The Week the Spreadsheet Didn't Show Up

# The Week the Spreadsheet Didn't Show Up *By Raffaele Cuccinotta* ## Here's the situation Every week I set out to find a pro football decision where the coach and the computer disagreed, lay out both sides, and ask whether the human or the model had the better case. This week, I went looking for the computer. I couldn't find it. ## What happened Sunday, Nov. 8, 2026, gave us two coaching decisions that drew heavy fire. In Indianapolis, Shane Steichen threw on third-and-8 with a minute-forty-five left instead of running to bleed clock. The pass was dropped, the Colts kicked a field goal, and Dallas drove for a winning touchdown. In Washington, Dan Quinn managed the Commanders' final drive against the Rams so clumsily that Pro Football Network called him "clueless" and fans demanded he be fired. Both decisions were covered as coaching failures. Both were framed, implicitly or explicitly, as moments where instinct overruled good sense. And in both cases — and in every other game from that Sunday — I could not locate a single published analytics model that had recommended any specific play call. ## What the model actually said Nothing. Or at least, nothing public. Pre-game models like numberFire, Squawka's Signal, and Bandwagon's ratings forecasted winners and spreads, not play-calling. Post-game win-probability charts from nflfastR (via Luck Adjusted) tracked the Colts-Cowboys game swing-by-swing, but the third-and-8 incompletion didn't even register among the biggest shifts. The field goal that followed it moved the needle nine percent. The incompletion itself was invisible to the model's ledger. For Quinn's final drive — two plays from the Rams' 31, eight seconds, never a throw to the end zone — there was no model verdict at all. The coverage was pure opinion. "Baffling." "Clueless." "Worst coach in the league." But no spreadsheet was cited. No bot was quoted. The NFL 4th Down Bot, which normally chirps in on these moments, stayed quiet. ## So what were we arguing about? Steichen, for his part, said he wanted the first down so he could finish the game with the ball. That is a coherent strategy. It failed because a receiver dropped the pass, not because the call was unsound. Quinn said he wanted a shot against a possible max blitz with a check-down option. The execution was poor. The design was defensible. In neither case did a coach "ignore analytics." The analytics, as far as I can tell, had not spoken. The debate was a shadow play — journalists and fans arguing about a controversy that existed only in the telling. ## The pattern This is not a one-week anomaly. I have now searched five major sports — NFL, NBA, MLB, NHL, soccer — for cases where a coach overrode a documented analytics recommendation and was demonstrably correct. I have found zero. What I am finding instead is a growing body of decisions that get covered as "analytics versus instinct" when no model recommendation is actually in the room. The Cowboys' two-point attempt in October had a 0.8 percent win-probability gap between kick and go — a toss-up the model called a lean, not a command. The Eagles' fourth-down field goal I audited three weeks ago sat inside a confidence interval wide enough to include the coach's choice. And now, two more "controversies" with no model behind them at all. ## What this means for the beat I started this column believing the honest answer was to give both sides a fair hearing. I am starting to worry that fairness requires me to say when one side hasn't shown up. If the spreadsheet is silent and we are still writing about coaches defying it, we are not covering a debate. We are manufacturing one. The model says one thing, the coach does another, so let's look at why. But when the model says nothing, and we pretend it spoke — that is worth auditing too. ## Sources - IndyStar: "Why Colts coach Shane Steichen threw the ball on third down with 1:45 to play" (Nov. 8, 2026) - IndyStar: "Why the Colts' letting the Cowboys score at the end of the game was the right decision" (Nov. 8, 2026) - Luck Adjusted: Cowboys at Colts, Week 9 2026 win-probability chart (nflfastR) - NFL.com: "Commanders' mystifying final drive symbolic of an offense that keeps misfiring" (Nov. 8, 2026) - SI.com: Commanders-Rams postgame transcript, Dan Quinn (Nov. 8, 2026) - Pro Football Network: "Dan Quinn Is Clueless" (Nov. 8, 2026) --- *Raffaele Cuccinotta covers the intersection of data and decision-making in professional sports. He is still trying to be fair to both sides.*