These kinds of images were likely made by a combination of limited available light, exposing for a workable shutter speed/aperture, and then controlling the final look in processing and especially printing.

From the answers, the most plausible approaches are:

• shoot in low light at the slowest shutter speed / widest aperture you can tolerate
• accept some underexposure on the negative
• sometimes rate the film faster and push-process it to compensate
• or simply print a thin negative darker in the darkroom

There’s no single “magic” amount of underexposure. The usable amount depends on the film, developer, and the look you want. In practice, photographers often chose exposure settings based on avoiding camera shake or getting enough depth of field, even if that meant the negative would be thin.

Also, “dark overall tone” is often a printing choice, not proof that the negative was heavily underexposed. Modern viewers often expect midtones to sit in the middle, but that’s an aesthetic convention, not a rule.

If using a lab, you could ask for push processing if you intentionally rated the film at a higher ISO/ASA than box speed. But the final low-key look is often best controlled when printing or scanning rather than relying on development alone.