← Back to scans

Will SNP win above 64 seats in the 2026 Scottish Parliament election?

KXSNPMAJORITY-26MAY07-A64 · Elections · 2026-04-24
45%
Agent
47%
Market Price
-2.0%
Edge
40%
Confidence
Volume: 21,841
Spread: 1.0c
Days to resolution: 378
Markets in event: 4
Final Rationale
I ACCEPT the critique's core point: my sq1 directly addressed the resolution question at 0.45, and incorporating sq2 (not required for the outcome) and sq3 (duplicative of sq1) artificially dragged my combined estimate down — this was double-counting. I also PARTIALLY ACCEPT the tactical voting critique, since YouGov's MRP already incorporates observed preferences. The underlying evidence (YouGov MRP: 67 seats/89% majority; More in Common MRP: 56 seats; SNP polling ~35.8% with Reform fragmenting opposition at 17.6%) genuinely supports a near-coin-flip around the 65-seat threshold. My corrected estimate of 0.45 is within 3pp of the market's 0.48, with the small remaining gap reflecting my slightly heavier weight on More in Common's downside projection and AMS's structural bias against majorities.
Price vs. prediction
How the market moved after the agent's call. Solid blue line is the actual market price; dashed line is the agent's static forecast at $t{=}0$.
$n = 58$ follow-ups
1. Planner Opus · Fermi decomposition + tool plan
Tools chosen: article_search web_search kalshi_data kalshi_orderbook wikipedia_lookup code_execution
Sub-questions (Fermi decomposition)
  1. Will the SNP win a majority of seats (65+) in the 2026 Scottish Parliament election based on current polling trends? (w=0.45)
  2. Will the SNP's vote share in 2026 exceed their 2021 performance (which yielded 64 seats)? (w=0.30)
  3. Will opposition vote fragmentation (Labour, Conservative, Green, Alba splits) be sufficient to allow SNP to win 65+ seats? (w=0.25)
Planner reasoning
This question asks whether SNP will win 65+ seats (above 64) in the May 2026 Scottish Parliament election, which is one seat beyond what they achieved in 2021 and constitutes a parliamentary majority (out of 129 total seats). Key analytical angles: (1) Historical base rate: SNP has crossed the 65-seat threshold only once (2011, with 69 seats) in three elections since 2011; in 2016 they got 63 and in 2021 they got 64. (2) Electoral mechanics: The Scottish Parliament uses the Additional Member System (AMS), which is semi-proportional and makes outright majorities difficult — the 2011 result was widely seen as a statistical anomaly. The D'Hondt regional list method actively penalizes large parties. (3) Current political environment: SNP has faced significant headwinds since 2021 — leadership turmoil (Sturgeon resignation, Humza Yousaf's short tenure, John Swinney's return), police investigations into SNP finances, Scottish Labour resurgence under Anas Sarwar, and broader UK political shifts post-2024 general election. (4) Polling trends need examination — recent Scottish polls and seat projections are critical. If SNP is polling significantly below their 2021 vote share, winning more seats becomes very unlikely. (5) Temporal analysis: The election is scheduled for May 7, 2026 (about 2 weeks away from today's date of April 24, 2026 — campaigns are in final stages). Results would be known quickly after election day. The key mechanism: SNP needs to both hold their constituency seats AND win enough regional list seats, which is structurally harder when they perform well on constituencies. The question resolves based on actual election results. Given AMS structural barriers, historical rarity of SNP majorities, and recent political headwinds, the prior should be below 50%, but current polling data is essential.
