Datasets:

AFOS-Analytics1
/

peru-2026-electoral-divergence

@@ -1,115 +1,120 @@
----
-license:
-- cc-by-4.0
-- apache-2.0
-language:
-- es
-- en
-- pt
-pretty_name: "AFOS · Peru 2026 Electoral Divergence"
-tags:
-- elections
-- peru
-- prediction-markets
-- polls
-- political-risk
-- divergence
-- open-data
-- latin-america
----
-![AFOS · Peru 2026 Electoral Divergence](banner.png)
-# AFOS · Peru 2026 Electoral Divergence Dataset
-🌐 **[English](#english) · [Español](#español) · [Português](#português)**
-Open dataset cross-referencing **opinion polls × prediction markets** for Peru's **2026 general election** (first round 12 April 2026; runoff 7 June 2026, Keiko Fujimori vs Roberto Sánchez), built in the same spirit as the AFOS Brazil 2026 dataset: sources are reported side by side with **explicit divergence**, not blended into a single average.
-Maintained by **[AFOS Analytics](https://afos-analytics.com)**. This is part of AFOS's expansion of its electoral-divergence method beyond Brazil. *No personal data — only public electoral information.*
----
-## English
-![Polymarket implied probability of winning over the campaign](odds-trajectory.png)
-![Market probability of winning versus poll vote share on the eve of the vote](odds-snapshot.png)
-**Contents (start with the polls):**
-| Path | Rows | Content |
-|------|------|---------|
-| `polls/peru-first-round-polls.csv` | 327 | First-round voting intention, **long format** (one row per candidate × poll), **all 14 candidates**, 36 polls, Jan→Apr 2026. |
-| `polls/peru-runoff-polls.csv` | 16 | Runoff head-to-head (Fujimori vs Sánchez), Apr→Jun 2026. |
-| `polls/peru-polls.json` | — | Full structured polls (first round + runoff) with methodology. |
-| `data/peru-market-odds-timeseries.csv` | 2,490 | Daily Polymarket win-probability per candidate (16 candidates, Dec 2025→Jun 2026) from the "Peru Presidential Election Winner" market. |
-| `data/peru-divergence-timeseries.csv` | 324 | **Market × poll divergence** per candidate — each first-round poll joined to the candidate's market odds on its date. |
-| `data/peru-poly-raw.json` | — | Raw Polymarket payload (event + per-candidate price histories), kept for provenance. |
-<sub>Market data fetched from Polymarket's gamma-api + clob (US-resolving function). Divergence covers the **first round**; no clean head-to-head runoff win-probability market exists to join the runoff polls.</sub>
-### ⚖️ Notable divergences (why divergence beats the average)
-The point of this dataset is the **gap** between what the market prices (probability of *winning*) and what polls measure (first-round *vote share*) — read across the full daily series, not a single snapshot.
-- **Rafael López Aliaga — the market's sustained favorite that the polls never matched.** For months the market priced his *win probability* at **40–55%** while his first-round vote-share polling stayed in the **low-to-mid teens (7–17%)**. He placed third and **missed the runoff** — the market's conviction ran well ahead of his actual support. A real, sustained divergence.
-- **Keiko Fujimori — polls ~16% × market ~22%:** steadier in both; she led the first round and advanced. The 7 June runoff against Roberto Sánchez ended in a virtual tie (~50.1% × ~49.9%, ~99.4% counted), with the JNE yet to proclaim an official winner.
-- **Ricardo Belmont — poll 9% × market 5.1%:** the market priced him *below* his vote share — never a contender in its eyes.
-- **A caution on noise:** outsider Carlos Álvarez briefly spiked to 31.6% in the market on a single day (5 Apr) against ~8% in polls, then fell to low single digits within days. Thin-market prints can diverge sharply *without* being a sustained signal — which is exactly why the full daily series matters, not one snapshot.
-**The reading:** the spread is the signal — but reading it means telling the *sustained* gap (López Aliaga) from transient noise (Álvarez). A blended average hides both.
-**Pollsters covered:** Ipsos Perú, Datum Internacional, CPI, IEP, CIT, Imasolu, CELAG, IDICE, CB Global Data (with publishing client where applicable).
-**Provenance & method:** poll figures are compiled deterministically (rowspan/colspan-aware HTML parser) from the public Wikipedia aggregation *"Opinion polling for the 2026 Peruvian general election"* and cross-checked against the [AS/COA poll tracker](https://www.as-coa.org/articles/poll-tracker-perus-2026-presidential-election); every figure traces to a named pollster's release. Market odds come from the public Polymarket markets. Nothing is imputed or smoothed; missing values are left blank.
