Polling

Polls are paid actions that simulate what would happen if the election resolved right now. They're the single most useful intelligence tool in the game, since they run the same vote math the turn processor uses, so what they predict is literally what you'd get. Two tiers: Quick and Full.

Quick poll

• Cost: 2 actions + ₳25,000
• Scope: Your current or about-to-declare race
• Returns: Topline vote share + your 5 strongest demographic groups + your 5 weakest

Quick Poll is the "should I run?" tool. Before declaring any candidacy, spend 2 actions and ₳25k to see whether you've got a realistic path. The topline is presented as a percentage-of-votes projection against the current field.

Full demographic poll

• Cost: 6 actions + ₳75,000
• Scope: Same race as Quick, deeper output
• Returns: Complete per-group breakdown across every demographic category

Full Poll is the "what do I need to fix?" tool. It tells you:

• Your vote share with every one of the 12 voter archetypes.
• Your Favorability per group.
• Turnout modifiers on each group (who's been canvassed / GOTV'd in this state).
• Where your appeal times reach times approval math is lowest: the gaps.

Use it mid-general to plan the closing sprint: which demographics to ad-target, which to canvass, which to accept as lost and move on from.

When to use each

What polls actually compute

For state races (House, Senate, Gov, State Senate, Commons, Bundestag constituency, Sangiin, Shūgiin):

Polls run the exact same group-level competitive allocation the turn processor uses, with the same FPTP spoiler rule for FPTP states. Per-archetype opponent effective favorability is used when available (the "what if my opponent's ad campaign is working on Suburban Professionals" case). Results reflect real vote math.

For presidential races:

Player polls are home-state projections only. The actual presidential vote model is different: it runs per electoral unit with party-weighted positions, steep party-org curves, state-lean adjustments, swing-state ground-game bonuses, and does not apply the FPTP spoiler step. Your poll shows what your home state would do; it does not forecast the Electoral College.

FPTP spoiler in polls

In FPTP states, polls show the spoiler effect's actual numbers. If you're a Democratic candidate in a race with a viable Green opponent, your poll shows:

• Your raw group-level allocation.
• The FPTP adjustment taking 4% of the Green's group votes and redistributing: either subtracting from you (if you're the closest major party) or leaving you alone (if you're the Republican instead).
• The final effective vote share.

This is the same math the turn processor will run when the election actually resolves. It's not a "worst case," it's the case.

Polling cooldowns and limits

• No hard cooldown on polls: you can poll multiple times per turn if you can afford the actions and funds.
• Diminishing strategic value. Polling the same race 3 times in a row wastes resources because your stats haven't moved much.
• Election must be active. You cannot poll an election that hasn't started the primary phase or that has resolved.

Reading a poll output

Quick Poll topline:

• Shows projected vote percentage for each candidate in the field.
• Your own projection is shown as both a percentage and a raw vote count.
• "Best 5 groups": demographics where your relative share is highest.
• "Worst 5 groups": demographics where your relative share is lowest. These are your ad-targeting and canvassing priorities.

Full Poll output:

• Per-archetype rows with size, turnout, your appeal, your reach, your approval, and the final allocated vote count.
• A separate column showing the same numbers for each opponent.
• A "gap analysis" section identifying the largest deltas.

Tactical uses

1. Declaring intelligence. Quick Poll before any primary declaration. If the topline shows you at 15% and an NPP at 65%, re-evaluate.
2. Mid-primary health check. One Full Poll 24-48 hours before the primary deadline to confirm you're winning the primary score. If you're close, a burst of Campaign actions in the final hours often seals it.
3. Early-general diagnostic. Full Poll at the start of the general. Identify the weakest 3 demographics. Invest ads/canvassing there for the middle of the general.
4. Closing sprint positioning. Quick Poll about 10 turns out from election end. If you're up 5+ points, coast; if you're tied or behind, the final 4 turns matter disproportionately, so prepare a burst.
5. Opposition scouting. When a competitor declares in your primary or general, a Full Poll shows their appeal breakdown too. Use it to understand where they'll pull votes from.

Common polling mistakes

• Skipping Quick Poll before declaring. You walk in blind. Almost always the wrong call.
• Over-polling. Three Quick Polls a day is expensive and the stats haven't moved. One poll per 6-12 turns is usually enough.
• Treating presidential polls as Electoral College forecasts. They're not: they're home-state projections. Use them to understand your home state's lean, not to project 270.
• Ignoring the "worst 5 groups" list. This is the most actionable output of any poll. Each of those 5 groups is a candidate for targeted ads or canvassing.

Related

• Stats & Actions: Poll action costs and stat effects.
• Election Mechanics: The vote math polls replicate.
• Demographics & Targeting: Group composition and appeal math.
• Canvassing: Acting on poll findings to boost turnout.
• Fundraising & Ads: Acting on poll findings to boost favorability.
• Campaign Strategy: Upgrades and long-game planning.