Pennant ChaseActivity 03FootballGrades 8–12

What a Pick Costs

Twenty-two starting jobs and one football. Some of those jobs touch the ball on every snap and some touch it four times a game. Your draft is a bet on which ones matter.

NameTeamLeague

Part One · Before the draft

Take a position

Football forces the trade-off into the open. You have to fill every position whether you want to or not, so every pick spent on a great player at one spot is a pick not spent somewhere else you're still required to fill.

Commit to a strategy in writing before you see the board. Circle one and defend it:

A. Quarterback first. He touches the ball on every offensive snap. Nothing else comes close to that leverage.

B. Trenches first. Linemen decide whether anyone else gets to do their job at all.

C. Spread the picks. No weak spots beats one great spot.

D. Something else — describe it.

Your strategy, and why you believe it will work

The leverage estimate

Before drafting, estimate how many times per game each of these touches the ball. Guess honestly — you'll check it later.

Predicted touches per game
Position	Your estimate	Actual, from your season	Difference
Quarterback (pass attempts + carries)
Lead running back (carries)
Top receiver (targets)
Third receiver (targets)
Kicker (attempts)

Part Two · The draft

Build the roster

The last two columns are the ones that make this activity work. Fill them in during the draft, not after.

Draft log
Pick	Player	Pos	Why you took him here	Best player you passed on
1
2
3
4
5
6
7
8

The idea to hold onto: replacement level

The right question at a pick is never "how good is this player?" It's "how much better is he than the player I could get at this position with a much later pick?"

If the best quarterback available is far better than the twelfth-best, and the best running back is barely better than the twelfth-best, the quarterback is the more valuable pick even if both are excellent players. Watch for that gap as the board thins out — it's the whole argument about positional value in one sentence.

Part Three · During the season

Keep the log

Five numbers a week. Yards per attempt and turnover margin are the two that will matter most at the end, so don't let those slide.

Team tracking log
Week	W	L	Pts for	Pts against	Yds / pass att	Yds / carry	Turnover margin

Part Four · After the season

Make the argument

A. Total versus rate

Build a league-wide table: every team's wins, total yards, yards per pass attempt, yards per carry, and turnover margin. Then make two scatterplots — total yards against wins, and yards per pass attempt against wins.

The two plots don't look the same. Why not? What is a rate stat measuring that a total isn't?

B. The questions

How strong is the relationship between turnover margin and wins in your league? Now the harder question: is turnover margin something a general manager can draft for, or is a lot of it luck? What evidence from your league would settle it?
Go back to your Part One touch estimates and fill in the actual column. Where were you most wrong, and how does that change what a first-round pick should be spent on?
Sort the league's eight first-round picks by how much they actually helped. How many of those picks would you make again today?
Find the teams that drafted a quarterback first and the teams that didn't. Compare their finishes. Then explain why this comparison is weaker evidence than it looks.
Take one specific pick from your draft log and calculate its cost — the player you actually took versus the best player you passed on. Use per-attempt numbers, not totals.
Running backs go cheap in most drafts. Was that correct in your league? Describe the league setting that would make it wrong.

C. The debate

Split into two sides and argue the resolution below. Every claim has to cite a number from your own league — not from anything you've read or watched.

Resolved: in a league like ours, the first pick should always be a quarterback.

Your three strongest pieces of evidence — and the one that most undercuts your side

Extension

What is the engine doing?

Football is the sport where it's most obvious that one play is a chain of decisions rather than a single roll. Write the chain out before you read the hints.

To simulate one play, the program probably has to…

Four links to think about

Run or pass? Statistics alone can't decide this. Down, distance, score, and time remaining all have to be in there. Find a fourth-and-1 in a box score and then a second-and-12 — does the play call look like it came from the same rule?
Who gets the ball? Another weighted draw. A back who took 60% of his old team's carries probably gets picked around 60% of the time. Check your own box scores: does the split among your backs look like their historical shares, rescaled?
What happened? Complete, incomplete, sack, interception — a set of probabilities that has to blend the quarterback's numbers with the defense's. If a great passer meets a great pass defense, where does the result land?
How many yards? Not a single number but a draw from a distribution. Most carries gain a little; a few gain a lot. What shape would that distribution have, and how would a player's yards-per-carry average set it?

The sim engine primer covers the math for blending an offensive rate with a defensive one.

Teacher guide

Running What a Pick Costs

Timing

One period to launch, 10 minutes a week for tracking, two periods at the end — one for the data work, one for the debate. Football seasons are short in games, so the calendar is easy; you set the number of games when the league is created.

The point of the unit

Two targets. The first is that rates beat totals: total yardage is a famously poor predictor of winning compared with yards per attempt and turnover margin, and students find that out from a table they built themselves. The second is replacement level — the idea that a player's value depends on what you'd have without him, not on how good he is in isolation.

The Part Four debate is where the assessment lives. The rule that every claim must cite the class's own league data does most of the teaching, because it blocks students from arguing with things they've heard on television.

Where classes get stuck

Correlation read as cause. Turnover margin correlates strongly with winning. Students will conclude that turnovers cause wins. Worth pushing on: teams that are ahead run more and throw less, which changes turnover chances. The arrow may point both ways.
Eight teams is not a data set. The QB-first comparison in question 4 will produce a confident answer from a sample of four versus four. Name that out loud; it's the most valuable minute of the unit.
Sports knowledge dominating. Students who watch a lot of football will argue from memory. The cite-your-league rule is the fix, and it also levels the field for students who've never watched a game.

Differentiation

Lighter: Drop the scatterplots and just rank teams by each stat, comparing each ranking to the standings. Same conclusion, no graphing.
Heavier: Compute correlation coefficients for four or five stats against wins and rank them by predictive strength. Then ask which of those a general manager can actually control at the draft — a genuinely open question.
Cross-curricular: The debate maps cleanly onto an argumentative writing or speech-and-debate unit, with the useful constraint that evidence has to be quantitative and self-generated.

Setting up the league

Email guy@pennantchase.com for a free class league — sport, class size, weeks available.
Send a spreadsheet of logins to have accounts created in bulk rather than having students sign up. The site blocks repeated sign-ups from a single IP address, and a class sharing a school network trips that protection.
Some school district mail servers block mail from pennantchase.com. If a reply never arrives, ask IT to allow the domain or write from a personal address.
Number of teams, eligible historical players, and games per season are all yours to set. For the debate to work you want at least six teams so the comparisons aren't trivially small.