The Organizer's Handbook.

What the create form asks

On the Pick'em page: a group name, your name as it should appear on the roster, the sport, and the term (Step 2). Then three switches that shape the season:

• Scoring — straight-up, against the spread, or confidence points (Step 3). Chosen at creation; it doesn't change mid-season.
• "We play for a pot" — turns on the pot note field and the paid list. The note is what shows on your group page ("what we play for — the organizer holds it").
• Add Luka as a house player — on by default; he plays under his printed rules (Step 8) and can never win a pot.
• Allow Luka Autopilot — on by default; turn it off here, or later in Commissioner tools, if your room wants every sheet hand-made.

Who can start what

Anyone with an account can organize one standard group, free, forever — straight-up scoring on weekly or daily slates — and anyone can join as many groups as they're invited to. Spread and confidence scoring are House scoring: they ride on the organizer's membership, and when the organizer is a House member the whole group plays with them, free players included. That's the Jesse rule: group features follow the organizer; personal perks follow the person. While membership is in its preview posture those scoring modes stay selectable, labeled as the preview they are; once membership opens, the gate follows the organizer's membership exactly.

One honest note about saving: signed in, your groups save to your account and follow you across devices. Signed out, a group lives on that device as a preview — the page labels it, and invite links resolve on that device — until you sign in.

3 days

For pro basketball, baseball, hockey, and soccer, the three-day window is the sweet spot for most crews: enough games to separate the field, few enough picks that a busy member survives it. Tonight-only is the original daily pool — in by dinner, settled by midnight. The 2-day window is a short race for groups that play in bursts.

Full week & one team's week

A full week of a daily sport is a standings race, not a sprint — reserve it for crews who genuinely want to pick every night, because that's what it asks. One team's week is the low-effort on-ramp: every game of a single team across its week — a home series, a road trip — a handful of games and one storyline.

Weekly, ride, or up front

Football's natural unit is the weekly slate — every game that week, the classic shape. Ride one team is the zero-maintenance format: each member picks a team before the season and its results are their score all year (members set their ride on the group page). Pick the whole season up front is one big preseason argument, then a year of watching it age. College football mirrors the weekly slate; the bowls are their own slate.

Straight-up winners

Pick who wins; a correct pick is worth 1 point. The final game's total-score tiebreaker separates equal sheets: closest prediction of the two teams' combined score wins the tiebreak, and it isn't part of the points. A tied game has no winner — nobody scores it, for any format.

Against the spread

Your pick has to beat the number, not just win: the picked side's score is measured against the game's spread, and an exact landing on the number is a push — no point for anyone. One honest line about today's slates: the clubhouse's odds feed posts moneylines, not spreads, so a spread group currently scores against a level (zero) line attached to each game — picking the winner, with tied games scoring for nobody — until sourced spread lines join the slate. The page never invents a spread.

Confidence points

Every game you pick also gets a confidence rank: on an N-game slate you hand out the numbers 1 through N, each exactly once, biggest number on your surest pick. A correct pick earns its rank; a wrong pick earns 0 no matter how big its rank was. The sheet enforces the system for you — choosing a rank another game already holds swaps the two — and it won't save while a picked game sits unranked. Standings order: points, then correct picks, then the tiebreaker.

The link and the code

The moment your group exists, its page carries an invite link with a one-tap copy button and a six-character join code. Text or email the link; anyone with it lands on a join screen, adds their roster name, and they're in. A player who lost the link can type the code into the "Have an invite code?" box on the Pick'em page instead. There's no approval queue and no roster cap in the software — the gate is simply who you send the link to.

Names and handles

Each player picks under the roster name they joined with, and members keep a contact handle on their profile so the group can actually reach each other — remind your crew to fill theirs in during week one, before the first "did anyone hear from—" moment. In a pot group, a member's payment handle shows on the roster next to their name, which is how friends match a payment to a player without the site ever touching the money.

Locks and daylight

Every pick locks at its game's start — not at a slate-wide deadline. Before a game locks, picks are private; the moment it locks, the group page shows everyone's pick for that game in one table, confidence ranks included in confidence groups. A member who never picks a game simply scores nothing for it; there are no auto-picks and no late edits, for anyone, including you.

