Understanding traditional seismic strategies is necessary to conduct appropriate restoration of historic buildings and city centers located in earthquake-prone areas. But, the fast disappearance of vernacular built heritage and of the associated popular knowledge and know-how, along with the spread of inappropriate maintenance works, make it very difficult. To overcome these obstacles, it is suggested to broaden the scope of observations by including both larger geographical areas and elements that do not appear as seismic features at first consideration. This paper illustrates the relevance and the limits of this approach by applying it to the case study of Yazd historic city center. Yazd is located in a moderate seismic area and is famous for its well-preserved historic city center, with a high number of Qajar adobe buildings in various conditions. The authors combined literature review, direct observation of numerous buildings and a deeper study of four Qajar houses to conduct the first part of their research. The data collected mainly relates to the geometrical features, the materials used and the interconnections between elements at different scales. Their analyses led to the elaboration of several assumptions on the seismic behaviour of the studied adobe buildings and vaults and on the distortions some features may have suffered. To validate these assumptions, two methodologies were used. Both are illustrated in this paper through the study of the insertion of baked bricks in adobe vaults. The first one is based on observation campaigns on damaged houses. The second one is based on experimental works, with tests conducted on the shear strengths of masonry made with mud mortars and adobes or baked bricks to better understand how builders benefited from mixing these materials in an adobe vault. Finally, the authors discuss the limits of the method and the extent to which restoration projects can benefit from it.