Note: What follows is a falsification matrix for the essay, “A Constraint Story: Why Colombia’s Election Defies Standard Forecasting.”
Testing the Constraint Theory Against Observable Data
LIVE UPDATE: January 26, 2026 – Prediction Market Analysis
Current Market Probabilities (Median Forecasts)
Source: Prediction market opened 01/16/26, data as of 01/26/26
| Candidate | Median Probability | Range (Low-High) | Essay Expectation | Status |
|---|---|---|---|---|
| Abelardo de la Espriella | 46% | 25-58% | ~35% parity with Cepeda | ⚠️ YELLOW FLAG – Stronger than predicted |
| Iván Cepeda | 31.5% | 25-44% | ~35% parity with De la Espriella | ⚠️ YELLOW FLAG – Weaker than predicted |
| Sergio Fajardo | 10% | 5-20% | ~14% (trailing but viable) | ✅ GREEN – Within expected range |
| Paloma Valencia | 5.5% | 3-20% | Negligible (<3%) | ⚠️ YELLOW FLAG – Higher than institutional collapse thesis suggests |
| Another candidate | 5% | 1-15% | <5% | ✅ GREEN – Field consolidated as predicted |
Critical Divergences from Essay Framework
1. De la Espriella Dominance (46% vs 35% predicted)
What the essay predicted: Near-parity between Cepeda and the right-wing candidate, with margins of 1-4 percentage points creating “unstable bipolar equilibrium”
What markets show: De la Espriella has opened a 14.5-point lead over Cepeda (46% vs 31.5%)
Possible Interpretations:
- A) Security crisis deepening faster than essay anticipated – If Total Peace failures accelerated, De la Espriella’s “firm military action” message gains urgency
- B) Cepeda’s “succession Rope” weaker than predicted – Inheriting Petro’s coalition without Petro’s personal appeal
- C) Markets overweighting early right consolidation – Prediction markets often lag structural analysis; this could correct
- D) Essay’s bipolar compression theory wrong – One pole is clearly stronger, not unstable parity
Matrix Implication: If this persists through Feb 15, it violates the “Terminal Attractor” prediction of 1-4 point separation
2. Cepeda Underperformance (31.5% vs 35% predicted)
What the essay predicted: Cepeda consolidates left to ~35% (hard ceiling of convinced left voters), stable through campaign
What markets show: Cepeda at 31.5%, struggling to reach even the predicted floor
Possible Interpretations:
- A) Petro’s approval deteriorated further – If approval dropped below 30%, Cepeda’s succession becomes toxic
- B) Left Rope fracturing – Some Pacto Histórico factions defecting or demobilizing
- C) Runoff math pricing – Markets may be discounting Cepeda’s first-round strength because they expect him to lose the runoff decisively
- D) Polling vs market divergence – Guarumo shows Cepeda at 33.6%, markets at 31.5%; which is right?
Matrix Implication: If Cepeda can’t break 35% in polling by Feb 15, the “hard ceiling” has become a “low ceiling” (~32-34%), suggesting the Left Rope is weaker than predicted
3. Valencia’s 5.5% (vs <3% institutional collapse thesis)
What the essay predicted: Traditional Centro Democrático figures polling at 3-5%, evidence of “institutional collapse”
What markets show: Valencia at 5.5% median (20% high), suggesting she has some consolidation potential
Possible Interpretations:
- A) Institutional Rope not dead, just weakened – Valencia has a base; question is whether she can expand it
- B) Markets pricing primary uncertainty – The 20% high suggests some forecasters believe she could surge post-primary
- C) Guarumo poll (6.9%, leading Gran Consulta) feeding market confidence – Markets reacting to her recent gains
- D) Anti-De la Espriella consolidation beginning – Establishment conservatives rallying to block outsider
Matrix Implication: Valencia’s performance in March 8 primary is now critical diagnostic. If she wins and consolidates >15% nationally, institutional collapse thesis weakens significantly
4. Fajardo at 10% (vs 9-10% predicted “terminal state”)
What the essay predicted: Fajardo stuck at 9-10%, demonstrating centrist strangulation
What markets show: Fajardo at 10% median, perfectly matching prediction
Interpretation: This is the essay’s strongest confirmation. Despite three consecutive presidential runs, despite being a “natural centrist” in a country that “identifies with the center,” Fajardo remains trapped at exactly the predicted ceiling. The Noose is strangling the center as predicted.
Matrix Implication: If Fajardo breaks 15% by Feb 15, the centrist strangulation mechanism is falsified. If he stays at 8-11%, the essay’s constraint logic holds.
Scorecard: Essay Predictions vs Market Reality
| Claim | Predicted | Market Shows | Status |
|---|---|---|---|
| Bipolar compression (1-4 pt spread) | Cepeda vs Right ~35% each | De la Espriella 46%, Cepeda 31.5% (14.5 pt gap) | 🚨 RED FLAG |
| Cepeda hard ceiling ~35% | Stable at 32-36% | At 31.5%, below floor | ⚠️ YELLOW FLAG |
| Institutional collapse (Valencia <3%) | Valencia negligible | