How increased criterion-related validity increases the odds of making good hiring decisions. 
Next
Back
In the lower left quadrant are true negatives, or applicants who did not score high enough on the test and who would in fact have been poor employees. Together, true positives and true negatives are often referred to as “hits.” These are accurate hiring decisions.
In contrast, the employees in the lower right quadrant are false positives, or applicants who passed the test but did not turn out to be good employees. 
On the upper left are false negatives, applicants who failed the selection procedure but who would have been good employees. Together, false positives and false negatives are referred to as “misses” because they are incorrect hiring decisions.
Notice what happens to our decisions when we increase the validity of the selection procedure to near perfect validity. Here the scores hug tightly to the regression line. In essence, there is nothing but true positives and true negatives (hits), and almost no misses at all.  In other words, increasing the validity to near 1.00 will allow us to make almost no errors in hiring. (Of course, perfect validity is something that rarely exists in the real world).
On the other hand, in this situation with 0 validity, the quadrants containing the true positives and true negatives are the same size as the quadrants containing the false positives and false negatives. In other words, we are just as likely to make a bad decision as a good decision when there is zero validity.
In summary, increasing the validity of a test leads to more good decisions. In contrast, with zero validity you are as likely to make a bad decision as a good decision. This is why increasing the validity of selection procedures is so important to organizations.
Replay
One system for considering how selection decisions are made is to think of decisions as falling into one of four categories or quadrants, as shown below. The figure shows a passing score on a test on the x-axis and a point of acceptable job performance on the y-axis. The figure shows a situation of moderate validity.  In the upper right quadrant are true positives, that is, applicants who pass the test and who turn out to be acceptable performers on the job.