Creating a model with limited reference and unknown depth

Hey everyone,
From most tutorials I have seen so far, models are created with a 2d reference image, usually side view, and front view sometimes. However, I’m up to a task of creating 3d models from very limited resources ( a front view only) and I’m looking for ways in which Blender may compensate this limitation by helping me determine the geometry via modifiers or something that can complete the shape. For example these two characters:

Z-leo: https://www.spriters-resource.com/fullview/37870/

Seven-force: https://www.spriters-resource.com/fullview/29710/

For the first one, I tried creating sketches to envision what he would look like from side view… They were no good as I couldn’t tell the depth here, nor could I find something with a similar shape.

For the second case, the issue is more complicated because it involves separate parts meant to be combined together to form different shapes themselves. I’m still experimenting with sculpting these shapes but once again my problem is limited perception depth.

Any insight would be really helpful.

Thank you

There’s no easy way for this but, if you start blocking the shapes- as in… for the red character- the middle of the chest is a sphere- and around it is a torus. And i think the thigh is like a stretched sphere from top and bottom… So you can worry about the other details later.

I think most of the time we waste alot of time thinking how to start without even initiating (which i also do alot)