Members can change any pick — or their tiebreaker number — right up to that game's start, and an Autopilot sheet can be overridden the same way. The sheet that locks is the sheet that counts.

Your voice in the room

The group page is part standings table, part room: members post in group talk, and as organizer you can pin any post as the announcement at the top. That pinned slot is your reminder tool — the deadline post, the "picks lock at the early game tonight" nudge, the ruling everyone needs to see. Use it instead of chasing people one by one; the software doesn't text anyone on your behalf, and this handbook won't pretend it does.

Entering results settles everything

As organizer, your group page has an Enter results panel: the two final scores for each game on the slate. Saving them settles the slate on the spot — winners, points, the tiebreaker's distance from the real total, and the standings all recompute from those scores. Entered a score wrong? Re-enter it and save again; the standings follow the corrected record. Your entered results are the group's official record — that's what makes the next part work.

Ties resolve in the open: equal points break on correct picks (confidence groups) and then on tiebreaker distance — the standings table shows the Δ next to every player once results exist, so anyone can check the math.

Settling disputes

Pools argue about three things: a score, a lock, and a house rule. The handbook answers: the score of record is the one in your results panel — if the room shows it was entered wrong, correct it and let the standings re-settle; a lock is the game's posted start, applied the same way to every sheet including yours; and a house rule your group wrote down (in its pot note or a pinned post) governs your group's game where it differs from a default — say which one you applied.

Make the ruling once, in group talk, and pin it. A pinned ruling ends most arguments because it's findable; an unpinned one restarts every Tuesday. When the argument is about the rules themselves, ask Luka — he settles rule questions from the house rules and will quote the one that applies.

The house player

Luka sits in your standings as a labeled house player. His picks run on one transparent knowledge rule — the listed favorite, else the home team — because the Lab's first model prices in-game states, not pregame picks, and he never borrows its percentage to dress up a pregame sheet. Toggle him any time in Commissioner tools on your group page. And the hard rule, enforced in the standings code itself: Luka can never win a pot. If he tops a pot group, the winner shown — and the person your group pays — is the highest human finisher.

Members' sheets, Luka's hand

A member can let Luka fill the games they haven't picked — same knowledge rule, same information a Take gives anyone. It's off unless a member turns it on, it's disclosed with a marker in the standings, the member can override any pick or rank before it locks, and in a confidence group Autopilot also sets the ranks by his printed rule. If your room bans it, flip "Allow Autopilot" off in Commissioner tools and no sheet in the group can use it — your room, your call.

On every game

Every game on the pick sheet carries an Ask Luka link that takes him the matchup and asks for his take: his pick, the sourced facts it rests on, and where he differs from the market number — in honest tiers (a lean, a solid read, a coin flip), never a "lock," never an invented percentage. That's a tool for your players, not an edge for anyone: he runs on the same information any member can ask him for.

Survivor

Each member picks one team per week. Win and you advance; a loss or a tie is a strike. At creation the organizer sets the group as 1-strike (one and done) or 2-strike. A member can never pick the same team twice in a season, and no pick by lock is a strike. Last member standing wins. Settles on: real game results.

Luka plays by the same rules and can never win a pot: if Luka outlasts the room in a pot group, the pot belongs to the last surviving human.

Pick-a-Few

The organizer sets how many games — Pick 5, say. Each member builds their own card of that many games from the week's full slate and picks the winners; the card is scored by wins, and most wins takes the week. When the group uses confidence scoring, confidence ranks can ride on the card. Settles on: real game results.

Weekly Honors

The group picks players for the week's real honors races — AFC Offensive Player of the Week, say, or a conference's Player of the Week in college. The pick list is real players from house data, and a correct pick earns the points the organizer set.

No source, no settle. A pick only settles when the league or conference's announced winner is entered by the organizer with the announcement source — name plus link — and the result prints its source for the whole group to check.

