KRSESQ: leeksfromchichis: It's not like they send these mars probes to randomly land.  The LZs are carefully chosen.

I recall hearing at the time that Viking 1's landing zone was selected because it was thought to be relatively level and obstacle-free and "safe and boring." Boy, were they surprised.

I found a NASA tech report: "This area was chosen because there were a minimal number of impact craters and, therefore, a minimal number of blocks that might be hazardous to the lander; radar data showed intermediate reflectivity values of about 8 percent, a value close to the Mars average."

It then goes on to note that, in retrospect, the distribution of hazardous blocks cannot be discerned from orbital imaging or Earth-based radar:

Nor was it possible, with the Viking orbiter image resolution at 1500-kilometer periapsis elevation, to delineate the hazardous block populations shown in the lander photographs. Higher resolution images are needed to determine such local hazards and to interpret the processes that formed the rocks seen at the landing site.
In retrospect, several pertinent observations about the use of the available instruments can be made. The orbiter imaging photographs were used to determine nearness to large impact craters and the inferred associated blocks; nearness to craters less than 100 meters in diameter could not be determined with this camera from this orbital altitude. Determination of surface roughness by orbital observation depends on both the actual distribution of roughness and the response of the instrument. It is quite possible to have surfaces that appear rough in the orbiter images but smooth in the lander images or smooth in the orbiter images but rough in the lander images. In the Lunar Orbiter and Apollo Programs, small blocks and slopes were observed on the Moon by f l y i n g appropriate focal-length cameras with ultra high resolution film at sufficiently low altitudes. Although Earth-based radar data could be used to avoid areas where reflectivity values were low and root-mean-square slope values were high, these data could not absolutely certify the absence of hazardous blocks. The Earth-based radar could be used to obtain depolarized echoes on Mars to help interpret block abundancies at a site. Later attempts to obtain depolarized echoes of the site (needed to resolve block frequency) were unsuccessful. It was also difficult to verify block size and frequency on the Moon from Earth-based radar (refs. 19, 20, and 21) although depolarized echoes were more easily obtained there and were used for interpretations of block frequencies (refs. 22 and 23). Infrared observations of the proposed landing sites under predawn and noontime conditions might have helped to resolve the nature of the radar anomaly at the A1NW site (23.4° north and 43.4° west); however, this would have required orbiatchanges. Time constraints and uncertainty about the applicability of the results ruled out any orbiatchanges. The most direct method for determining block populations would be a much higher resolution (about 1 meter as was used for lunar site certification) imaging system supported by IR and radar observations. A radar altimeter for confirming the actual elevation would also be useful.