Upsets, Perfect Records and One Brutal Reality Check

Week 3 Had Absolutely No Interest in Being Predictable

Three weeks in, and the league is already starting to develop a personality.

Darius35 keeps scoring like somebody forgot to tell him fantasy football is supposed to be difficult.

eduardobass7 keeps winning close games.

Bosscoe followed a 171.50-point eruption by falling below 100.

ALL BALLZ ITCH finally broke into the win column by taking down the team that entered the week as one of the league's biggest scorers.

And after starting 2–0, both horns75 and Goonies Never Say Punt finally discovered that the standings do, in fact, contain a loss column.

Week 3 produced 1,187.32 total points, an average of 118.73 per team. That's the lowest-scoring week of the young season so far.

Darius35 apparently viewed that as a personal challenge.

He posted a league-best 147.48, while Goonies Never Say Punt fell all the way to 97.86, the week's lowest score.

We also got one legitimate upset, one mild surprise, a four-point finish, two decisive wins, and enough explosive bench performances to make hindsight look considerably smarter than it actually was.

In other words:

Fantasy football remains undefeated.

Week 3 Awards

---

Biggest Upset

ALL BALLZ ITCH WAS GIVEN A 40% CHANCE. APPARENTLY THAT WAS 40% TOO MUCH DOUBT.

ALL BALLZ ITCH 136.32, Bosscoe 98.94

Bosscoe entered Week 3 in a very different place than he left Week 2.

One week earlier, he had dropped 171.50 points and climbed to the top of the league in total scoring.

This week, he was projected for 128.29 against ALL BALLZ ITCH's 120.57, carrying a 60.0% win probability.

Reasonable.

Then the games started.

And ALL BALLZ ITCH apparently took the projection personally.

Brock Bowers scored 27.60.

Garrett Wilson scored 26.70.

Ja'Marr Chase scored 24.80.

Kenneth Walker scored 21.30.

Even kicker Will Reichard dropped 16.00.

Those five spots alone produced 116.40 points.

Bosscoe's entire lineup scored 98.94.

That's the matchup.

That's the paragraph.

We could probably stop there.

But it gets worse for Bosscoe.

This wasn't even an owner-decision disaster.

His submitted lineup was exactly the projection-optimal lineup before the games. There were 0.00 projected points being carelessly left behind.

The lineup made sense.

The players simply didn't cooperate.

Amon-Ra St. Brown scored 11.90 against a 19.47 projection.

Jalen Coker scored 3.80 against 14.54 expected points.

The New England defense was projected for 6.54 and finished at minus-1.00.

Even Lamar Jackson came in below expectation.

Bosscoe missed his team projection by 29.35 points, the largest shortfall in the league.

Meanwhile, ALL BALLZ ITCH beat expectation by 15.75.

That's a swing of more than 45 points relative to the forecast.

And the winning lineup wasn't even perfect.

Jaylen Warren scored 20.60 on the bench, and DK Metcalf had 12.10. A perfect hindsight lineup pushes ALL BALLZ ITCH all the way to 161.52.

So yes, Bosscoe could have squeezed another 8.94 points out of his own roster after the fact.

It still wouldn't have mattered.

This wasn't a coaching failure.

This was a 40% underdog walking into the week, getting huge performances everywhere, and turning a projected loss into a 37.38-point beatdown.

ALL BALLZ ITCH is finally on the board at 1–2.

Bosscoe falls to 1–2.

And after scoring 171.50 one week and 98.94 the next, Bosscoe has officially completed the full fantasy-football emotional experience.

---

Another Favorite Goes Down

GOONIES STARTED 2–0. THEN THEIR BENCH OUTSCORED HALF THE STARTING LINEUP FOR FUN.

Estevan059 109.44, Goonies Never Say Punt 97.86

This wasn't a massive upset.

Goonies entered as only a modest favorite: 53.5% win probability, with a projected 119.13–116.46 edge.

Basically, the numbers were saying:

"We think Goonies wins, but don't put the house on it."

Good advice.

Estevan059 got 25.14 from Bo Nix, 21.90 from Derrick Henry, and 20.20 from CeeDee Lamb. Nobody completely detonated, but those three gave Estevan enough stability to survive several quieter performances elsewhere.

The final total was only 109.44, seven points below projection.

