The survivorship bias here reflects the positive-results bias in research itself. One neither wants to nor can publish negative results, despite their potential

The survivorship bias here reflects the positive-results bias in research itself. One neither wants to nor can publish negative results, despite their potential to be more valuable than massaged and hacked positive results. Not only is this unfair to unlucky research paths, but it also breeds a social media-like positive feedback loop wherein the individual researcher feels alone in failure.