I have failed you…or at least, I have deceived you…
At the end of my Reddit post for the last FGW update of 2021, I alluded to making a few posts in the offseason:
I definitely had those posts in mind. In my mind is where they stayed.
I have an excuse, though! As of May I’m officially a father of 3! Which is obviously great but, man, gonna be doing this on even less sleep than last year. Fuck me, right? Well, at least that’s what my wife said, about a year ago for sure. A bit less lately. Kids, man…I should probably hurry up before one starts crying.
How is FGW Calculated? Is it the same in 2022?
FGW has a new tweak for 2022 – call it the Derrick Henry adjustment. But first, for the uninitiated:
Fantasy Games Won (FGW) is the approximate number of Head-to-Head fantasy matchups a player has won you in a given season. Consequently, it is a superior metric to points scored for reflecting on the impact players have on your fantasy team.
FGW is calculated in 3 easy steps:
- Convert Points Scored to “Impact“
Start by assuming that a player scoring the average points for his position has no impact to your chance of winning in a week. For example, I have the average quarterback starting for a fantasy team scoring 21.1 points in a week. If your QB scores 21.1 points, that leaves you at 50% chance of winning your matchup in that week with all else being equal.
If your quarterback scores more or less than 21.1, you use the standard deviation of points scored by the typical opponent to modify that 50% number. I have a standard deviation of 17.6 in my book. If your QB scores 38.7, that’s 1 standard deviation above the mean, and per the normal curve, you go from 50% chance of winning your matchup to 84%. That’s an impact of 0.34.
- Throttle Impact by % Started
I try to approximate % Started by using a combination of Raw % Started data and Expert Rankings.
You can’t use Raw % Started as a proxy for legitimate fantasy managers because inactive managers will skew the metric – see last year when Christian McCaffrey was still started in 12% of leagues in Week 4 when he was out for the week.
You also can’t use Expert Rankings because, no matter how good your equation is to go from Expert Rankings to % Started, Expert Rankings don’t tend to get updated when Game-time decisions are made for injured players. Especially for the surprise inactives.
Thus for % Started, I use either Raw % Started or an Expert Rankings equation – whichever result is further from 50% started. So we get 0% for the injured McCaffreys from Expert Rankings, but also more accurate data on the top end where most well-established high end players have a Raw % started around 99%…and most Expert Rankings won’t reveal that fact.
You multiple your Impact Score by % Started. So an impact of 0.34 for a 50% started player comes out to 0.17.
- NEW FOR 2022 – Replacement Player Impact (The Derrick Henry Factor)
Quick Story from 2021:
Derrick Henry was a beast in 2021, until he got hurt in Week 8. After that, because his % Start was always 0, his FGW total was locked, and he maintained a high score for the rest of the season.
Anyone that has played fantasy football knows that when your first round pick goes down, you tend to have a hole on your team. You spent a lot of draft capital on that bad boy and you’re starting a 10th rounder in his place now.
It’s not as big of a deal when your 7th rounder gets hurt, since you were on the fence about starting him anyway. But that first rounder – man, that hurts.
I have taken a stab at calculating replacement player score as a function of a player’s draft slot by running a regression on last year’s data. So Jonathan Taylor has a replacement player score of 10.5, while Cordarrelle Patterson is at 12.4.
I did look at the impact of week of the season, but that effect was surprisingly small and didn’t hold up in the regression analysis.
The final step is to take this replacement score, convert it to Impact like you did in step 1, and multiply it by the % Not Started of the player in question. So a replacement score of 17.5 for a QB becomes an Impact of -0.08, and multiplying by the % Not-Started brings the Derrick Henry Adjustment to -0.04. Final FGW for this hypothetical player is 0.13 for the week.
Enough With The Book Report! Who Won Week 1?
- Justin Jefferson is the man. Although, I can’t shake the fact that he was the clear #2 to Ja’Marr Chase at LSU. So Justin Jefferson is da man. That’s a bit less prestigious.
- Patrick Mahomes throwing to a corps of receivers you end up about 5 seasons into your Madden simulation where you keep letting the computer do your draft and free agent signings.
- Over/under for % of his 2022 points Saquon has scored so far…50%. There’s a good chance he’s already scored more than half of his points for the year.
- Josh Allen throwing to…Gabe Davis? Who goes from Gabriel to Gabe?
- Ja’Marr Chase with enough points to maintain the man status.
- I’m just going to mention AJ Brown here because DeVonta Smith, what a jerk that guy is. Anyone who followed FGW last year knows that DeVonta usually hurt your team when you played him, but fantasy experts kept saying all offseason how his season was pretty good. No reason why he’d continue to kill your lineup, right? He can take a step, it’s year 2! Nope. Still a jerk.
- Kyle Pitts is making me wonder if there’s a thing about Falcons #1 targets and a permanent inability to score touchdowns. He’s doing the Julio Jones thing.
- Dawson Knox at 74% started is very strange. I know he’s pretty decent, he’s just also Dawson Knox.
- Have a Zeke story in a minute.
- Rodgers starting off 2022 just like 2021, it seems.
- Same with Allen Robinson! Owner of the Allen Robinson Bear Pelt, ARob is undoubtedly the star of the 2021 bad fantasy players. He’s at it again! I cannot wait to track his % started over the next few weeks.
- Meanwhile, Mike Williams doing Mike Williams things.
I have to say, I’m pretty thrilled to have Season 2 going. Because Season 1 almost got stonewalled before it even got going, all because of a comment I made that was a bit too insightful for some:
After some time building Karma in AskReddit…
I was back online. But never will I forgive Zeke. He had too many fanboys and fangirls ready to vote me down. The time of the Zeke Positive Game Script is over. The time of the Byron Cobalt positive game script has begun.