Hi guys,

So a couple of weeks ago I made a post discussing various ways to rank and compare player performance in Roto formats, given that there isn't a straight-cut system to do so like there is for points leagues (you can find that post here). In the post, I used excel to examine different methods for ranking players from their 2020 stats, and all three had their successes and their flaws. Many people commented about the use of the Standings Gain Points method (SGP), which I was unaware of and basically was a better system than any of mine, so I have been playing with that since.

Now that I have a pretty strong grasp of how the SGP method works and how it can compare to players fairly accurately, both between hitters and pitchers, but one thing still bothered me. In terms of comparing purely player talent, this method seemed to work. But isn't a top 5 catcher much more valuable than a top 5 SS? In my opinion, catchers seem to be amazingly undervalued, given both requirements on team rosters and the talent drop off after the elite players at the position. So I set out to solve that issue.

Here's what I came up with: I want to take the SGP score of each player, and subtract what I would consider being the value of a replacement-level player at that position. Replacement level, in fantasy terms, would be the best player that would be on a starting lineup. So in my 10 team league with 2 C, a CI, MI, U, and 5 OF spots, the replacement level players would be the 21st C, around the 16th 1B, 2B, 3B, SS, and around the 51st OF. It won't be exact rankings because there can be different positions to fill the CI, MI, and U spots differently, but it will usually be around there for each position. Now that I have those values, I simply subtract the positional replacement scores from the SGP scores to find each players' value above replacement (VAR).

The 2020 results were interesting, to say the least. Using just the SGP method, the top 5 players of 2020 were Marcell Ozuna, Trea Turner, Jose Abreu, Freddie Freeman, Jose Ramirez, all of which I would for the most part agree with. After taking into account the positional adjustments, Ozuna was still #1 but was now followed by Mookie Betts, Abreu, Ramirez, and Freeman. The positional adjustment valued OF a bit higher and lowered the value of SS and 1B. This was likely due to the talent level at SS being arguably an all-time high lately, and additionally many perenially elite OF having below-average seasons (much of that can be attributed to the small sample size of 2020).

Perhaps the most interesting, but also most expected, adjustment on the hitting side was at catcher. Salvador Perez, who ended 2020 as the most valuable fantasy catcher, ranked as the 31st overall hitter. After the adjustment, he jumped all the way to 10th. There were similar increases for all of the top catchers (Travis D'Arnaud went from 38 to 11, J.T. Realmuto jumped from 44 to 16), so as a result, 5 catchers were among the top 30 most valuable hitters. This was the exact result I expected to see, both due to the depleted talent level at catcher, along with the fact that my league starts 2 catchers, so the replacement level catcher would be the 21st catcher, much worse than the replacement of any other position (most other positions were around the 75-125 range; the replacement catcher was Wilson Ramos, the 240th best hitter in fantasy).

I feel that this method is a much better tool when assessing player value, given that I believe a strong hitting catcher can bring so much more value to your team than anywhere else. I'm not sure if I am now ready to draft Realmuto or Perez as one of the top 15 hitters, but it just gives me something else to consider.

One aspect that I did not adjust for was players with multiple eligibilities. I simply could not think of a good way to account for players that can be eligible at multiple positions without throwing off all of the other replacement levels, so I used only the position each player played the most of in 2020 or their last season in the majors. Please let me know if you have an effective way to account for this, as I am constantly updating my model for this upcoming season.

Thanks for reading, and obviously discuss in the comments your opinions on this ranking format, or if you have any other questions or comments. I obviously don't know everything (I didn't even know what SGP was about a month ago), so please let me know if there is anything I missed. Thanks!