To solve a problem we have to describe what is happening.
We define the key components and dimensions of the system:
- What is the market?
- Who are the players?
- What service are they providing?
- The way we express a problem, and the terms in which we define it, are critical to how we solve it.
Once we define the key variables, we can start to quantify them – to express them in numerical terms. This allows us to answer questions like: How many objects are there? Are they increasing, or decreasing in number? If so, how quickly? These types of questions are simple, but fundamental to describing a problem.
Visualising a data set helps us to describe what is happening, and allows us to build an intuition about the relationships between its key variables. A single well-designed chart can often demonstrate a problem exists, and suggest a solution.
Charts are powerful. We like making them.
Once we have described a problem we can understand it through the construction of models. These models are a latticework on which a new more advanced intuition can grow
A model is a simplification of the system it seeks to represent. It is predicated on assumptions – eg, buyers and sellers have perfect information, agents are rational, markets are frictionless etc.
Making assumptions means making choices as to which characteristics of the system we simplify, and which we try to capture. Good choices ensure that our models are relevant.
A model is a representation of a system. Factories, highways, and power lines can all be represented as numbers, symbols, and pictures. An entire supply chain can be represented as a formula: what goes into the process, and what comes out.
Models allow us to distill a complex system into its principal components, the parts that matter. We can then gain a better understanding of the relationship between these key elements.
Models allow us to simulate – to test and investigate a system without affecting it. In the virtual world of the model, we can build a highway, shut down a power station, and even merge two rival firms.
A simulation is a laboratory for us to test our intuition, extend our understanding, and explore the limits of a system.
A model that can help us understand a problem is powerful. But models can take us beyond understanding alone. They can help us solve a problem, and make better decisions.
Endgame helps our clients make predictions:
- What will a carbon tax mean for electricity prices?
- How will batteries affect electricity networks?
- What will happen when two competitors merge?
- We build models that answer these questions, and so help our clients make predictions.
Many problems lend themselves to optimization. We have extensive experience in operations research and are highly familiar with a range of optimization tools and programming languages (ie, AIMMS, GAMS, AMPL, Python + Gurobi). We routinely prototype optimization models in a matter of days, and have particular expertise in electricity market modelling.
Ultimately, our models and advice help our clients make better decisions:
- We help governments formulate better policies
- We work with regulatory bodies to make better rules
- We help businesses make better strategic decisions.
Models help us move beyond intuition, beyond ‘gut instinct’ towards robust decision-making
// contact us
Are you trying to solve a problem?