The paper tests this hypothesis and finds that there's probably a causal effect:

"...these results demonstrate that the screening effect may have played a role, but it appears insufficient to entirely account for the observed difference between near misses and narrow wins."