Some time ago I found a paper, written I believe by Conway, in which the author describes different operations we may perform on combinatorial partizan games and how these operations can be used in order to endow games with a partial order.
For example, the most classical operation is the sum $G+H$ of games $G$ and $H$ in which the players play by choosing a component each time and playing a legal move in that component. Then we may say that a game $G$ is better $H$ for the left player if any time $L$ can win on $H+X$ she can also win on $G+X$ and we write $G\geq H$ (I'm simplifying a bit).
This paper predates ONAG (1976) and (then) also Winning Ways. I found this paper iteresting because it justifies a bit more why we define the classical partial order on games like we do and also considers what happens if we use different operations (for example one may require the players to play in both components each time).
Any help is really appreciated (and I will download interesting papers at once in the future).