Early 2015 Projections: Adjustment for Position

Earlier this week, Jeff Zimmerman presented Early 2015 Hitter Projections using Steamer and/or ZIPS averaged ROS projections. The main contingency at this time: all values are set to 600 plate appearances. If I had all the time in the world, I would go through the list and manually adjust the PA based on lineup position, career PA/G, etc, but I’m not that much of a Mensch.

The next day, Mike Podhorzer highlighted some of the surprises ranked in the top 30. Again the 600 PA contingency is clear as Rajai Davis, Jarrod Dyson and Corey Dickerson make the list although if Dickerson doesn’t get platooned, I (and Mike) think he’ll surpass expectations. His splits page tells us there is no good reason to platoon him.

In Mike’s intro, he also referenced that there is no adjustment for position in Jeff’s SGP rankings. That’s where this post comes in.

Using Zach Sanders’ Fantasy Value Above Replacement (FAVRz) approach to fantasy valuation, I associated z-scores to each 5×5 category: HR, SB, R, RBI and BA. The entire sample was used for the z-scores since everyone’s stats were extrapolated to 600 PA. I won’t get too into the approach since you can read Zach’s 4-part strategy and improvements, but I will mention the following as there is a small amount of subjectivity:

• In part 2, you can read how Zach defines replacement level (based on the number of players drafted within each position), but for this, with advice (and zeal) from Zach, I went with slightly updated numbers per position: C=14; 1B=24; 2B=18; 3B=18; SS=18; OF=72. The rest of the 23 man rosters (x 12 teams) would be filled with 76 SP and 36 RP.

After we get all the 5×5 z-sums, the following players with their associated z-sums are our “Replacement Players” at each position:

As you can see, SS is our “most scarce” position followed by 2B in this format where middle infielders are required and a 2^nd catcher is not. Prior to next season, I (or other contributors) can run the FAVRz for different league/roster formats. Let’s just say off the top of my head, in 2-catcher formats, the new “Replacement Level” is John Hester according to these rankings at a -.49 z-sum meaning every catcher would have .49 added to their z-sums instead of 1.20 subtracted as noted in the grid next to J.P. Arencibia above.

Sorting by (Ascending) Position:

• The embedded file at the bottom of this post is sorted by the best FAVRz, which belongs to SS, Troy Tulowitzki with 600 PA. However, if you sort by column C/ “Position,” you will see that within each position I highlighted the replacement player in yellow. I randomly bolded some names in each position under replacement level that I like and should jump in value if we were to decrease the PA totals for some of the names that may surprise you above replacement level.

Position-related contingencies:

• Each player is associated to one position. I didn’t incorporate position flexibility, but if they wind up in a different position, you can simply adjust their FAVRz (columns H & I) by the other position(s)’ replacement value. Their additional positions are hidden in columns D and E. Simply download the embedded file and unhide if you would like.
• Carlos Santana – I’m making him a contingency on his own. I decided to stick him at 3B since A) he’s 3B eligible in 99% of leagues next year; B) 3B is a more scarce position than 1B and C) he won’t be catcher eligible in many leagues unless he plays there in 9 more games this year. How much more scarce is 3B relative to 1b? Simply look at the above “Replacement Players” grid.
• There are a few DH’s I stuck in the 1B category – David Ortiz and Zach Walters for example. Chances are Zach Walters may gain some type of infield eligibility in many of your leagues, but right now he only has 3 games at 3B, 4 games at MI, and 4 games in the OF. Therefore, I considered him a DH at the moment and simply stuck him at 1B. After-all, the Indians also have Lonnie Chisenhall, Carlos Santana, Francisco Lindor and Jason Kipnis consuming the middle/left side of the infield. He has spent the most time thus far at DH (11). David Ortiz is not a 1B, but I’m not going to de-value just him. Let others de-value him as a veteran utility. I kept positions I somewhat adjusted in red font (download the file and position 1/column C is shown; if applicable, position 2/column D and position 3/column E are hidden).
• I debated on what to do for the 1B-OF combos of the world. Let’s just say I put Brandon Moss and Steve Pearce in the 1B slot and Mark Trumbo and Allen Craig in the OF slot. I did this both objectively based on playing time next year and also to adjust the replacement level for 1B and OF as I deemed appropriate: if I put all of these guys under OF, it turned 1B into a more scarce position than 3B based on the replacement level.

Finally, here is the embedded grid with Jeff Zimmerman’s initial SGP value in the 3rd column followed by their 5×5 z-sums in the 4th column and their position-adjusted, “FAVRz” static values in the 5th column. Again, you can adjust the PA in column 6, which will adjust all the z-scores, but the FVARz will still be static/adjusted by the same highlighted replacement player in yellow. You can manually adjust that aspect in column H (non-static FAVRz) once downloaded. Feel free to comment below for support if I baffled you. Again, it’s sorted initially by the 5th column that you see: static FAVRz/Fantasy Value Above Replacement:

In addition to the positional adjustments highlighted in red, I also noted some of the surprises that Mike Podhorzer and a reader mentioned. Basically, through a positive BABIP regression, Alfonso Soriano doesn’t kill you in the BA department while providing lofty counting stats – again, with the assumption he somehow gets 600 PA, which he won’t. Similarly, but to a greater degree and at SS, Javier Baez mashes his way into the top 20. The same goes for George Springer.

…If only Michael Cuddyer would stay healthy in Colorado.