VotePredictor Elections

Forecast a 2026 race

forecast_race

Read-onlyIdempotent

Current VotePredictor forecast for one 2026 US election: win probability, predicted margin, whether it comes from polls or fundamentals, how far it has moved this cycle, and the candidates. Accepts a state name or code ('Ohio', 'OH') or a district ('CA-12'). Returns an explicit refusal for races where a candidate has left the ballot and no forecast is published.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "where": "<where>"
    },
    "name": "forecast_race"
  }
}

Input Schema

Name	Required	Description	Default
`where`	Yes	Where the race is: a state name ('Ohio'), a two-letter state code ('OH'), or a House district ('CA-12'). Pass what the user said.
`office`	No	Which office: SEN for U.S. Senate, GOV for governor, HOUSE for a U.S. House district. Any case; 'senate' and 'governor' also work.

Output Schema

Name	Required	Description
`url`	Yes	Cite this
`as_of`	Yes	Date of the model run behind this number
`office`	Yes
`rating`	Yes	Safe D / Likely D / Lean D / Toss-up / Lean R / Likely R / Safe R
`source`	Yes	'polls' where a poll-driven forecast exists, otherwise 'fundamentals' (lean, incumbency, national environment)
`n_polls`	Yes
`p_basis`	Yes	What set the win probability. 'shop_ratings': an unpolled House district whose probability blends in the rating shops' current ratings while pred_margin stays the lean model's, so the two can point different ways. Say so rather than treating it as an inconsistency.
`race_id`	Yes	Canonical id, {year}_{OFFICE}_{location}
`summary`	Yes	One-paragraph answer suitable for reading aloud
`location`	Yes	State code, or district id for the House
`p_dem_win`	Yes	Democratic win probability, 0.01-0.99. Never 0 or 1: an election that has not happened is not certain.
`state_name`	Yes
`pred_margin`	Yes	Predicted Dem-minus-Rep margin in points
`dem_candidate`	Yes
`model_version`	Yes
`rep_candidate`	Yes
`actual_matchup`	Yes	The contest actually on the ballot, when it is not Democrat vs Republican
`movement_basis`	Yes
`movement_points`	Yes	Change in the PUBLISHED win probability, in points, positive toward Democrats, since the current model began (the day after its latest revision). Null when the current model has under two days on file.
`no_forecast_code`	Yes	independent_contender: no Democrat is on the ballot and an independent challenges the Republican (nobody left the ballot; the model has no slot for an independent). nominee_changed: the polls on file ask about a former nominee. no_major_party_nominee: a side has nobody yet.
`forecast_published`	Yes	False when this site declines to publish a probability for the race
`no_forecast_reason`	Yes	Why nothing is published, in words. Do not substitute your own estimate.

Schema Changelog

Changes observed during successful MCP inspections.

Changed1 schema field changedNov 2, 2026

changedOutput schema / properties / movement_points / description

Previous value: -"BACKCAST change in win probability, in points, positive toward Democrats: today's model rerun on the polls available at each past date. Not how the published number moved; do not describe it as the forecast's history."New value: +"Change in the PUBLISHED win probability, in points, positive toward Democrats, since the current model began (the day after its latest revision). Null when the current model has under two days on file."

Changed11 schema fields changedOct 29, 2026

changedInput schema / properties / office / description

Previous value: -"Which office: SEN for U.S. Senate, GOV for governor, HOUSE for a U.S. House district."New value: +"Which office: SEN for U.S. Senate, GOV for governor, HOUSE for a U.S. House district. Any case; 'senate' and 'governor' also work."

changedInput schema / required

Previous value: -[
-  "where",
-  "office"
-]New value: +[
+  "where"
+]

addedOutput schema / properties / actual_matchup

Added value: +{
+  "anyOf": [
+    {
+      "type": "string"
+    },
+    {
+      "type": "null"
+    }
+  ],
+  "description": "The contest actually on the ballot, when it is not Democrat vs Republican"
+}

addedOutput schema / properties / as_of

Added value: +{
+  "anyOf": [
+    {
+      "type": "string"
+    },
+    {
+      "type": "null"
+    }
+  ],
+  "description": "Date of the model run behind this number"
+}

addedOutput schema / properties / model_version

Added value: +{
+  "anyOf": [
+    {
+      "type": "string"
+    },
+    {
+      "type": "null"
+    }
+  ]
+}

addedOutput schema / properties / movement_basis

Added value: +{
+  "anyOf": [
+    {
+      "type": "string"
+    },
+    {
+      "type": "null"
+    }
+  ]
+}

changedOutput schema / properties / movement_points / description

Previous value: -"Change in win probability in POINTS since the first snapshot this cycle, positive toward Democrats"New value: +"BACKCAST change in win probability, in points, positive toward Democrats: today's model rerun on the polls available at each past date. Not how the published number moved; do not describe it as the forecast's history."