-**License (dual):** **data** → CC BY 4.0 (`LICENSE-CC-BY-4.0`); **code/scripts** → Apache 2.0 (`LICENSE-APACHE-2.0`), matching the repo root and the Hugging Face mirror. Underlying poll numbers are facts released by the named pollsters; the Wikipedia aggregation is CC BY-SA. Please attribute **AFOS Analytics** and the **original pollsters**.
-**Cite:** *AFOS Analytics. Peru 2026 Electoral Divergence Dataset. Hugging Face, 2026. CC BY 4.0.* (see `CITATION.cff`)
-**Disclaimer:** observational research. Not investment advice, not voting guidance.
----
-## Español
-Dataset abierto que cruza **encuestas × mercados de predicción** para la **elección general del Perú 2026** (primera vuelta 12 abr 2026; segunda vuelta 7 jun 2026, Keiko Fujimori vs Roberto Sánchez), con **divergencia explícita** entre fuentes en lugar de un promedio único.
-- `polls/peru-first-round-polls.csv` — intención de voto en primera vuelta, formato largo, **los 14 candidatos**, 36 encuestas (ene→abr 2026).
-- `polls/peru-runoff-polls.csv` — segunda vuelta cara a cara (Fujimori vs Sánchez), 16 encuestas.
-- `data/peru-market-odds-timeseries.csv` / `data/peru-divergence-timeseries.csv` — probabilidad de Polymarket por candidato y divergencia mercado × encuesta.
-### ⚖️ Divergencias destacadas (por qué la divergencia supera al promedio)
-Lo importante es la **brecha** entre lo que valora el mercado (probabilidad de *ganar*) y lo que miden las encuestas (*voto* de primera vuelta) — leída en toda la serie diaria, no en una foto.
-- **Rafael López Aliaga — el favorito sostenido del mercado que las encuestas nunca confirmaron.** Durante meses el mercado valoró su *probabilidad de ganar* en **40–55%** mientras su voto de primera vuelta se quedaba en la **franja baja-media de la decena (7–17%)**. Quedó tercero y **no llegó al balotaje** — la convicción del mercado iba muy por delante de su apoyo real. Una divergencia real y sostenida.
-- **Keiko Fujimori — encuestas ~16% × mercado ~22%:** más estable en ambos; lideró la primera vuelta y avanzó. El balotaje del 7 jun frente a Roberto Sánchez terminó en empate técnico (~50,1% × ~49,9%, ~99,4% escrutado), con el JNE aún sin proclamar un ganador oficial.
-- **Ricardo Belmont — encuesta 9% × mercado 5,1%:** el mercado lo valoró *por debajo* de su voto.
-- **Una advertencia sobre el ruido:** el outsider Carlos Álvarez subió brevemente a 31,6% en el mercado en un solo día (5 abr) frente a ~8% en encuestas, y cayó a un dígito en pocos días. Los precios de mercados poco líquidos pueden divergir con fuerza *sin* ser una señal sostenida — por eso importa la serie diaria completa.
-**La lectura:** la brecha es la señal — pero leerla implica distinguir la divergencia *sostenida* (López Aliaga) del ruido transitorio (Álvarez). Un promedio oculta ambas.
-**Encuestadoras:** Ipsos Perú, Datum Internacional, CPI, IEP, CIT, Imasolu, CELAG, IDICE, CB Global Data. **Fuente:** agregación pública de Wikipedia + tracker AS/COA; cada cifra remite a la publicación de una encuestadora con nombre. **Licencia:** CC BY 4.0 (atribuir a AFOS Analytics y a las encuestadoras originales). Investigación observacional; no es asesoría de inversión ni orientación de voto.
----
-## Português
-Dataset aberto cruzando **pesquisas × mercados de previsão** para a **eleição geral do Peru 2026** (1º turno 12/abr; 2º turno 07/jun, Keiko Fujimori × Roberto Sánchez), com **divergência explícita** entre fontes. Pesquisas (14 candidatos, 36 do 1º turno + 16 do 2º turno) compiladas deterministicamente da agregação pública da Wikipedia + tracker AS/COA; odds do Polymarket. Licença CC BY 4.0 (atribuir AFOS Analytics + institutos originais). Pesquisa observacional; não é recomendação de investimento nem orientação de voto.
