The mainstream of economics is named neoclassical economics. Its success in 20th century economics is obvious and it remains dominant even in 21st century. One of the reasons why neoclassical economics so succeeded is the appropriate employment of mathematical methods. Two pillars of the neoclassical framework are optimization and equilibrium. Both permitted to transform economic problems into tractable mathematical formulations. But, the cost of this use of mathematics was not cheap. It enclosed the economics to the simple world of calculus and put aside many of important questions that economics should attack. Neoclassical way of explanation seems to be flourishing but it is an effort to make scarecrows from in one field to in another. They may be impressive for naive people, as the explanation is full of high mathematics. But, this is not the real way to develop economics. Problem setting itself is designed such that the problem thus obtained is soluble. By consequence, there are full of ridiculer explanations. To cite an example, the RBC (Real Business Cycle) theory is one. This "theory" claims that business cycle is caused by the technical shocks the economy incurs. How does it explain the present world wide depression? Is financial crisis also a technical shock? I cited many other examples in the text, but the main question is not this point. RBS is only an example, but of what? It is a symptom of the crisis of economics as discipline. Now it is the time to seek a breakthrough. We should build a new economics different from neoclassical economics. There are many schools in economics and all but few are for it. But the question is how? Most of the endeavors that schools of economics tried failed. Is it possible after all? The attempt is not an easy one, of course. It requires a firm strategy. It is not a question of changing this hypothesis to another. The trouble that neoclassical economics faces is that it is enclosed in a circle of mathematical tractability. This circle is very loose and hard to see it, but it is strong enough that it never permits to go beyond it, as far as economics continues to use mathematics as main methodology. By consequence, the new strategy must be as powerful and as omnipotent as mathematics. What I propose as such a candidate is MAS (Multi-Agent simulation). MAS may liberate us from the yoke of mathematical tractability. Process analysis becomes the major field of MAS. By mathematics, analysis of processes was difficult except for special types of time development, i.e. differential and difference equations. In MAS, varieties of people can be included. People of different income, of different taste, of different opinion, and so on. Goods can be also very different. Firms may be different too. Evolution of behaviors can be implemented. MAS may open a new stage of economics. It is time to bet for it.