It is a good question. And I think you could get different answers because it is a thing that is dynamic and still evolving. Fielders have always "shifted" their positioning but at one time it was based on observed and recorded tendencies, it still is but modern technology has added a lot more data to the numbers being crunched and teams are emboldened to go beyond what they at one time weren't.
In order to judge the veracity of this, or any analysis that states simply that the "shift" saved, or cost X number of net runs one would first have to know what constitutes an "instance". Once that is defined, then each and every ground ball hit during a shift has to be judged as:
1) hit into the shift for an out that would not have happened without the shift
2) hit away from the shift for a hit that would not have happened without the shift
or
3) would make no difference
Then, one must look at 1 and 2 and determine if the allowance or prevention of a hit led to a run or the prevention of a run that would otherwise have happened. Then, of course, one sums up for the net effect.
Seems like some room for variance. I would like to see the methodology before I could even hazard an opinion.
