Going from 10 MPG to 20 MPG saves far more fuel than going from 30 MPG to 50 MPG — even though the second improvement looks "bigger" (20 MPG gained vs 10 MPG gained) — because MPG is the inverse of what actually matters: fuel consumed per distance traveled
The previous articles on this site covered MPG/L-per-100km conversion, EV efficiency, and hypermiling. This article addresses a well-documented cognitive phenomenon sometimes called "the MPG illusion" — where comparing fuel economy via "miles per gallon" (or km per litre) leads to systematically wrong intuitions about fuel saved, because MPG is non-linear relative to actual fuel consumption.
The core issue: MPG is "distance per fuel," but fuel consumption is "fuel per distance"
"Miles per gallon" answers: "how far can I travel on one gallon?"
But the question that actually determines fuel costs/consumption is the inverse: "how much fuel do I use to travel a given distance?" — e.g., "how much fuel does this car use to travel 10,000 miles?"
Because MPG is a ratio (distance/fuel), and fuel-used-for-a-fixed-distance is its inverse (fuel/distance) — equal changes in MPG do NOT correspond to equal changes in fuel used for a fixed distance.
Worked example: comparing two "10 MPG improvement" scenarios
Scenario A: a car improves from 10 MPG to 20 MPG.
Fuel used for 10,000 miles:
- At 10 MPG: 10,000 ÷ 10 = 1,000 gallons
- At 20 MPG: 10,000 ÷ 20 = 500 gallons
- Fuel saved: 500 gallons
Scenario B: a car improves from 30 MPG to 40 MPG (a smaller MPG increase than Scenario A's 10-MPG jump, but let's also check a similarly-sized, 10-MPG jump for direct comparison — and then a larger jump, 30→50, to make the point vivid):
Fuel used for 10,000 miles:
- At 30 MPG: 10,000 ÷ 30 ≈ 333.3 gallons
- At 50 MPG: 10,000 ÷ 50 = 200 gallons
- Fuel saved: ≈ 133.3 gallons
Despite Scenario B's MPG improvement (30→50 = +20 MPG) being twice as large, in MPG terms, as Scenario A's (10→20 = +10 MPG) — Scenario A saves far more fuel (500 gallons vs ~133 gallons) — nearly 4x more fuel saved, from an MPG improvement that was half the size, in raw MPG-point terms.
Why this happens: the "1/x" relationship
Mathematically, fuel used for a fixed distance D is D / MPG — and 1/x is a highly non-linear function, particularly at low values of x — the same absolute change in MPG (e.g., "+10 MPG") corresponds to a much larger change in 1/MPG when starting from a low MPG value than when starting from a high MPG value.
Going from 10→20 MPG: 1/10 = 0.1 → 1/20 = 0.05 — a change of -0.05 (gallons per mile)
Going from 30→50 MPG: 1/30 ≈ 0.0333 → 1/50 = 0.02 — a change of -0.0133 (gallons per mile)
The first change (-0.05) is roughly 3.75x larger than the second (-0.0133) — despite the second involving a larger "MPG points gained" figure (+20 vs +10) — this directly reflects the "fuel saved" comparison above (the ratios are consistent — roughly 500/133.3 ≈ 3.75, matching the 0.05/0.0133 ratio).
Why "L/100km" (or similar "fuel per distance" metrics) avoid this illusion
L/100km (litres per 100 kilometers) — covered in the original unit-conversion article as a common metric in many countries — directly represents "fuel per distance," not "distance per fuel" — it's the metric that's already in the "right" form for additive/linear comparisons.
Re-doing the above example in L/100km-style terms (using consistent, though illustrative, units — the specific numbers would differ from the MPG figures due to unit-conversion, but the relationship is what matters here): a car using, say, "10" units of "fuel-per-distance," improving to "5" units — represents a "halving" of fuel consumption — directly, and intuitively, "twice as efficient."
A car using "3.33" units, improving to "2" units" — represents a smaller proportional improvement (a reduction of roughly 40%, vs the first example's 50% reduction) — and additive differences in "fuel-per-distance" correspond directly to "how much fuel is saved, for a given distance" — unlike MPG, where "additive" differences (MPG points gained) don't correspond directly to "fuel saved."
Practical implications: comparing vehicles, and policy discussions
When comparing two vehicles' fuel economy (e.g., "should I buy the 25 MPG car or the 35 MPG car, given the price difference?") — *calculating the actual fuel/cost difference, for your expected annual mileage, by converting both MPG figures to "fuel used for X miles" (i.e., applying the 1/MPG relationship, as in the worked example above) — provides the figure that actually matters for the cost-comparison decision — rather than "comparing MPG numbers directly," which, as demonstrated, can significantly misrepresent the actual magnitude of the difference.
For policy/aggregate discussions (e.g., "improving fuel economy standards from X MPG to Y MPG") — the same non-linearity applies: an improvement from a low baseline MPG (e.g., for larger/heavier vehicles, which tend to have lower baseline MPG) represents a larger absolute fuel-savings impact, for the same "MPG points improved," than an equivalent "MPG points" improvement applied to vehicles with a higher baseline MPG — this is part of why fuel-economy policy discussions sometimes frame targets/improvements in terms of "gallons per mile" (or similar "fuel-per-distance" framings), rather than "MPG," specifically to avoid the MPG-illusion-driven misperception of where improvements would have the largest actual impact.
How to use the Fuel Economy Converter on sadiqbd.com
- For comparing vehicles' actual fuel costs: convert both vehicles' MPG (or whatever unit they're expressed in) to a "fuel per fixed distance" figure (e.g., gallons for 10,000 miles, or using L/100km directly if that's the available unit) — this figure, not the raw MPG numbers, represents the quantity that scales linearly with actual fuel cost
- For understanding "where does an MPG improvement matter most": recognize that improvements at the low end of the MPG scale (improving a low-MPG vehicle) represent larger absolute fuel savings, for the same "MPG points" improvement, than equivalent improvements at the high end — this applies whether considering individual vehicle choices or broader fleet/policy contexts
Frequently Asked Questions
Is "the MPG illusion" specific to MPG, or does it apply to km/L too? It applies to any "distance per fuel" metric (which includes km/L, not just MPG) — the underlying issue is the "ratio" direction (distance/fuel, rather than fuel/distance), not the specific units involved. "L/100km" (and similar "fuel-per-distance" metrics) avoid the illusion — but "km/L" (distance-per-fuel, like MPG, just in different units) would exhibit the same non-linear, "illusion"-prone relationship to actual fuel consumption as MPG does — the fix is using a "fuel-per-distance" metric, not simply "using metric units instead of imperial."
Is the Fuel Economy Converter free? Yes — completely free, no sign-up required.
Try the Fuel Economy Converter free at sadiqbd.com — convert between MPG, L/100km, and km/L, and calculate actual fuel usage for any distance.
