tl;dr: The landing site was changed late in the mission planning sequence (after the Critical Design Review). As a result of the landing site change, the lander went over a crater which caused the distance to the surface to change significantly. This was detected by the laser altimeter, but the altimeter reading was discarded because of the rapid change (assumed to have failed).
As a result, "the lander was still about five kilometers above the surface when the lander’s computer believed it was on the surface. The lander continued to descend at a slow rate of about one meter per second under its thrusters until it exhausted its propellant. The lander then went into freefall, crashing into the surface at a speed of more than 100 meters per second."
As Scott Manley demonstrates, this is painfully familiar to players of Kerbal Space Program. You’re heading toward a landing site and overshoot into a giant crater, and now instead of being 5m from the ground you are 15m… 20m… 80m… low fuel… maybe if you leap out the hatch and use your jetpack you can save at least one of them…
Noteworthy: The video mentions that this was the 3rd very-recent lunar lander to crash because of software failures. Compare that to the Apollo program, with human pilots, and a lethally dangerous training regimen for those pilots - https://en.wikipedia.org/wiki/Lunar_Landing_Research_Vehicle
Which training regimen was credited by Neil Armstrong - who very nearly died in one of the LLRV crashes - as absolutely essential to the success of the Apollo moon landings.
Suggestion: Near-future unmanned lunar landers should default to remote control by a heavily-trained pilot on the Earth, and only be allowed to manage the final ~5 seconds of their landings.
This reminds of what happened to a Mars lander when one software team was doing calculations with the metric system and another team was using English measurements.
tl;dr: The landing site was changed late in the mission planning sequence (after the Critical Design Review). As a result of the landing site change, the lander went over a crater which caused the distance to the surface to change significantly. This was detected by the laser altimeter, but the altimeter reading was discarded because of the rapid change (assumed to have failed).
As a result, "the lander was still about five kilometers above the surface when the lander’s computer believed it was on the surface. The lander continued to descend at a slow rate of about one meter per second under its thrusters until it exhausted its propellant. The lander then went into freefall, crashing into the surface at a speed of more than 100 meters per second."