Laboratory experiments to simulate landscapes and stratigraphy often suffer from scale effects, because reducing length- and time scales leads to different behaviour of water and sediment. Classically, scaling proceeded from dimensional analysis of the equations of motion and sediment transport, and minor concessions, such as vertical length scale distortion, led to acceptable results. In the past decade many experiments were done that seriously violated these scaling rules, but nevertheless produced significant and insightful results that resemble the real world in quantifiable ways. Here we focus on self-formed fluvial channels and channel patterns in experiments. The objectives of this paper are 1) to identify what aspects of scaling considerations are most important for experiments that simulate morphodynamics and stratigraphy of rivers and deltas, 2) to establish a design strategy for experiments based on a combination of relaxed classical scale rules, theory of bars and meanders, and small-scale experiments focussed at specific processes. We present a number of small laboratory setups and protocols that we use to rapidly quantify erosional and depositional types of forms and dynamics that develop in the landscape experiments as a function of detailed properties, such as effective material strength, and to assess potential scale effects. Most importantly, the width-to-depth ratio of channels determines the bar pattern and meandering tendency. The strength of floodplain material determines these channel dimensions, and theory predicts that laboratory rivers should have 1.5 times larger width-to-depth ratios for the same bar pattern. We show how floodplain formation can be controlled by adding silt-sized silicaflour, bentonite, Medicago sativa (alfalfa) or Partially Hydrolyzed PolyAcrylamide (a synthetic polymer) to poorly sorted sediment. The experiments demonstrate that there is a narrow range of conditions between no mobility of bed or banks, and too much mobility. The density of vegetation and the volume proportion of silt allow well-controllable channel dimensions whereas the polymer proved difficult to control. The theory, detailed methods of quantification, and experimental setups presented here show that the rivers and deltas created in the laboratory seem to behave as natural rivers when the experimental conditions adhere to the relaxed scaling rules identified herein, and that required types of fluvio-deltaic morphodynamics can be reproduced based on conditions and sediments selected on the basis of a series of small-scale experiments.