was an exact tie at $0.00 because its four balances and four minimums were equal, which makes the snowball's tie-breaker pick the same order the avalanche would.
How much does the snowball method cost compared with the avalanche?
Across the twelve portfolios the snowball paid $30,074.90 more interest in total, against $155,178.74 of snowball interest — 19.4%. Per portfolio that ranged from $0.00 to $8,822.67. The mean is pulled up by the two largest portfolios; the middle of the twelve is $1,342.41.
Is the snowball-vs-avalanche gap bigger when the debt is bigger?
Not by itself. The gap tracks the rate spread multiplied by how long you stay in debt. A $33,500 portfolio with a 7-point spread cost $1,237.13, while a $25,000 portfolio with a 16-point spread cost $4,556.54. Three portfolios share a 10-point spread and their gaps run from $1,447.69 to $8,822.67 — a 6x difference from payoff length alone.
Can the snowball and the avalanche produce exactly the same result?
Yes. The $24,000 portfolio with four $6,000 balances at 24%, 22%, 18% and 15% and four $150 minimums finished at month 44 for $8,036.56 under both methods, to the cent. Equal balances and equal minimums make the snowball's higher-rate tie-breaker produce the same order as the avalanche.
Do the two methods ever finish in the same month?
In 3 of the 12. The $5,000 two-card portfolio, the six store cards and the equal-balance portfolio all cleared in the same month under both methods, so there the snowball cost dollars but not a single month. In the other nine the avalanche finished 1 to 16 months sooner.
Are these real people's debts?
No. They are twelve portfolios built from common card mixes, from $5,000 to $60,000, and every balance, rate and minimum is printed in the table so you can re-run them. No user data is used and no individual's numbers appear here. The arithmetic is the site's own payoff engine, spot-checked against the live API.