You can guard against the metrics being inaccurate by calibration and anchoring. And also verification. (and watch the watchers too) It is obviously an active

You can guard against the metrics being inaccurate by calibration and anchoring. And also verification. (and watch the watchers too) It is obviously an active process, much like guarding democracy against demagoguery. Attempts to subvert the system can usually be spotted early. A person designing right secondary metrics that result in good primary endpoints is also meritorious. The usual problems are either complacency, where meritorious people stop being so and are not being degraded, stagnation - lack of further development of merit or unnecessarily high barriers of entry. Merit when developed should accrue over time, yet it should be possible for a "nobody" to get in at high levels.