Fortunately for Estevan059, Goonies had a much rougher afternoon.

Justin Jefferson scored 5.20.

Dalton Kincaid scored 3.80.

The Seattle defense scored 2.00.

Rhamondre Stevenson had 6.80.

That lineup finished 21.27 points below its projected total and produced the lowest team score of Week 3.

And then we looked at the bench.

Oh boy.

Michael Wilson: 25.90.

Tyler Shough: 24.80.

Matthew Golden: 21.00.

Aaron Jones: 14.20.

Denver defense: 14.00.

If we build the perfect lineup after knowing all the answers, Goonies jumps from 97.86 to 151.30.

That's a 53.44-point hindsight swing.

It looks catastrophic.

It also isn't remotely that simple.

Before kickoff, the lineup Goonies submitted was only 2.27 projected points below the best possible projected combination.

Read that again.

The postgame numbers scream, "HOW DID YOU LEAVE ALL THOSE POINTS ON THE BENCH?"

The information available before the games mostly shrugged and said, "This lineup is fine."

That's exactly the difference between a bad decision and a bad result.

Michael Wilson was projected for 12.05.

Matthew Golden was projected for 11.18.

Denver's defense was projected below Seattle.

Tyler Shough was projected behind Josh Allen.

Then football happened and all of those assumptions caught fire at once.

Estevan059 improves to 2–1.

Goonies falls to 2–1.

And somewhere, five players on the Goonies bench are probably asking whether they can start their own fantasy team next week.

---

Nail Biter

KRISTIANP1 FINALLY GOT A BREAK FROM THE BLOWOUTS. UNFORTUNATELY, HE STILL DIDN'T GET A WIN.

eduardobass7 122.70, kristianP1 118.66

The forecast liked eduardobass7 here.

Projected score: 121.70–113.18.

Win probability: 61.0% to 39.0%.

The final score?

122.70–118.66.

For once, the projection didn't get embarrassed.

eduardobass7 finished almost exactly where expected, beating his projection by one point.

kristianP1 outperformed his number by 5.48 and made the favorite sweat.

A lot.

The difference was only 4.04 points.

eduardobass7 once again rode Jaxon Smith-Njigba, who followed last week's 42.50 explosion with another monster 35.36.

Tyler Warren added 14.70.

Houston's defense scored 13.00.

Parker Washington contributed 13.00.

It wasn't the same offensive fireworks show as Week 2, but it was enough.

kristianP1 answered with a balanced lineup of his own.

Davante Adams scored 20.70.

Jared Goff had 19.36.

Javonte Williams scored 18.30.

George Pickens added 15.20.

Chuba Hubbard had 15.00.

Five solid performances.

Still four points short.

Naturally, there was one player on the bench ready to make the loss hurt a little more.

Wan'Dale Robinson scored 18.70.

Bucky Irving scored 7.80 in the starting lineup.

Swap those two after the fact and kristian wins.

Easy, right?

Not exactly.

Before kickoff, Bucky was projected for 14.77.

Wan'Dale was projected for 8.83.

Starting Bucky wasn't just defensible.

It was the clear projection-based choice.

This is becoming kristianP1's season.

Last week, a pile of bench points made the lineup look terrible after the games even though most of the decisions were reasonable beforehand.

This week, one hindsight swap would have produced the first win, but the weaker-performing starter was projected almost six points higher entering the matchup.

You can do the right thing and still get the wrong result.

Fantasy football has built an entire industry around that sentence.

eduardobass7 moves to 3–0.

kristianP1 falls to 0–3.

But that 0–3 deserves a little context.

The last two losses have come by 3.38 and 4.04 points.

That's 7.42 combined points separating kristian from a very different-looking record over the past two weeks.

Unfortunately, Sleeper does not award victories for emotional damage.

---

Statement Win

DARIUS35 WAS A 70% FAVORITE. THEN HE SCORED 147.48 JUST TO MAKE SURE WE UNDERSTOOD WHY.

Darius35 147.48, horns75 117.00

Sometimes the favorite wins.

Sometimes the favorite wins and spends three hours explaining why the probability should have been higher.

Darius35 entered Week 3 with a 70.0% win probability, easily the strongest favorite on the slate.

Projected score:

Darius35 122.23.

horns75 106.30.

Actual score:

147.48–117.00.

