SAG 2021 — Week 11 Report
“I arose, I settled myself before the orb, and I gazed into its crystalline depths… and what do you think I saw gazing back at me?”
This week fellow SAG-heads we climb into the divination attic of Commissioner Clue, attempt to read our future in the tea leaves (remembering to swirl the cup three times with our left hand), prophesies things to come in the tarot, and divine each team’s lot with the help of The Dream Oracle by Inigo Imago. That’s right, the SAG LAB+ has become so powerful that it is now attempting to forecast the future. Step inside the “Road to the Leonard”, powered by SAG LAB+.
A note of warning before proceeding. I’m not a statistician. The following is just one man’s attempt to make sense out of what is educated guessing on weekly player performance. I welcome all feedback on how to improve this. Please also note, not all feedback is equal. For example: if a reader had, say a PhD in Economics from a world-renowned academic institute, their opinion would carry more weight than a wine salesman, no matter how good a tune that wine salesman may be able to belt out.
The Leonard according to Yahoo.
Each and every week we put our faith and hitch our prospective fortune on the mysterious machine that is the Yahoo score predictions. It builds us up to knock us down, gives us hope when there is none, and gradually as the games tick over spell out where we end up. This is a modern-day internet torture device. Hours of time is spent perched upon the toilet, wondering if the predicted 8.23 for Chase Claypool is worth the risk compared to the forecast 9.04 for Christian Kirk.
Behind this sits a wizard of oz like phantom intelligence, an algorithm that plays dice with our sanity. So, let's put it to work and let the Yahoo Projection Gods forecast out the season from here.
- Note: Data current as of 23/11/2021. Will only take into account lineups set by GMs for future weeks and not any adjustments that may be made in the future.
Quite the interesting outlook this spits out. Django is going on a MONSTER run, winning the last four and clinching a week off. Cainers squeak out a close win in week 13 to jump into the wildcard week, sending Gypos to the Leonard even though they win their final game of the regular season.
We get the Leonard we all want in this future. Plus, we have some serious chaos merchants in the wildcard week.
But how much faith do we put in Yahoo? Let's see what data the SAG LAB+ can do for a couple of different models.
The Leonard according to Current Form
Current form. What is this? Just as players can hit hot streaks, so to can SAG teams. We’ve seen the JAB Heater and the Cainer Coaster this year. In an attempt to approximate this, SAG LAB+ has developed the Four Week Rolling Form (4WRF) metric. This looks at how many standard deviations a team’s score was from their mean, and sums this over the last 4 weeks. Currently the values are:
These values have been then used to adjust each teams mean score to account for trend. For example: Gypsies are -3.14, their adjustment is -31.4% of their standard deviation away from their mean. The result is each team's average score, adjusted for form. Confused? Thats ok. Far too much time has gone into this, so ‘trust the process’. At the end of the day, none of this will be correct any way.
Let’s bring up the Current Form forecast:
My, my, my. Now we have some conversation fuel! Bandits have completely shat the bed and this week's reasonable score ranks as a blip in the system after spending three weeks in double digits. Cainers on the other hand pinch two wins and climb into the top four and Patches’ season is rewarded with a week off.
There are 5,328 other variants and attempts to predict the season results from the SAG LAB+ monkeys, scattered on whiteboards, dot matrix printouts and filling scrap heaps all over the place. What do we know at the end of the day? JABs finish top. Rubbers finish bottom. The rest is still to play for.
Onto the recaps.
Channel4 defeated by Patches
Pats defense set the tone early on this one. Trash points on TNF ended up giving an early 28.85 head start to the Average Joes and even though the WR corps their didn’t fire CMcC was back, Chubb did Chubb, and Herbert did Herbert. Dak was incredibly bad for the News Team and wasn’t even in positive territory at the half. Didn’t help that the Cowboys couldn’t catch COVID on Miami spring break, let alone a pass on the numbers in this one.
Rubbers defeated by Bandits
This result was decided at the opening bell of the Rubber’s Fire Sale Player Bonanza last week. To only end up with one doughnut was an impressive performance. Bandits got the skill player points everywhere, but the QBs were shaky. It wouldn’t have mattered this week, but there would be a few brains trust meetings going on south of Bass Strait to get onto this for the sprint to the playoffs.
Co-Cainers WOLFMANNED by Jabs
Another week, another Wolfman. He is getting quite the work out this year. The Green Bay stack and Ekler going nuts just wasn’t enough though as while the JABs had plenty of great performances (Cam, Mixon and Evens were solid or better), There is really one man that we need to talk about. J Tay BLEW UP in one of those individual games of the season. 200+ yards from scrimmage and 5 TDs, this guy was playing a different game. To pull it out against the highly fancied Bills team was an A+ performance. How's this for his last 8 games points— 17, 30, 27, 15, 18, 34, 18, 50. 13 TDs over this period. Insane.
TacCorp defeated by Gypsies
A big game for the current ladder. Gypos slip out of the current Leonard zone, pushing Taco down to the dreaded 7 spot. It's a relatively comfortable 7 spot these days with Rubbers effectively putting the cue in the rack. The Kansas City Stack had another off-day fantasy wise, despite claiming a win at home against the Cowboys. Elsewhere, this game is littered with disappointment and both GMs will be happy to put this one in the rear view and get on with next week.
Three weeks to go.
3rd and 8,
Commish Clue.
