That's the only empirical way, yes. We can also just talk about what it would mean in theory, though. In this case, we'd say that AlphaGo is well trained to the

That's the only empirical way, yes. We can also just talk about what it would mean in theory, though. In this case, we'd say that AlphaGo is well trained to the training data set sampling distribution but that may be far from the actual world game distribution.