How does backward induction help solve negotiation problems, according to Rubinstein's model?

It involves anticipating future moves and outcomes to make optimal decisions at each stage, leading to subgame perfect equilibria.