House Questions

A weekly card of yes/no questions about the group's own slate, generated from the real games: "Does the home team win the Monday night game?" "Will at least one underdog win this week?" "Does AP No. 1 win this week?" The organizer chooses which questions run. Answers lock at the first relevant kickoff and settle mechanically from the finals, with the resolving fact printed on each settled question — "Final: Bills 24, Chiefs 27 — the home team lost."

It is a pool among friends in the group — never a market — and it runs on the same pot law as everything else on this page.

Squares

The classic 10×10 grid for one big game — built for the Super Bowl, ready for any single game on a slate. Create the pool, choose the game, and set how many squares each member may claim. Members tap squares to claim them, or take random ones; the organizer can top a member up the same way. The two digit axes are drawn at random only when the organizer draws them — until the draw, the axes show "?" and nobody can cherry-pick a row or a column. Squares still unclaimed at the draw belong to the house, marked House on the grid — House squares win nothing; the organizer's pot note says what the room agreed happens then.

Winners settle mechanically at the end of Q1, halftime, the end of Q3, and the final (overtime rolls into the final): the organizer posts the game's score by quarter from the box score, and the pool prints each winning square with the score that paid it. Settles on: the game's quarter scores, posted on the pool page.

Brackets

Members post their full bracket — or their round-by-round re-picks, your call as organizer — in the group room before the first game tips; the organizer scores them against the bracket as rounds land and posts the table. Settles on: tournament results.

Do the homework before you pick

The clubhouse is built so the homework takes ten minutes: the Rail carries the lines and the movement for the games on your slate, and the Closing Lines archive shows where numbers finished in past slates — the single best teacher of what a line means by the time a game starts. For pro football, the Lab explains how its model reads a game state; its in-game number is a different question from a pregame pick, and Luka will say so if you try to borrow it. Then tap Ask Luka on the games you're stuck on — his take shows its work, so you can agree with the reasoning or fade it on purpose.

How to hand out the big numbers

Rank math first, because it decides everything: on a 14-game slate the ranks total 105 points, and the top three ranks alone (14 + 13 + 12) are worth 39 — more than the bottom seven ranks combined (28). So confidence is really a question about your three surest reads. Put the biggest numbers on listed favorites you actually agree with — games where the market's read and yours point the same way — and let the true coin flips carry the small numbers; a coin flip with a 3 on it costs you almost nothing when it lands wrong, and a favorite carrying a 4 is where slates go to die.

Late-slate strategy: ahead vs behind

Once everyone's picks turn visible at lock, the pool becomes arithmetic. If you're ahead, the room's obvious big-rank games are your friends: matching the field's heavy favorites protects your lead, because the field can't gain on games where you all score together. If you're behind, matching the field is how you stay behind — you gain only where your sheet differs and is right, so shift your bigger ranks toward the games where the room's ranks look thin, and accept the variance: a chased slate is won by two or three differentiated ranks landing, not by ten safe ones. Neither move is a gamble on a game; both are just reading the points table you can already see.

The tiebreaker is a totals guess, not a winner pick

The tiebreaker asks for the final game's combined score, and it's scored on distance — how far your number lands from the actual total. Direction doesn't matter: two over is exactly as good as two under. So don't anchor on your pick in that game; think about the two offenses and the pace they play at, then guess the total as honestly as you can. It's the only number on the sheet where being a little wrong in either direction costs the same — most players waste it by predicting the score of the game they want.

Home, rest, and the listed favorite

Home teams win more than road teams in every sport on the slate — that's real, and it's also already inside the number: the listed favorite is usually the home team because of it. The homework edge is the second layer: in the daily sports, a team on the second night of back-to-back games, or at the end of a long road trip, is not the same team the season record describes. The slate shows you every start time and matchup — check who's rested before you spend a big rank on a tired favorite, and don't pay full price for a home edge the line already charged you for.

Path planning beats this week's lock

Survivor's real game is the calendar: you can spend each team once, so burning the league's best team in week one buys one win at the price of your best future asset. Plan two or three rounds ahead — find the weeks where your remaining teams have their softest matchups, work backward, and spend this week's pick on the weakest team that still projects as a winner. The players who last aren't the ones who picked the most winners; they're the ones who never had to pick a bad team in a bad week.