-### ⚖️ Divergências em destaque (por que a divergência supera a média)
-O ponto é a **diferença** entre o que o mercado precifica (probabilidade de *vencer*) e o que as pesquisas medem (*voto* de 1º turno) — lida na série diária inteira, não numa foto.
-- **Rafael López Aliaga — o favorito sustentado do mercado que as pesquisas nunca confirmaram.** Por meses o mercado precificou a *chance de vencer* dele em **40–55%** enquanto o voto de 1º turno ficava na **casa baixa-média da dezena (7–17%)**. Ficou em 3º e **não foi ao 2º turno** — a convicção do mercado ia bem à frente do apoio real. Divergência real e sustentada.
-- **Keiko Fujimori — pesquisas ~16% × mercado ~22%:** mais estável nos dois; liderou o 1º turno e avançou. O 2º turno de 07/jun contra Roberto Sánchez terminou em empate técnico (~50,1% × ~49,9%, ~99,4% apurado), com o JNE ainda sem proclamar vencedor oficial.
-- **Ricardo Belmont — pesquisa 9% × mercado 5,1%:** o mercado o precificou *abaixo* do voto.
-- **Um alerta sobre ruído:** o azarão Carlos Álvarez deu um spike breve de 31,6% no mercado num único dia (5/abr) contra ~8% nas pesquisas, e caiu pra um dígito em poucos dias. Preços de mercados pouco líquidos podem divergir forte *sem* ser sinal sustentado — por isso a série diária completa importa.
-**A leitura:** a diferença é o sinal — mas lê-la é distinguir a divergência *sustentada* (López Aliaga) do ruído passageiro (Álvarez). A média esconde as duas.
----
-**Sources / Fuentes:** Pollsters (Ipsos Perú, Datum, CPI, IEP, CIT, …) · [Wikipedia aggregation](https://en.wikipedia.org/wiki/Opinion_polling_for_the_2026_Peruvian_general_election) · [AS/COA poll tracker](https://www.as-coa.org/articles/poll-tracker-perus-2026-presidential-election) · Polymarket. Column definitions in [`DATA_DICTIONARY.md`](DATA_DICTIONARY.md).

+---
+license:
+- cc-by-4.0
+- apache-2.0
+language:
+- es
+- en
+- pt
+pretty_name: "AFOS · Peru 2026 Electoral Divergence"
+tags:
+- elections
+- peru
+- prediction-markets
+- polls
+- political-risk
+- divergence
+- open-data
+- latin-america
+---
+![AFOS · Peru 2026 Electoral Divergence](banner.png)
+# AFOS · Peru 2026 Electoral Divergence Dataset
+🌐 **[English](#english) · [Español](#español) · [Português](#português)**
+Open dataset cross-referencing **opinion polls × prediction markets** for Peru's **2026 general election** (first round 12 April 2026; runoff 7 June 2026, Keiko Fujimori vs Roberto Sánchez), built in the same spirit as the AFOS Brazil 2026 dataset: sources are reported side by side with **explicit divergence**, not blended into a single average.
+Maintained by **[AFOS Analytics](https://afos-analytics.com)**. This is part of AFOS's expansion of its electoral-divergence method beyond Brazil. *No personal data — only public electoral information.*
+---
+## English
+![Polymarket implied probability of winning over the campaign](odds-trajectory.png)
+![Market probability of winning versus poll vote share on the eve of the vote](odds-snapshot.png)
+**Contents (start with the polls):**
+| Path | Rows | Content |
+|------|------|---------|
+| `polls/peru-first-round-polls.csv` | 327 | First-round voting intention, **long format** (one row per candidate × poll), **all 14 candidates**, 36 polls, Jan→Apr 2026. |
+| `polls/peru-runoff-polls.csv` | 16 | Runoff head-to-head (Fujimori vs Sánchez), Apr→Jun 2026. |
+| `polls/peru-polls.json` | — | Full structured polls (first round + runoff) with methodology. |
+| `data/peru-market-odds-timeseries.csv` | 2,490 | Daily Polymarket win-probability per candidate (16 candidates, Dec 2025→Jun 2026) from the "Peru Presidential Election Winner" market. |
+| `data/peru-divergence-timeseries.csv` | 324 | **Market × poll divergence** per candidate — each first-round poll joined to the candidate's market odds on its date. |
+| `data/peru-poly-raw.json` | — | Raw Polymarket payload (event + per-candidate price histories), kept for provenance. |
+<sub>Market data fetched from Polymarket's gamma-api + clob (US-resolving function). Divergence covers the **first round**; no clean head-to-head runoff win-probability market exists to join the runoff polls.</sub>
+### ⚖️ Notable divergences (why divergence beats the average)
+The point of this dataset is the **gap** between what the market prices (probability of *winning*) and what polls measure (first-round *vote share*) — read across the full daily series, not a single snapshot.
