As a reader of this site, you may be familiar with my analysis of how well the various baseball projection systems performed each year for fantasy baseball purposes. I came to the conclusion early in my time of doing that analysis that simply averaging together the various projections can yield great results. That revelation eventually lead me down the path of trying to find the optimal weighting of the projections to achieve even better results. That work has led to the creation of my Special Blend of projections and the 2015 version of those projections are now available.
I analyzed each statistic and the how well the projections performed in each of those statistics so that I could average the projections together but weight them differently for each statistic. This Special Blend has performed better than any of the projections that are involved within it. For 2015, I tweaked the weightings a little bit more now that I have last year’s projections added to my dataset to research. Those tweaks should make this year’s special blend of projections even more special.
The projection sources that go into the Special Blend are:
For each statistic in the projections, a combination of some of those projections is used and weighted based on how well those projections have performed in the past. Without the hard work of all of those involved in making those initial projections, this special blend would cease to exist. Many kudos to them.
While these projections have most of the big names for fantasy baseball this year, it’s possible you may see some names missing from this list. This will happen if one of the projection sources did not provide a projection for a player. That is often the case with rookies or debuting international players. All sources involved here must have a projection for a player in order for this special blending to work.
The projections for 2015 are listed below with a column for a calculated WERTH roto value for a standard 5×5 roto league. You can also access them through a Google Doc here and you are able to download them from there.
Enjoy browsing and utilizing these projections! I will be adding them to my fantasy baseball cheatsheets in the next release.
Updated: 03/15/2015
Brian Scully
03/02/2015 at 1:44 PMI've really been looking forward to these. Thanks for all the hard work. One comment: Eric Hosmer is listed 9 times on several different teams and positions. I think there may have been a copy-paste issue or row insert issue that carried his name over in place of other players.
Luke
03/02/2015 at 2:48 PMWhoa, good catch! I'm not 100% sure why that happened initially but I just made the changes to get rid of those phantom Hosmers.
Jtm
03/02/2015 at 3:58 PMMany thanks, Luke! I understand Steamer had kind of a down year in 2014. Can you talk about how that affected its weighting in your 2015 projections?
Luke
03/03/2015 at 3:04 AMGood question. There were some slight changes to the weighting of each of the stats, just based upon having another year of data, but I did use historical data too and Steamer is still a primary source in a lot of different stats because of how strong it has performed historically.
Andrew F
03/03/2015 at 3:28 PMCurious as what you weighted and where, particularly in terms of playing time. Did you rely exclusively on MORPS/Fans for playing time? If so, how did you translate their playing time estimates into other system's projections?
Thanks for all the hard work!
Unknown
03/03/2015 at 7:47 PMThanks, Mr. Cheatsheet!
Quick Question/possible bug – I'm in the middle of a month-long slow draft. My league plays 4 OFs and I drafted Jose Bautista and Adam Jones thus far. On the "Team Summary" tab, Adam Jones is being listed as a bench player rather than a starter. This isn't a huge deal, but it is also screwing up the "Live Standings" tab in turn.
Luke
03/03/2015 at 8:15 PMIs this on the newest sheets that I loaded this afternoon?
Luke
03/03/2015 at 11:59 PMI did a combination for playing time as well (weighting and averaging a few different sources). Then, all stats that I'm using for the combination here are adjusted to a per PA or per IP level and then multiplied by the number of projected PA or IP to give the full projection.
Ben Gellman-Chomsky
03/04/2015 at 6:40 PMGreat work as ever, Luke, but whither Kris Bryant?
Ben Gellman-Chomsky
03/05/2015 at 1:40 AMDon't mind me, just read the last paragraph.
Joel Terry
03/07/2015 at 9:10 PMLuke, were these projections actually last updated over a year ago?
Luke
03/07/2015 at 10:34 PMLol, no. I meant 2015