GE vs Samsung refrigerators: an incident-report comparison

What this GE vs Samsung comparison is based on#

The evidence here is two federal complaint records of very different sizes: 265 CPSC SaferProducts incident reports naming a component on a GE refrigerator between 2011 and 2026, and 1,556 doing the same for a Samsung refrigerator across the same window. Every share belongs to its own column, and neither total counts refrigerators sold or in service.

Two records with the same headline#

Both columns are led by the ice maker, and in both it is far ahead of everything else. Samsung's row takes 46% of its record — 715 filings — and GE's takes 24%, or 64 filings. That is 22 points, the largest gap on the page, and it separates two records that nevertheless agree completely about what their complaints are mostly about.

That agreement is worth stating first, because it survives every caveat that follows. Whatever else these two records disagree on, the ice maker is the leading named component in both, by a margin that no reclassification of a few reports could close.

Where the resemblance stops#

Below the top row the two columns are not comparable in the same way, because one of them can resolve small components and the other cannot.

Samsung's second row is the control board at 5%, 72 filings. Its third is the compressor at 4%, 66 filings. Its fourth is the evaporator at 4%, 58 filings, and its fifth the thermostat or sensor at 3%, 39 filings. Every one of those rows rests on dozens of reports, which means that column has a genuine profile below its headline.

GE's second row is the control board at 2% — six filings. Its third is the compressor at 2%, five filings. Its fourth is the wiring or power cord at 2%, four filings. Those are small counts, and a small count is not a small measurement; it is barely a measurement at all.

So the gaps in the middle of this page — 3 points on the control board, 3 on the evaporator, 2 on the compressor, 2 on the thermostat — all have the same structure. The Samsung side is measured and the GE side is a handful of reports. The arithmetic is right and the comparison is lopsided.

What a very large record buys#

Samsung's 1,556 filings buy precision, and it is worth being clear that precision is all they buy. A 1% row there is nine to eighteen reports, so the shape of that column is stable and will not shift. But a larger complaint record with no denominator is not closer to a failure rate than a small one — it is a better-measured share of a self-selected set.

That distinction matters most at the top. The 46% ice maker row will not move. Whether it reflects how often ice makers fail or how often their failures are worth reporting is a question no number of additional filings can answer, because a leak or a flood produces a federal report far more readily than a quiet fault does.

Why the size gap changes what each column is for#

These two records are not two versions of the same object, and treating them as such is the main way a page like this goes wrong. One is a well-measured description of a very large complaint set. The other is a thin description of a much smaller one. They can be laid side by side, but they answer different questions.

The Samsung column answers: within this complaint record, how is attention distributed across components? It answers that in detail, down to rows of nine filings, and the answer is stable.

The GE column answers only: which component do these complaints overwhelmingly name? Beyond that it has nothing to say, because the difference between six filings and five filings and four filings is not a ranking. A reader who wants to know GE's second-most-reported refrigerator component will not find a defensible answer here, and a page that supplied one would be inventing it.

That is also why the direction of the middle gaps should not be over-read. Every one of them leans Samsung, which looks like a pattern until you notice that a column whose rows are three to six filings has almost nowhere to lean except down.

The rows that genuinely match#

Three rows show no measured difference: the condenser at 1% on both sides, the motor at 1% on both, and the heating element at 0% on both. The wiring or power cord differs by a single point in GE's direction, four filings against nine.

Both tails end in single reports. GE's carries a drive belt and a water valve or line at one filing each; Samsung's carries a drain pump or hose at one. Those are the edge of the classification rather than parts of a refrigerator profile.

Reading the pair together#

The ice maker is the first stop on either brand, and on a Samsung it is more nearly the only stop the record names. What differs is the second half of each column: Samsung's offers the control board, the compressor and the evaporator with real counts behind them, and GE's offers nothing with more than six filings. Neither ranks the two brands, because neither counts refrigerators.

GE vs Samsung refrigerators: what the reports mention#

Both columns are shares of their own corpus, and the corpora differ in size. GE: CPSC SaferProducts incident reports for GE refrigerators, 2011-2026, n=265. Samsung: CPSC SaferProducts incident reports for Samsung refrigerators, 2011-2026, n=1556. A difference no larger than one report on either side is marked "within one-report noise".

About the corpus: these are CPSC SaferProducts reports — incidents someone chose to report, not units sold — so the shares on this page are shares of the reports and never a failure rate. The corpus is weighted toward safety events: 12,007 of its 20,581 records (58%) are fire, smoke, burn or melt reports, which over-represents the components that burn and under-represents the ones that merely stop working.

GE refrigerators: the reported component split#

GE refrigerators — every share in this table is a share of: CPSC SaferProducts incident reports for GE refrigerators, 2011-2026, n=265.

Shares of reports, not failure rates — see the corpus note above.

Samsung refrigerators: the reported component split#

Samsung refrigerators — every share in this table is a share of: CPSC SaferProducts incident reports for Samsung refrigerators, 2011-2026, n=1556.

Shares of reports, not failure rates — see the corpus note above.

How to use these numbers#

What this comparison can settle: which components the reported incidents of each brand actually concentrate on, for this appliance. What it cannot settle: which brand fails more often. That would need units sold, and no public source states it. Use the component split to plan for the repair each brand is most reported for, not to rank the two.