Darius didn't just meet expectations.

He cleared them by 25.25 points, the largest positive projection swing of the week.

And the reason was not subtle.

Jahmyr Gibbs scored 41.40.

Brock Purdy scored 31.28.

Christian Watson scored 22.60.

That's 95.28 points from three players.

Good luck.

horns75 actually had a respectable week.

He beat his own projection by 10.70 points.

George Kittle exploded for 26.20 after being moved into the starting lineup.

Matthew Stafford delivered 22.90.

Jacksonville's defense scored 17.00.

Deebo Samuel added 15.40.

In plenty of weeks, 117 keeps you competitive.

This was not one of those weeks.

There is one decision horns75 can reasonably stare at.

Drake London scored 28.40 on the bench.

And unlike several of the hindsight traps we've already discussed, London was also projected well beforehand at 13.47. Xavier Worthy, who started and scored 4.00, carried a 10.63 projection.

So yes, there were legitimate points available there.

Put London's 28.40 in for Worthy's 4.00 and horns75 reaches 141.40.

That's a massive improvement.

And still a loss.

Darius35 scored 147.48.

Even the clearest lineup mistake on the other side doesn't erase what happened.

Darius is now 3–0, has scored a league-leading 431.66 points, and is averaging an absurd 143.89 points per week.

Three games is still a small sample.

But if you're looking for the early team everybody else should be watching, the scoreboard is making a fairly loud argument.

---

First Win, Finally

KING IROK SAID ENOUGH OF THIS 0–2 NONSENSE AND HANDLED BUSINESS

King iRok 134.08, Prestige WorldWide 104.84

King iRok entered as a modest 54.9% favorite.

Nothing dramatic.

No massive projected edge.

Just a slight expectation that he should win.

He did considerably more than that.

Bijan Robinson scored 35.30.

Tee Higgins scored 21.00.

Jordan Love delivered 19.48.

Minnesota's defense put up 17.00.

Trey McBride added 16.50.

And suddenly the team that entered Week 3 at 0–2 had 134.08 points and a very comfortable first victory.

The best part for King iRok is that one lineup change worked beautifully.

Drake Maye was in the earlier lineup and ultimately scored 5.76.

Jordan Love got the start and produced 19.48.

That's a 13.72-point gain.

Not every move hit.

Luther Burden scored 19.80 on the bench, while Courtland Sutton finished with 7.60 after moving into the lineup.

But the quarterback decision more than softened that blow, and the final margin was 29.24 points anyway.

Prestige WorldWide has a much uglier box score to look at.

Christian McCaffrey scored 21.60.

Chris Olave had 19.70.

Kicker Harrison Mevis somehow chipped in 16.00.

After that, the lineup thinned out quickly.

TreVeyon Henderson scored 3.90.

Malik Nabers had 7.60.

Saquon Barkley had 9.00.

Sam LaPorta scored 7.40.

And then came the bench.

Joe Burrow: 22.58.

Jordan Addison: 20.00.

Terry McLaurin: 19.70.

A perfect hindsight lineup gets Prestige WorldWide all the way to 143.38.

That makes the final lineup look brutal.

But once again, context matters.

Between the earlier forecast and kickoff, Prestige WorldWide moved from Joe Burrow to Bryce Young and from Terry McLaurin to TreVeyon Henderson.

Combined, those two changes actually increased the projected lineup total by 0.93 points.

After the games?

They cost 23.74 actual points.

That's not brilliant coaching magically becoming stupid overnight.

That's the exact same decisions looking completely different once we know the answers.

King iRok gets the win he needed and moves to 1–2.

Prestige WorldWide falls to 0–3.

And if the schedule itself could receive hate mail, Prestige WorldWide would probably have already sent several pages.

More on that in a moment.

---

The Bench Report: Hindsight Had Another Fantastic Week

The loudest Week 3 bench performances:

• Drake London — 28.40 on horns75's bench
• Michael Wilson — 25.90 on Goonies Never Say Punt's bench
• Tyler Shough — 24.80 on Goonies Never Say Punt's bench
• Joe Burrow — 22.58 on Prestige WorldWide's bench
• Matthew Golden — 21.00 on Goonies Never Say Punt's bench
• Jaylen Warren — 20.60 on ALL BALLZ ITCH's bench
• Jordan Addison — 20.00 on Prestige WorldWide's bench
• Luther Burden — 19.80 on King iRok's bench
• Terry McLaurin — 19.70 on Prestige WorldWide's bench
• Wan'Dale Robinson — 18.70 on kristianP1's bench

That list looks ugly.

