On literally every project I've seen that enforces a % threshold, that % threshold has directly led to bad quality tests. This is more true the higher the % is, with the worst being 100%.
You can argue that it _shouldn't_ be this way, and I would agree, but it is that way. Perhaps in part because developers are humans, and even humans with the best initial intentions will game metrics with a large enough sample size over a long enough time horizon.
You can argue that it _shouldn't_ be this way, and I would agree, but it is that way. Perhaps in part because developers are humans, and even humans with the best initial intentions will game metrics with a large enough sample size over a long enough time horizon.