Premier League Rest-of-Season Player Projections

With the Premier League finally coming out of its Coronavirus pandemic suspension next Wednesday, I thought it would be a good time to work on rest-of-season projections to help those who are looking to make late runs in their season-long leagues, hold onto their top spots or who are starting new leagues for the congested schedule. This felt like a very good idea until I actually started doing the work, but that wasn't going to deter me from trying.

There are obviously plenty of obstacles when trying to project what is going to happen, or in actual terms, provide median outcomes for future events, but the biggest issue in all of it was estimating how much each player was actually going to be on the pitch. Projecting substitutions on a week-to-week basis is enough of a hassle, but we're now confronted with five possible subs per match, plus a condensed schedule that will force lineup rotation much more than we're used to. But again, this would not deter me from trying.

Below is a breakdown of my outfield player projections for the remaining 92 Premier League matches of the 2019/20 season, specifically using the standard scoring rules on Fantrax, one of the more popular draft fantasy Premier League games. I am not going to break down my entire process, but I will say that I attempted to project each game on its own and then distribute the totals to each individual player based on their projected minutes. For example, I projected team totals of individual stats for each Manchester City match for their remaining 10 matches and then distributed the totals among the players based on how many minutes I expect them to play. As you can imagine, the minutes estimates are mostly guesses, but there is real math behind the distribution of the stats to each player. So, if I project someone like Mohamed Salah for 600 minutes over Liverpool's remaining nine games, his projections would change if I adjusted them down to 550. Additionally, each of his teammates would have their projections adjusted, not just the players who have to make up those 50 minutes.

Each column below is sortable, so you can see who I project to score the most goals, create the most chances or win the most tackles. Additionally, you can sort by team to see the distribution throughout the roster. I am very open to any feedback when it comes to the minutes and am happy to re-run the numbers if there's a general consensus that I am way over-estimating or under-estimating certain players. Just remember that if minutes are removed from one player, they have to be redistributed the others. I will also make any changes if injuries occur or if I just flat out missed someone because I've been staring at numbers in a spreadsheet for more hours than my eyes (and brain) are used to.

One last point: I didn't make any manual adjustments to the numbers. There were some figures that jumped out to me because they felt odd, but I didn't want my personal bias to come into the numbers, even though I'm the one who developed the formulas that gave me the figures. Any manual adjustments would have felt unfair, so I decided to avoid it (that's my easy way of saying "that's just what the numbers say!" when one looks wonky).

As always, if you have any questions or comments, please feel free to leave them below or hit me up on Twitter @RotoWireAndrew. 

Legend: FPTS = fantasy points, G = goals, A = assists, SOG = shots on goal, CC = chances created (key passes), ACRO = accurate open-play crosses, DR = successful dribbles, AW = aerials won, DIS = dispossessed, TKLW = tackles won, INT = interceptions, CLR = claerances, YC = yellow cards, GC = goals conceded, CS = clean sheets