2. Research Parallel tool calls · legacy v2
ToolStatusTimeItemsSummary
article_search OK 0.7s -
web_search OK 61.6s -
kalshi_data OK 1.5s -
kalshi_orderbook OK 0.2s -
wikipedia_lookup OK 1.8s -
code_execution OK 0.0s -
3. Evidence Synthesis Sonnet · 17 items · legacy v2
#StrengthCredibilityDirectionSourceClaimPriced In
1 STRONG 82 ↑ UP web_search YouGov MRP poll (fieldwork March 23 – April 8, 2026) projects SNP winning 67 seats — a narrow majority — with SNP winning a majority in 89% of simulations. Yes
2 STRONG 82 ↓ DOWN web_search More in Common MRP poll (fieldwork March 24 – April 10, 2026) projects SNP winning only 56 seats, well below the 65-seat majority threshold, remaining largest party but in minority. Yes
3 STRONG 80 ↑ UP web_search PollCheck 5-poll moving average as of April 20, 2026 shows SNP on 35.8% vote share, with Labour and Reform UK tied at 17.6% each, Conservatives and Lib Dems at 10.6% each. Yes
4 MODERATE 90 ↓ DOWN code_execution Historical base rate of SNP winning 65+ seats is 1 out of 3 elections (2011: 69 seats; 2016: 63; 2021: 64), giving a raw base rate of 33.3%. Yes
5 STRONG 92 ↓ DOWN wikipedia_lookup The Scottish Parliament uses the Additional Member System (AMS/MMP), which is semi-proportional and structurally makes outright majorities difficult for large parties due to the compensatory regional list mechanism. Yes
6 MODERATE 78 ↓ DOWN web_search YouGov's head of elections Patrick English warned that 'much could change between now and May 7, with tactical voting potentially playing a significant role,' noting that in 60 of 66 projected SNP constituency wins, combined opposition votes outnumber SNP votes. Yes
7 MODERATE 68 ↑ UP article_search As of January 2026, Scottish Labour trails the SNP far behind in the polls and could face a fight to stay in third place, indicating SNP's dominant polling position entering the campaign. Yes
8 STRONG 78 ↓ DOWN web_search SNP is polling at approximately 35.8% on the constituency vote as of April 20, 2026; in 2021 they won 47.7% of constituency votes and 40.3% of regional list votes yielding 64 seats — current polling suggests a lower vote share than 2021. Yes
9 STRONG 80 ↓ DOWN web_search More in Common projects SNP on 56 seats (down from 64 in 2021), explicitly stating they will 'lose seats' compared to 2021 — implying vote share is lower than 2021 performance. Yes
10 MODERATE 75 ↑ UP web_search YouGov MRP projects SNP on 67 seats (up from 64 in 2021), suggesting their vote share distribution — even if lower on raw percentage — could be more efficiently spread to win more seats under AMS due to fragmented opposition. Yes
11 STRONG 82 ↑ UP web_search Reform UK surged from 0.2% in 2021 to ~17.6% in 2026 polling, splitting the right-of-centre vote and significantly fragmenting the opposition landscape compared to 2021. Yes
12 STRONG 80 ↑ UP web_search Reform UK is projected to win approximately 20 MSPs in YouGov's median estimates, becoming the second-largest party and splitting the right-of-centre vote with the Conservatives (10.6%), fragmenting opposition substantially. Yes
13 MODERATE 75 ↑ UP web_search Labour polls at only 17.6% — compared to around 22% in 2021 at Holyrood — indicating the main opposition challenger is weaker than in 2021, enabling SNP to hold more constituency seats. Yes
14 STRONG 80 ↓ DOWN web_search YouGov notes that in 60 of 66 projected SNP constituency wins, combined opposition votes (Labour + Conservative + Reform + Lib Dem) outnumber the SNP — suggesting tactical voting could consolidate opposition and unseat SNP constituency MSPs. Yes
15 MODERATE 75 ↓ DOWN web_search More in Common projects the Greens winning 8 seats including first constituency wins, and together with SNP's projected 56 seats still being one seat short of majority — showing Green fragmentation from SNP does not help SNP win 65+. Yes
16 MODERATE 72 ↑ UP web_search John Swinney's SNP recovered in polls after post-2024 general election lows, benefiting from unpopularity of both Keir Starmer and Anas Sarwar, stabilizing SNP support heading into the campaign. Yes
17 STRONG 72 ↑ UP kalshi_data Kalshi prediction market currently prices SNP winning above 64 seats at 48%, up +21% over the past 30 days, reflecting a significant shift toward YES as MRP polls emerged. Yes
Information Gaps
  • No detailed breakdown of SNP's 2026 constituency vs. regional list polling separately — critical for AMS seat projections
