Fridge efficiency: better full or nearly empty?

[LifesBetterWhenYoureBeating]

I've heard from several knowledgeable people that a fridge is more efficient when kept reasonably full than when nearly empty. It makes sense to me as any food/drinks would retain heat longer than air and thus the temp inside the fridge would take longer to drop and cause the thermostat to start the compressor again. Is that really what's going on?

Now, when I'm at anchor and there are grocery shops nearby I never have that much stuff to fill my fridge so I keep a dozen bottles of water there in hope that they'd make it more efficient, although I haven't confirmed whether or not that works. But now I've started wondering if there's anything that would be better than water bottles for this purpose? Maybe Styrofoam blocks? Something else? Or is the efficiency gain not worth the trouble?

[rslifkin]

Full will generally be more efficient than empty, as the compressor cycles will be longer and less frequent (takes longer to warm up and also to cool back down). Which means you get more cooling time for each pump-down cycle the compressor goes through, leading to slightly better efficiency.

However, if the fridge isn't full, lining the sides with additional insulation to take up some volume and reduce heat loss would be better than adding more food (to slow heating and cooling). But adding more contents is likely more practical in most cases.

In most cases, the full vs empty difference in efficiency won't be massive, so I wouldn't go crazy worrying about it.

[PaulCrawhorn]

Thermal MASS if the contents is what gives efficiency gains from less frequent on/off cycling.

Smallish water bottles gives good flexibility, add as the initial contents are depleted, bottles of air or coir will do nothing.

But to the degree this mass is warm when you put them in, the ROI requires a longer period, so only makes sense if you are days away from restocking.