+- **Rafael López Aliaga — the market's sustained favorite that the polls never matched.** For months the market priced his *win probability* at **40–55%** while his first-round vote-share polling stayed in the **low-to-mid teens (7–17%)**. He placed third and **missed the runoff** — the market's conviction ran well ahead of his actual support. A real, sustained divergence.
+- **Keiko Fujimori — polls ~16% × market ~22%:** steadier in both; she led the first round and advanced. The 7 June runoff against Roberto Sánchez ended in a virtual tie (~50.1% × ~49.9%, ~99.4% counted), with the JNE yet to proclaim an official winner.
+- **Ricardo Belmont — poll 9% × market 5.1%:** the market priced him *below* his vote share — never a contender in its eyes.
+- **A caution on noise:** outsider Carlos Álvarez briefly spiked to 31.6% in the market on a single day (5 Apr) against ~8% in polls, then fell to low single digits within days. Thin-market prints can diverge sharply *without* being a sustained signal — which is exactly why the full daily series matters, not one snapshot.
+**The reading:** the spread is the signal — but reading it means telling the *sustained* gap (López Aliaga) from transient noise (Álvarez). A blended average hides both.
+**Pollsters covered:** Ipsos Perú, Datum Internacional, CPI, IEP, CIT, Imasolu, CELAG, IDICE, CB Global Data (with publishing client where applicable).
+**Provenance & method:** poll figures are compiled deterministically (rowspan/colspan-aware HTML parser) from the public Wikipedia aggregation *"Opinion polling for the 2026 Peruvian general election"* and cross-checked against the [AS/COA poll tracker](https://www.as-coa.org/articles/poll-tracker-perus-2026-presidential-election); every figure traces to a named pollster's release. Market odds come from the public Polymarket markets. Nothing is imputed or smoothed; missing values are left blank.
+**License (dual):** **data** → CC BY 4.0 (`LICENSE-CC-BY-4.0`); **code/scripts** → Apache 2.0 (`LICENSE-APACHE-2.0`), matching the repo root and the Hugging Face mirror. Underlying poll numbers are facts released by the named pollsters; the Wikipedia aggregation is CC BY-SA. Please attribute **AFOS Analytics** and the **original pollsters**.
+**Cite:** *AFOS Analytics. Peru 2026 Electoral Divergence Dataset. Hugging Face, 2026. CC BY 4.0.* (see `CITATION.cff`)
+**Disclaimer:** observational research. Not investment advice, not voting guidance.
+---
+## Español
+Dataset abierto que cruza **encuestas × mercados de predicción** para la **elección general del Perú 2026** (primera vuelta 12 abr 2026; segunda vuelta 7 jun 2026, Keiko Fujimori vs Roberto Sánchez), con **divergencia explícita** entre fuentes en lugar de un promedio único.
+- `polls/peru-first-round-polls.csv` — intención de voto en primera vuelta, formato largo, **los 14 candidatos**, 36 encuestas (ene→abr 2026).
+- `polls/peru-runoff-polls.csv` — segunda vuelta cara a cara (Fujimori vs Sánchez), 16 encuestas.
+- `data/peru-market-odds-timeseries.csv` / `data/peru-divergence-timeseries.csv` — probabilidad de Polymarket por candidato y divergencia mercado × encuesta.
+### ⚖️ Divergencias destacadas (por qué la divergencia supera al promedio)
+Lo importante es la **brecha** entre lo que valora el mercado (probabilidad de *ganar*) y lo que miden las encuestas (*voto* de primera vuelta) — leída en toda la serie diaria, no en una foto.
+- **Rafael López Aliaga — el favorito sostenido del mercado que las encuestas nunca confirmaron.** Durante meses el mercado valoró su *probabilidad de ganar* en **40–55%** mientras su voto de primera vuelta se quedaba en la **franja baja-media de la decena (7–17%)**. Quedó tercero y **no llegó al balotaje** — la convicción del mercado iba muy por delante de su apoyo real. Una divergencia real y sostenida.
+- **Keiko Fujimori — encuestas ~16% × mercado ~22%:** más estable en ambos; lideró la primera vuelta y avanzó. El balotaje del 7 jun frente a Roberto Sánchez terminó en empate técnico (~50,1% × ~49,9%, ~99,4% escrutado), con el JNE aún sin proclamar un ganador oficial.
