The YouTube tier list
There’s a guy on YouTube — you’ve seen him, you’ve maybe even subscribed to him — and he makes solar panel “tier lists.” He sits in front of a green screen wearing a polo shirt and ranks panels from S-tier to D-tier based on a single number: efficiency. Maxeon goes in S-tier because it’s 22.8%. Some Tier-2 Chinese brand goes in B-tier because it’s 20.4%. He says these numbers with confidence as if they’re the same number, measured the same way, comparable.
They are not the same number measured the same way. They might not even be measuring the same thing. The question of what efficiency actually means in solar panels is hilariously, painfully more complicated than YouTube comparison videos let on.
What efficiency actually measures
At the most basic level: solar panel efficiency is the percentage of sunlight energy hitting the panel that gets converted into electrical energy. If 1000 watts of sunlight hits a panel and the panel outputs 200 watts, the panel is 20% efficient. Simple.
Except: there are at least four different things people call “efficiency” in this industry, and they’re not interchangeable.
- Cell efficiency: How efficient an individual photovoltaic cell is in lab conditions. NREL tracks these. Best lab cells hit 27–30% for single-junction silicon, 33%+ for multi-junction.
- Panel (module) efficiency: How efficient the assembled panel is. Lower than cell efficiency because of spaces between cells, frame, glass losses, and connection losses. Top residential panels are 20–23%.
- Inverter efficiency: How much DC the inverter loses converting to AC. Typically 96–98%.
- System efficiency: The whole installation. After temperature derating, soiling, wire losses, and inverter losses, real-world system efficiency for a 22% panel might be 16–18%.
YouTube tier list guy is showing you panel efficiency under Standard Test Conditions (STC) — 1000 W/m² of sunlight, 25°C cell temperature, AM1.5 spectrum. STC is roughly the conditions of a sunny but cool spring morning. It’s nowhere near the conditions your panels operate in 95% of the time.
The Shockley-Queisser limit and why 25% is the magic number
In 1961, two physicists named William Shockley and Hans Queisser calculated the theoretical maximum efficiency of a single-junction silicon solar cell. The answer was about 33%. That’s the absolute ceiling — physics says you cannot make a single-junction silicon cell more efficient than that, period.
The reasons boil down to: sunlight contains photons of many different energies, and a silicon cell can only use photons within a specific range. Photons with too little energy pass through. Photons with too much energy get absorbed, but only the “right” amount of their energy goes to making electricity — the rest becomes heat.
The current commercial record for production single-junction silicon panels is around 25.5% (Maxeon’s IBC architecture, Longi’s HPBC). NREL’s lab record for a heterojunction silicon cell is 27%. Both are pushing against physics. The reason 25% is “the magic number” is that getting from 22% to 25% required ten years and billions in R&D — and getting from 25% to 33% requires breaking past current architectures into tandem cells where you stack two different semiconductors. That’s perovskite-on-silicon territory.
Where this gets complicated
Real-world efficiency is always lower than STC efficiency. Always.
Temperature: A panel rated 22% at 25°C might drop to 19–20% at 65°C (which is what panels reach on a hot summer day). That’s the temperature coefficient — every panel has one, usually -0.30 to -0.40%/°C for modern mono. Phoenix loses 4–5% efficiency to heat across a year compared to STC.
Soiling: Dust, bird droppings, pollen, wildfire ash. A typical residential install loses 2–5% to soiling annually if not cleaned. Industrial installs in dusty climates lose more.
Spectral mismatch: STC uses AM1.5 spectrum. Your panels see different spectra at different times of day and seasons. Most panels are optimized for AM1.5, so you lose a couple percent when the actual spectrum differs.
Angle of incidence: STC measures at perfect perpendicular sun. Your panels rarely get that. Fixed-mount loses 5–15% per year compared to ideal.
Stack all of these and a 22% panel can deliver 16–18% real-world system efficiency. The YouTube tier list comparing 22.8% and 22.6% is comparing numbers that disappear into the rounding error of how your roof actually operates.
What to actually do with this
If you want to compare efficiency meaningfully, look at three things instead of just the headline number:
- Temperature coefficient. Lower (less negative) is better. A panel with -0.30%/°C will outperform a -0.40%/°C panel in any climate hotter than 30°C, even with slightly lower STC efficiency.
- NOCT or PTC rating. Some manufacturers publish a “Nominal Operating Cell Temperature” or PVUSA Test Conditions rating using more realistic conditions (45°C, 800 W/m²). These numbers are 12–15% lower than STC but more useful for predicting actual output.
- Degradation rate. Premium panels degrade at 0.25–0.40%/year. Budget panels degrade at 0.50–0.70%/year. Over 25 years, that compounds — premium at year 25 outputs ~91% of original; budget outputs ~84%. The 2% headline difference at year 1 gets dwarfed.
The headline efficiency number on a brochure is roughly the equivalent of a horsepower number on a car spec sheet — directionally meaningful, terrible for predicting real-world performance.
FAQs
If 33% is the theoretical limit, how does anyone claim higher?
Multi-junction cells stack different semiconductors absorbing different parts of the spectrum. Tandem perovskite-on-silicon has hit 33.9% in lab. Not commercially available for residential yet, but coming.
Is a 22% panel really worth more than an 18% panel?
If you have limited roof space, yes. If you have abundant space, mostly no — the 18% panel is cheaper per watt and the system output difference is small. Calculate watts-per-dollar, not efficiency.
Why do some manufacturers report “cell efficiency” and others “module efficiency”?
To make their numbers look better. Cell efficiency is always higher. If a spec sheet doesn’t specify, assume cell efficiency. Module efficiency is the number you actually want.
Does panel orientation affect efficiency?
Not the efficiency rating — that’s measured perpendicular to the sun. It affects real-world energy output. South-facing (northern hemisphere) at latitude-adjusted tilt is optimal.
Will efficiency keep improving?
Single-junction silicon will improve slowly — probably 0.1–0.2% per year until it hits a ~26% practical ceiling. Tandem perovskite-silicon will jump to 30%+ in commercial residential within 5 years. Worth waiting? Probably not — the energy yield of a 22% panel installed today is meaningful electricity for 25 years.
The landing
YouTube tier list guy is selling you certainty in a domain that doesn’t have any. Solar efficiency is a real spec, but it’s one of fifteen that determine how much electricity comes off your roof. The honest answer to “which panel is more efficient” is usually “both, but in different ways, and within rounding error of each other in real conditions.” Buy on temperature coefficient, degradation rate, and warranty. Ignore the half-percent gaps that brochures fight over. And if you find yourself on a tier list video for more than four minutes, close the tab.