  • No data on whether tactical voting coordination is actually occurring between Labour/Conservative/Reform/Lib Dem voters in key marginal constituencies
  • No post-April 20 polling data to confirm whether trends have shifted in the final two weeks before the May 7 election
  • No information on Alba party polling or seat projections — their split from SNP on the list could affect SNP regional list performance
  • No seat-by-seat marginality data to assess how many constituencies are genuinely competitive between SNP and nearest rival
  • No information on campaign events, debates, or late-breaking issues (scandals, endorsements) that could shift votes in the final two weeks
  • Conflicting MRP models (YouGov: 67 seats / More in Common: 56 seats) with no clear resolution — the 11-seat gap between models represents massive uncertainty
Key Uncertainties
  • The two MRP models (YouGov projecting 67 seats with 89% majority probability vs. More in Common projecting 56 seats) are in sharp conflict — which model is more accurate is the central uncertainty
  • Tactical voting: If opposition voters coordinate in key marginal constituencies, SNP could lose many of the 60 seats where combined opposition outnumbers SNP, collapsing the majority
  • Reform UK's actual vote share on election day — if they underperform polls (as new parties sometimes do), the right-of-centre vote consolidates back to Conservatives and Labour, potentially hurting SNP in constituencies
  • Whether SNP's lower raw vote share (~35.8%) compared to 2021 (~47.7% constituency) translates to fewer constituency seats even with fragmented opposition
  • Final campaign dynamics in the last two weeks before May 7 — debates, scandals, or momentum shifts that could move polling by 2-3 points in either direction
  • AMS list seat mechanics: SNP winning more constituencies actually reduces their regional list seat entitlement under D'Hondt, making a net majority very sensitive to small changes in constituency wins
Market Context
Price: 0.48 · 7d: +0.00 · 30d: +0.21
Volume: MODERATE · Depth: HIGH
Reliability: MODERATE-HIGH — reasonable volume (1133 contracts/day average), high orderbook depth (24,555), significant 30-day price movement likely driven by MRP poll releases. The sharp +21pp rise in 30 days reflects the YouGov MRP publication. Current price near 48% suggests the market is pricing near-coin-flip uncertainty, consistent with the conflicting MRP models.
4. Independent Forecaster Opus · No market price · legacy v2 Did NOT see market price
Independent estimate: 37% (40% confidence)
base rate
38%
35% conf
evidence driven
38%
40% conf
contrarian
34%
45% conf
Spread
4pp
Low disagreement
sq1: Will the SNP win a majority of seats (65+) in the 2026 Scottish Parliament election based on current polling trends?
45%
Base rate: SNP achieved 65+ only once in 3 Holyrood elections since 2011 (33%). AMS structurally penalizes large parties. Two MRP models conflict sharply: YouGov projects 67 seats with 89% majority probability (pushes UP ~25%); More in Common projects 56 seats (pushes DOWN ~15%). The 11-seat gap is enormous. Current 35.8% vote share is far below 2021's 47.7% constituency performance (pushes DOWN). But Reform splitting the right-of-centre vote massively fragments opposition (pushes UP). YouGov's warning that opposition votes exceed SNP in 60/66 wins suggests tactical voting risk. Averaging the two credible MRPs gives ~61 seats, just below majority. With high uncertainty and MRPs roughly bracketing the threshold, I estimate ~45%.
sq2: Will the SNP's vote share in 2026 exceed their 2021 performance?
8%
Base rate for incumbent party holding vote share after 17+ years in power is quite low (~20-25%). Current polling shows SNP at 35.8%, roughly 12 points below 2021's 47.7% constituency share and ~4-5 points below their list share. No credible poll shows SNP matching 2021. Both MRPs agree vote share is lower. Final two weeks unlikely to produce 10+ point swing. Estimate: 8%.
sq3: Will opposition vote fragmentation be sufficient to allow SNP to win 65+ seats?
48%
This is essentially asking the same question as sq1 from the fragmentation angle. Reform at 17.6% vs. 0.2% in 2021 is dramatic fragmentation favoring SNP in FPTP constituencies. Labour weakened to 17.6% (vs ~22% in 2021). YouGov's model — which accounts for fragmentation — projects majority. But tactical voting could collapse this (60/66 seats have combined opposition > SNP). Greens taking Green-leaning SNP voters also hurts. This mechanism is genuine but not certain to deliver. Estimate: 48%.