It also proves why context matters.

Some misses were visible beforehand.

Drake London over Xavier Worthy is the cleanest example. London carried the better projection and then demolished Worthy's actual score.

Others were basically impossible to criticize before the games.

Wan'Dale Robinson was projected well below Bucky Irving.

Goonies' submitted lineup was only 2.27 projected points away from optimal despite leaving more than 50 hindsight points unused.

Prestige's late pair of lineup changes actually improved the expected total slightly before costing almost 24 actual points.

Postgame optimization is easy.

You already know who scored.

Real lineup decisions are made while everybody is still guessing.

That's what makes this stupid hobby so much fun.

---

Three Weeks In: Two Perfect Teams, Two Winless Teams and One Very Misleading Record

Darius35 Has Put the League on Notice

Three games.

Three wins.

431.66 points.

Nobody has scored more.

Darius is averaging 143.89 per week, and the Week 3 lineup wasn't even hindsight-optimal. There were another 26 points available after the fact.

Nobody needs to start handing out trophies in September.

But 3–0 plus the league's highest scoring total is about as clean an early résumé as you can build.

eduardobass7 Is Quietly Right There

The other 3–0 team isn't winning the same way.

eduardobass7 has scored 393.24, nearly 40 fewer than Darius, but has now survived multiple competitive matchups and keeps finding enough.

Week 2 was the big upset.

Week 3 was the four-point escape.

Different path.

Same 3–0 record.

Bosscoe Might Be the Most Misleading 1–2 Team in the League

Bosscoe is 1–2.

Bosscoe has scored 397.70 points.

That's the second-highest total in the entire league.

More than undefeated eduardobass7.

More than Goonies.

More than everybody except Darius.

He also owns a +26.54 point differential.

This is what happens when a 171.50-point week and a 98.94-point week live on the same résumé.

Bosscoe isn't playing like a typical 1–2 team.

He's playing like a roller coaster with a fantasy roster attached.

horns75 Is 2–1 With a Negative Point Differential

horns75 has scored 336.18 and allowed 337.76.

That's a minus-1.58 differential.

Record?

2–1.

No complaints will be accepted from anyone who believes style points matter.

kristianP1 Is 0–3, but the Last Two Have Been Brutal

The record is ugly.

The recent margins are not.

Week 2: lost by 3.38.

Week 3: lost by 4.04.

That's 7.42 total points across two losses.

A couple of catches, one defensive touchdown, one random kicker swing, and this record could look very different.

It doesn't.

But it could.

I'm sure that's extremely comforting.

Prestige WorldWide's Schedule Needs to Calm Down

Prestige has allowed 458.44 points in three weeks.

That's an average opponent score of 152.81.

One hundred fifty-two point eight one.

Every week.

For three weeks.

Prestige has made lineup decisions that deserve scrutiny, sure.

But when your opponents are averaging a score that would have been the second-highest total in Week 3, perhaps the universe could contribute a little help too.

---

Final Word

Week 3 gave us a little bit of everything.

ALL BALLZ ITCH finally got its first win by taking a 40% chance and turning it into a 37-point statement.

Darius35 stayed perfect and somehow made a 70% win probability look conservative.

eduardobass7 stayed perfect the stressful way.

King iRok finally broke through.

Goonies Never Say Punt discovered that a loaded bench is only useful if you somehow know the future.

Bosscoe discovered that scoring 171 one week does not roll over like unused cellphone data.

kristianP1 discovered, for the second week in a row, that "reasonable decisions" and "victory" are not the same thing.

And Prestige WorldWide discovered that opposing teams remain extremely rude.

After three weeks, only Darius35 and eduardobass7 remain undefeated.

Only kristianP1 and Prestige WorldWide remain winless.

Everybody else is already packed into the messy middle where one big week can completely change the conversation.

Which is exactly where we want this league.

Three weeks down.

A lot of season left.

And if Week 3 taught us anything, it's that the projections are useful, the decisions matter, and none of us actually know what's about to happen.

Bring on Week 4.