There are probably pure-win half steps, like the kind of farming where you plant in the seasonal area you always come back to at a certain time of the year, as you follow the herds, or the kind where game is so plentiful you can afford to settle, hunt, and dabble in farming vegetables beside your settlement (such as in the American Pacific north west). Farming seems to be tied to settlement. Farms stabilize settlements; settlements nurture farms. And farms domesticate crops, making farming easier and supporting a larger population.

In the Mesopotamia region, there were settlements in the rainy hills where the local wildlife was conveniently easy to domesticate but farming was hard. Those moved down centuries later into the rainless flood plain between the Tigris and Euphrates, where only group effort could ensure irrigation, and group surpluses were needed to stave off bad harvests, but farming worked well. The "Ubaid period" (neolithic) was pretty egalitarian, but centralization emerges in the "Jemdet Nasr period" and kingship in the "early dynastic period" (Sumerian for king is "lugal", "lu"=man, "gal"=big, and initially it seems to have been just a word for "boss"). With centralization and kingship, empires follow fast. Civilization was co-existing with non-farming groups, but civilization tempts even non-farmers to switch from hunting to raiding. Sumer got sacked repeatedly by nearby tribes.

I am thinking there was a demographic transition point, probably quite early, when the number of people that could be kept alive - not as healthy, but alive - by farming or equally by raiding the surplus of farmers, exceeded the carrying capacity of the local game and wild plants. At that point walking away from the fields was not possible. Therefore agriculture has a ratchet effect.