addedOutput schema / properties / no_forecast_code

Added value: +{
+  "anyOf": [
+    {
+      "enum": [
+        "independent_contender",
+        "nominee_changed",
+        "no_major_party_nominee",
+        "unknown_race",
+        "unreadable_location"
+      ],
+      "type": "string"
+    },
+    {
+      "type": "null"
+    }
+  ],
+  "description": "independent_contender: no Democrat is on the ballot and an independent challenges the Republican (nobody left the ballot; the model has no slot for an independent). nominee_changed: the polls on file ask about a former nominee. no_major_party_nominee: a side has nobody yet."
+}

changedOutput schema / properties / no_forecast_reason / description

Previous value: -"Why nothing is published, e.g. a candidate has left the ballot. Do not substitute your own estimate."New value: +"Why nothing is published, in words. Do not substitute your own estimate."

addedOutput schema / properties / p_basis

Added value: +{
+  "anyOf": [
+    {
+      "enum": [
+        "polls",
+        "fundamentals",
+        "cook_rating",
+        "shop_ratings"
+      ],
+      "type": "string"
+    },
+    {
+      "type": "null"
+    }
+  ],
+  "description": "What set the win probability. 'shop_ratings': an unpolled House district whose probability blends in the rating shops' current ratings while pred_margin stays the lean model's, so the two can point different ways. Say so rather than treating it as an inconsistency."
+}

changedOutput schema / required

Previous value: -[
-  "race_id",
-  "forecast_published",
-  "no_forecast_reason",
-  "office",
-  "location",
-  "state_name",
-  "p_dem_win",
-  "rating",
-  "pred_margin",
-  "source",
-  "n_polls",
-  "dem_candidate",
-  "rep_candidate",
-  "movement_points",
-  "url",
-  "summary"
-]New value: +[
+  "race_id",
+  "forecast_published",
+  "no_forecast_reason",
+  "no_forecast_code",
+  "actual_matchup",
+  "office",
+  "location",
+  "state_name",
+  "p_dem_win",
+  "rating",
+  "pred_margin",
+  "source",
+  "p_basis",
+  "n_polls",
+  "dem_candidate",
+  "rep_candidate",
+  "movement_points",
+  "movement_basis",
+  "as_of",
+  "model_version",
+  "url",
+  "summary"
+]

First observedSep 1, 2026

TDQS

A4.1/5.0

Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare readOnly/idempotent/non-destructive, so safety is handled. The description adds genuinely new behavioral facts: it says whether a forecast derives from polls or fundamentals, and that certain races return an explicit refusal when a candidate has left the ballot — a non-obvious edge case an agent needs to interpret output correctly.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The output payload is front-loaded in the first clause and the refusal caveat is the last sentence, so the ordering is sensible and nearly every clause carries information. The opening sentence is dense with enumerated fields, but nothing is filler.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

An output schema exists, so the description needn't explain return structure, and the refusal behavior plus input formats round out the operational picture. The one real gap is not acknowledging that a state can hold multiple 2026 races requiring the `office` discriminator, which affects correct invocation.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100% and both parameters are fully documented, so the schema carries the load. The description reinforces the `where` formats with the same examples the schema uses, which is mild redundancy, and never mentions the `office` parameter at all — an agent must open the schema to learn SEN/GOV/HOUSE exists.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource ('Current VotePredictor forecast for one 2026 US election') plus the exact payload: win probability, margin, source, movement, candidates. This is clearly distinguishable from the aggregate-sounding sibling chamber_outlook and from member_profile.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It gives concrete usage context by naming accepted input forms ('Ohio', 'OH', 'CA-12') and the refusal case, so an agent knows the tool's operating envelope. It stops short of explicitly contrasting with chamber_outlook for multi-race or aggregate questions, so the alternative routing is left to inference.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All VotePredictor Elections tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources

Need Help?