What is FGW? Check it out here:
While everyone was cleaning up mountains of wrapping paper and tissue paper in order to have a tidy New Year’s Eve, I left the tissue paper for the dog to eat and focused on cleaning up the FGW metric in order to have a tidy playoff calculation.
Like the aforementioned wrapping paper, however, I ended up taking an aspect of the metric that needs attention, made a pile, and shoved it in the garage.
Two attempted updates to the metric this week. The successful one first:
1. Active In Playoffs Percentage
The intent of FGW is to measure how many games a player has won you in fantasy. If you’re not in the playoffs, a player cannot win you (or lose you) any games anymore.
I would agree with the assertion that playoff time is the time to transition away from games won to % chance of winning a championship – and I plan to add that next week. The calculations go hand in hand. But as long as the metric is for games won, this is the way to do it, as far as I’m concerned.
Similar to how I’ve grabbed somewhat of an “average” league by using 12-team half-PPR as the scoring system, I’m picking a playoff format that covers various formats seemingly as well as I can: 4-team playoffs out of 12, starting in Week 16. The reasons I’m not doing a 6-team playoff are because A. 50% chance of making playoffs is really the high end of what you’d expect in a league, and B. The bye week can get really complex. I’m betting that the bye week complexity is a juice-not-worth-the-squeeze situation.
To calculate playoff odds, I’m using a rough baseline that’s admittedly not based on a ton of data, but is tough to argue with being close to reality:
- 8-7 record: 33% chance of making playoffs
- 9-6 record: 66% chance of making playoffs
- 10-5 record: 100% chance of making playoffs
Then, given the average team performance of 7.5 wins, I’m adding a player’s dFGW through week 15 to 7.5 to give a “new average” for teams that owned that player.
The final calculation is giving a % chance of reaching 10 wins (>9.5), 9 wins (8.5 to 9.5), and 8 wins (7.5 to 8.5), and adding the associated chances of making the playoffs.
Example: Cooper Kupp
Regular Season dFGW = 2.45
Average Win Expectation = 10.05 (holy crap)
Chance of 9.5 wins or more: 58%
Chance of 8.5 to 9.5 wins: 17%
Chance of 7.5 to 8.5 wins: 12%
Playoff odds: 58% * 100% + 17% * 66% + 12% * 33% = 74% chance of owners being active with Cooper Kupp in the playoffs
- Obviously there’s still a chance that a 7-8 team makes the playoffs, or a 10-5 team misses them, but the chances are so low that they’re negligible if you’re using a probability distribution to approximate win totals, anyway.
- Fantasy season results are a bounded distribution (e.g. an average expected win total of 15 obviously doesn’t have a normal distribution associated with it, as you can’t go above 15 wins), but the averages we’re dealing with are close enough to 7.5 that I didn’t feel the need to incorporate a bounded distribution.
- For players that were likely dropped a lot during the season, it’s unlikely they were on the same team all season. For example, you probably just picked up Justin Jackson, rather than held onto him all year as a handcuff. This could probably be approximated with ownership percentages, but I’m going to punt this one for now an hope to capitalize on better field position in the future.
2. True Replacement Player Calculation
When a player like Derrick Henry goes down in week 8, you don’t get to just slot in an average starter. That crater in your lineup is likely filled by a sub-optimal starter. Thus, defaulting to 0 Fantasy Games Won is unfairly optimistic for players that get hurt – it’s more fair to attribute negative Fantasy Games Won to reflect this replacement.
This one I’m going to punt until next year. It’s too complex and you have to fucking nail it, too, or the whole house of cards comes crashing down:
Some of the reasons why my attempt at this incorporation failed this week:
- Not all drops in % start vs. 100% are equal. For example, a first round pick with a low % start is likely getting replaced with a lower round pick, but a lower round pick with a low % start is likely getting replaced with a higher pick. So while it’s fair to say “when you’re not starting Derrick Henry, you’re starting a below average player”, it’s also fair to say “when you’re not starting Alexander Mattison, you’re probably starting a high draft pick”. So you need a way of reflecting this nuance.
- Once you’ve figured out the above nuance, a player that was never started all year should have no impact to your chances of winning the playoffs. FGW needs to approach zero as draft capital approaches zero, so that dFGW (draft-adjusted FGW) for free agents is zero. So a lowly drafted player that maybe gets started once or twice shouldn’t automatically get a lower score than Taco Wallace from the Packers’ practice squad.
I’ll put it this way – I will address indirect FGW when I get this figured out.
Back to week 16!
Who Won The Week?
- The MANDREWS with another 20+ point week. He might come in with one of the highest % chance of winning championship numbers when it’s all said and done.
- OH SHIT DEEBO’S BACK
- This week in “where the fuck did that come from?”, Tee Higgins was doing his best Ja’Marr Chase impression. Notably the lowest Active % on this list, though his % started number means he’s probably active in a lot of consolation brackets.
- Back in 2018, I was taking a shit at work when I got an alert that Kareem Hunt had been cut by the Chiefs. I immediately picked up Charcandrick West, and as a result, didn’t have a roster spot to handcuff Melvin Gordon. Lo and behold, one of the best fantasy teams I’ve ever run lost in the playoffs to a Justin Jackson/Jaylen Samuels combo at RB. Jackson finally paid me back for that this week.
- Nick Chubb, another candidate for Consolation Backet MVP.
- This seemed like such an obvious week for the Herbert/Allen stack. F.
- If you started Russell Gage or Tyreek Hill, you knew this was a possibility. F.
- Meanwhile, James Robinson owners got screwed for the wrong reason, as he looked fine before the injury. F.
- With their active percentages of 74% and 67%, respectively Kupp and Taylor are the only players with the mathematical possibility of being the MVP this year. And it’s getting dicey for Taylor – with a rough Week 16, his active % is about to get slashed quite a bit.
- I got far enough with the “true replacement” factor to see that Derrick Henry would be somewhere between 1.1 and 1.3 for FGW if it had been incorporated. So, still worthy of a first round pick, if a bit of a stretch for #3.
- Is Deebo going to be a first round pick next year? It feels like Jefferson will be one, for some reason I don’t know if Deebo gets there. Mixon’s probably a top 5 pick. Andrews up to at least round 4.
- Kind of crazy that Terry McLaurin has averaged 75% started all year, when no other Project Mayhem WR is above 42%.
- For the 20% of Dak Prescott owners in the playoffs, and the 54% of them that actually started him, Dak had a great game this week!
They Were The Best Of Picks, They Were The Worst Of Picks
- Mark Andrews is up to the 5th best draft pick of the year. And he did it largely with a QB that had me trying to remember the guy Piazza replaced on the Mets.
- It’s still very strange to see Kittle on this list. I swear, I’m not counting pancake blocks for 1 point each.
- Not the list I’d have expected to see the Ole Miss wide receiver teammates on at the beginning of the year.
- Is the universe trying to teach us a lesson with McCaffrey and Barkley here? They’re 25 and 24, but have performed like fantasy assets 5 years older. Is the lesson that fantasy football is a hobby for the insane?
The Full List
- Judging by the fact that the best pick in Round 6 this year was arguably the guy who never played (Gus Edwards)…probably. We’re probably all insane.
Good luck to the survivors! Super weird to be playing fantasy football on New Year’s Day. We now get to confuse our championship years with the regular season, just like the NFL does!