+- **Ricardo Belmont — encuesta 9% × mercado 5,1%:** el mercado lo valoró *por debajo* de su voto.
+- **Una advertencia sobre el ruido:** el outsider Carlos Álvarez subió brevemente a 31,6% en el mercado en un solo día (5 abr) frente a ~8% en encuestas, y cayó a un dígito en pocos días. Los precios de mercados poco líquidos pueden divergir con fuerza *sin* ser una señal sostenida — por eso importa la serie diaria completa.
+**La lectura:** la brecha es la señal — pero leerla implica distinguir la divergencia *sostenida* (López Aliaga) del ruido transitorio (Álvarez). Un promedio oculta ambas.
+**Encuestadoras:** Ipsos Perú, Datum Internacional, CPI, IEP, CIT, Imasolu, CELAG, IDICE, CB Global Data. **Fuente:** agregación pública de Wikipedia + tracker AS/COA; cada cifra remite a la publicación de una encuestadora con nombre. **Licencia:** CC BY 4.0 (atribuir a AFOS Analytics y a las encuestadoras originales). Investigación observacional; no es asesoría de inversión ni orientación de voto.
+---
+## Português
+Dataset aberto cruzando **pesquisas × mercados de previsão** para a **eleição geral do Peru 2026** (1º turno 12/abr; 2º turno 07/jun, Keiko Fujimori × Roberto Sánchez), com **divergência explícita** entre fontes. Pesquisas (14 candidatos, 36 do 1º turno + 16 do 2º turno) compiladas deterministicamente da agregação pública da Wikipedia + tracker AS/COA; odds do Polymarket. Licença CC BY 4.0 (atribuir AFOS Analytics + institutos originais). Pesquisa observacional; não é recomendação de investimento nem orientação de voto.
+### ⚖️ Divergências em destaque (por que a divergência supera a média)
+O ponto é a **diferença** entre o que o mercado precifica (probabilidade de *vencer*) e o que as pesquisas medem (*voto* de 1º turno) — lida na série diária inteira, não numa foto.
+- **Rafael López Aliaga — o favorito sustentado do mercado que as pesquisas nunca confirmaram.** Por meses o mercado precificou a *chance de vencer* dele em **40–55%** enquanto o voto de 1º turno ficava na **casa baixa-média da dezena (7–17%)**. Ficou em 3º e **não foi ao 2º turno** — a convicção do mercado ia bem à frente do apoio real. Divergência real e sustentada.
+- **Keiko Fujimori — pesquisas ~16% × mercado ~22%:** mais estável nos dois; liderou o 1º turno e avançou. O 2º turno de 07/jun contra Roberto Sánchez terminou em empate técnico (~50,1% × ~49,9%, ~99,4% apurado), com o JNE ainda sem proclamar vencedor oficial.
+- **Ricardo Belmont — pesquisa 9% × mercado 5,1%:** o mercado o precificou *abaixo* do voto.
+- **Um alerta sobre ruído:** o azarão Carlos Álvarez deu um spike breve de 31,6% no mercado num único dia (5/abr) contra ~8% nas pesquisas, e caiu pra um dígito em poucos dias. Preços de mercados pouco líquidos podem divergir forte *sem* ser sinal sustentado — por isso a série diária completa importa.
+**A leitura:** a diferença é o sinal — mas lê-la é distinguir a divergência *sustentada* (López Aliaga) do ruído passageiro (Álvarez). A média esconde as duas.
+---
+**Sources / Fuentes:** Pollsters (Ipsos Perú, Datum, CPI, IEP, CIT, …) · [Wikipedia aggregation](https://en.wikipedia.org/wiki/Opinion_polling_for_the_2026_Peruvian_general_election) · [AS/COA poll tracker](https://www.as-coa.org/articles/poll-tracker-perus-2026-presidential-election) · Polymarket. Column definitions in [`DATA_DICTIONARY.md`](DATA_DICTIONARY.md).
+## Structural context (World Bank)
+Beyond the divergence data, this dataset ships `data/peru-structural-context.csv`: official, open World Bank indicators that frame the country — **governance** (Worldwide Governance Indicators, 0-100 scale) plus **economy & education** (World Development Indicators: population, GDP, GDP per capita, inflation, public education spending, expected years of schooling). These are annual structural indicators that contextualize the country; they do **not** predict the electoral outcome. Columns are documented in `DATA_DICTIONARY.md`.