2016 Prospect Rankings

This past season may end up going down as the year of the prospect. So many good rookies got their promotion to the majors like Kris Bryant, Carlos Correa, Noah Syndergaard, and Lance McCullers. It is time to look at finding the next batch of rookies. By using just a handful of minor league stats, I have created basic prospect rankings for 2016.

I have played around with prospect rankings for a while and just recently publish my 2015 ZOBRIST rankings which looked at older potential hitter breakouts. I have had some personal player rankings using just the player’s stats, but nothing worth publishing. Well, I finally took the few step forward by stealing some ideas from a couple of my fellow writers. First, I was able to add position values to the rankings after helping Carson Cistulli work on his minor league WAR values. The final piece came after reading a recent article by Tony Blengino at ESPN where he gives credit to the player’s age compared to the level’s average age.

With these additions and bit finding the best weight for each category, I have come up with a prospect ranking system which I believe holds up darn good. For hitters it uses wOBA, most common position played, and age compared to the rest of the level. For pitchers, it is K%-BB%, the percentage of games as started (vice relieving), and age compared to the rest of the level. The player’s value combines his performance at all levels he was in during the time frame.

For each player, I came up with an unregressed value which shows the player’s value no matter how little they played. Then, I regressed the value a bit to come up with a  second value. Finally, I put the values on a 100 scale where a top player value is 100. For the 100 scale, I used the highest one-year total I could find. I can and did run the analysis for multiple years’ worth of data, so the scaled value can go above 100 in these occasions.

The process and results seem to hold up, but I did  have a few possible issues show up. First, the position used is the position the player was at the most. It doesn’t know if the player was good at the position and may need to move off of it. The second issue is injuries especially with pitchers. If a highly touted prospect like Kyle Zimmer is unable to stay off the DL, they get hammered quite a bit. Third, players just drafted usually don’t have enough playing time to get highly ranked. The final change I could add is some form of prospect pedigree using  their draft slot and/or prospect ranking. I am torn about using any prospect ranks as I am trying to find players people may have missed.

So here are four lists with all the data downloadable from here. Finally, let me know if you have any questions especially if you think any of the three weights I used (age, position, talent) were mis-weighted. That is it for now and tomorrow I will look at where some players in the Arizona Fall league are ranked.

Hitters: 2015 Only

Hitters: 2013-2015

Pitchers: 2015

Pitchers: 2013-2015



Print This Post

Jeff writes for RotoGraphs, The Hardball Times, Rotowire, Baseball America, and BaseballHQ. He has been nominated for two SABR Analytics Research Award for Contemporary Analysis and won it in 2013 in tandem with Bill Petti. He has won three FSWA Awards including on for his MASH series. In his first two seasons in Tout Wars, he's won the H2H league and mixed auction league. Follow him on Twitter @jeffwzimmerman.

newest oldest most voted
MustBunique
Member

Remember when you posted this awesome information? That was awesome.

MustBunique
Member

I’m high on Arcia as well. What do you see his PT situation being in 2016?

brad
Guest
brad

oh and Berrios rank gives me good feelings too. keeper!

brad
Guest
brad

oh I love seeing Arcia higher than Crawford
i made the trade for Arcia more SB’s gotta have speed

MustBunique
Member

Agreed, which brings me to a thought. There is probably some value related to how soon a player will play MLB games. Maybe players with a higher probability of playing a lot next year deserve a bump? I realize there is a lot of uncertainty related to this but some sort of probability scale could be developed and applied. Oh, and thanks for ruining my productivity this afternoon. Always enjoy your posts, this one is especially interesting.