Each spring, we run the top 100 prospects based on the ZiPS projection system. These differ from Keith Law's because they're solely data-driven. While there are lots of ways to be clever about getting scouting information into data, a computer can't see all and doesn't know if, say, a pitcher's stats are weak because he was experimenting with a new grip on a pitch or if a slugger was playing through a sore wrist. But computers do see a lot of things, and they're quite good at sifting through large amounts of data to find patterns.
Where the computer projections are most interesting tends to be where they differ, so I want to focus on the players that ZiPS sees differently than Keith does. A lot of times, Keith drops the law on ZiPS, Judge Dredd-style, such as in the case of A.J. Reed or Jose Peraza -- both players ZiPS liked last year -- or Brendan Rodgers, a player ZiPS was cautious on. But ZiPS also scores some wins, liking Trea Turner (No. 11 vs. No. 28), Jonathan Gray (No. 33 vs. unranked) or Steven Matz (No. 10 vs. No. 37) from a data-driven standpoint.
One thing that you'll find different is that ZiPS tends to like low-ceiling guys more if they look to be decent contributors -- players like Kolten Wong. A few players are unranked in ZiPS simply because they don't have any professional (or very limited) play and/or no college-level experience. A computer projection system without data is like a cheesesteak without meat.
Note: Full Top 100 ZiPS prospect rankings are at the bottom of this column.