Show full reasoning trace
base rate: SNP has won 65+ seats once in three elections since 2011 (33%). AMS proportional mechanics make majorities rare — only one majority in all Holyrood history. Base rate ~30-35%.
evidence updates: YouGov MRP (89% majority probability, 67 seats) pushes strongly UP. More in Common MRP (56 seats) pushes strongly DOWN. Reform UK fragmenting opposition pushes UP substantially. Lower raw SNP vote share (35.8% vs 47.7%) pushes DOWN. Tactical voting risk in 60/66 seats pushes DOWN. These largely offset, with the two MRP models bracketing the threshold.
combination method: Weighted average with adjustment for correlation and the fact that sq2 (exceeding 2021 vote share) is nearly impossible but not actually required for SNP to win 65+ seats given opposition fragmentation.
final: Conflicting high-quality MRP evidence creates genuine uncertainty around the 65-seat threshold. Averaging YouGov (67) and More in Common (56) gives ~61-62 seats — just below majority. With 2 weeks of campaign remaining and tactical voting risk, I estimate 38% probability of 65+ seats. Confidence is LOW-MEDIUM given the sharp MRP disagreement.
ensemble: {'ensemble_count': 3, 'perspectives': {'base_rate': {'p_yes': 0.38, 'confidence': 0.35}, 'evidence_driven': {'p_yes': 0.38, 'confidence': 0.4}, 'contrarian': {'p_yes': 0.34, 'confidence': 0.45}}, 'spread': 0.04, 'individual_p_yes': {'base_rate': 0.38, 'evidence_driven': 0.38, 'contrarian': 0.34}}
5. Devil's Advocate Sonnet · legacy v2 Market price revealed
Math Error Reasoning Flaw
Challenges
  1. The question asks about 'above 64 seats' which means 65+, but the forecaster should double-check this framing. More in Common projects 56 (clearly below), YouGov projects 67 (clearly above). The averaging approach of 61 seats being 'just below majority' is reasonable, but the forecaster inconsistently reports the final number (0.367 combined, 0.38 in reasoning trace, 0.45 in sq1). This inconsistency suggests the combination method isn't well-defined.
  2. The combination method is unclear. sq1 directly asks the resolution question at 45%, but the combined forecast is 36.7% — lower than sq1. Why would incorporating sq2 (which isn't required for the outcome, as the forecaster acknowledges) and sq3 (essentially the same question) pull the estimate DOWN from sq1? This looks like double-counting of downside factors.
  3. Simple averaging of two MRPs (YouGov 67 + MiC 56)/2 = 61.5 seats treats them as equally reliable. YouGov has a longer track record with MRP methodology in UK elections; weighting should potentially favor it, though MiC's recent Scottish work is also credible. The forecaster doesn't justify equal weighting.
  4. The forecaster notes tactical voting risk (60/66 seats have combined opposition > SNP) as pushing DOWN, but YouGov's MRP *already accounts for* observed voting patterns. Treating this as an additional downside on top of the MRP projection may be double-counting.
  5. With 2 weeks until the election (May 7), late polling movement uncertainty is real but bounded. The forecaster applies this as a downside factor, but late movement could go either direction — this should widen uncertainty, not specifically lower the point estimate.
Suggested adjustment: +7pp
The combined forecast of 0.367 is internally inconsistent with sq1's 0.45 estimate, which more directly answers the resolution question. sq2 is acknowledged as not required for the outcome, and sq3 is essentially duplicative of sq1. The combination appears to incorrectly drag the estimate below the direct sub-question estimate. Adjusting upward toward sq1's estimate (~0.43-0.45) fixes this internal inconsistency. This happens to align with the market (0.48), but the rationale is the math error, not market conformity.
Market comparison: Forecaster: 36.7%, Market: 48%. The 11pp divergence is partly explained by an internal inconsistency in the forecaster's combination method — their own sq1 direct estimate (45%) is much closer to the market. The market likely reflects a cleaner aggregation of the same MRP evidence. The divergence is not justified by new information the forecaster has; it appears to stem from combination methodology issues.
6. Aggregator Opus · final calibrated forecast
45% (40% confidence)
I ACCEPT the critique's core point: my sq1 directly addressed the resolution question at 0.45, and incorporating sq2 (not required for the outcome) and sq3 (duplicative of sq1) artificially dragged my combined estimate down — this was double-counting. I also PARTIALLY ACCEPT the tactical voting critique, since YouGov's MRP already incorporates observed preferences. The underlying evidence (YouGov MRP: 67 seats/89% majority; More in Common MRP: 56 seats; SNP polling ~35.8% with Reform fragmenting opposition at 17.6%) genuinely supports a near-coin-flip around the 65-seat threshold. My corrected estimate of 0.45 is within 3pp of the market's 0.48, with the small remaining gap reflecting my slightly heavier weight on More in Common's downside projection and AMS's structural bias against majorities.