This is the second installment of our projection series. We will be posting a series of articles in which we project the roto stats for all players who could be useful in mixed leagues. The projections and “Roto Ratings” for each player are available to those who subscribe to our premium content via our “Front Office” package. 

Big thanks to my buddy Brian Sager (@TheRealSAG) for helping me develop this idea and for talking through the whole thought process with me.

When analyzing a player’s Fangraphs page, you sort of automatically project the stats that you think the player will have in the upcoming season. But because you can’t memorize loose projections for 300+ players, you have to repeatedly go back to a player’s page and go through the mental process of projecting them again. But as a service to our premium content subscribers, I have decided to do the projections myself and make them available on the site.

The first step in the process is simply to project a range of possible outcomes for each player while assuming he plays a full season. To project those possible outcomes I use a variety of stats. To project batting average I factor in plate discipline skills (K%, BB%, Contact%, Swing%, Z-swing%, O-Swing%) and batted ball profiles (LD%, GB%, FB%).  To predict home runs I again use the batted ball profiles as well as HR/FB rate from past years. To predict runs, RBI and steals I consider past performance in those categories, stolen base success rate, and where a player will be hitting in the lineup. I also used 155 games as my standard “full season” to account for regular days off and the occasional day-to-day injury. For example, here is my projection for Evan Longoria assuming he plays 155 games:

But you can’t assume all players are going to play 155 games. So the next step is to approximate how many games you think the player will miss and then fill in those games missed with the stats of the type of guy you might find on the waiver wire. To find that replacement level, I took the stats from the third basemen that were owned in less than 70% but more than 30% of ESPN leagues at the end of last year and averaged their stats. A replacement level guy at third base will give you the following stat line over 155 games:

The next step is to take the stats you think you’ll get from a player for the amount of games you project him to play once you factor in injury risk. For example, I projected Longoria for 128 games, so I multiplied all my projections for him by 0.83 (135 games is 83% of 155 games). Then I took my replacement level stat line and multiplied all those numbers by the remaining 0.17. Then you add those two numbers together to get the final stat line you’d expect to get from 128 games from Longoria and 27 games from a replacement level player. Longoria’s projections after accounting for injury risk and a replacement player look like this:

After I got my final stat line, I decided to come up with a formula to use the projections to do rankings. This system I came up with is admittedly crude, but I think it does a pretty decent job of ranking the players.

I plan on ranking 200 hitters, so I took the 200 hitters with the most plate appearances last season and created tiers for each roto category. For example, 20 guys hit above .307 last year. The next 20 guys hit between .293 and .306. So if I projected a guy to hit .308, I assigned him 10 points for average. If I projected him to hit between .293 and .306, I assigned him nine points, etc, etc. Because I projected ranges, I used the midpoint to see which tier someone fit into. I projected David Wright to hit between .283 and .291. The midpoint there was .287. That fell within the third tier so I assigned Wright eight points for average. Below are the tiers I used:

After I assigned a player a point total for each individual category I added them all up and gave each player a score which I am calling their “Roto Rating.” Below are the projections and Roto Ratings for my top 20 third basemen. Enjoy!

[am4show have=’p4;p7;p3;’ guest_error=’Front Office’ user_error=’Front Office’ ]

[/am4show]