Decision Desk TSR
Decision Desk TSR was founded during the November 2025 Class 2 Mod Election by Reddit user PickleArtGeek to "call mod elections and props." Our team consists of the most renowned number experts TSR has to offer.
BaseCalc pools every pollster tracking a race into one weighted average; ProbCalc turns it into win probabilities. Write-in candidates are discounted 45% (✎). Methodology →
Running 20,000 simulations…
Rolling 3-Day Average — All Pollsters
BaseCalc & ProbCalc
Decision Desk TSR's own composite tools pool every pollster tracking a race (Spinner Insights included, alongside EggPolls, campaign-run polls, etc.) into one weighted average and a Monte Carlo win-probability model. This is DDTSR's aggregation work, separate from any single pollster's own releases — see Spinner Insights Polls for SI's individual poll releases. The full write-up is Understanding ProbCalc.
Write-ins and dropped candidates
Candidates running as a declared write-in (flagged per-election, e.g. Class 7 July's MaxFlares) get a 45% methodology discount applied to their polled share before entering the average, chart and ProbCalc — write-in support is well known to poll hotter than it performs at the actual ballot. They are marked with ✎ in the average row. Raw poll numbers in the table are left untouched.
Candidates absent from the most recent poll are treated as having dropped out (or not yet declared) and excluded from the current average, bars and ProbCalc — their earlier polling still shows on the trend line up to the point they stopped being tracked.
How we call races
Every race Decision Desk TSR calls is marked with the method used, visible on TSR Elects:
- Outright — a candidate has secured a majority (or, under FPTP, an insurmountable plurality) and no further tabulation can change the outcome.
- Auto-runoff (RCV) — under ranked-choice contests, once first-choice totals are final we simulate the instant-runoff elimination order to project a winner before every round is manually re-tabulated.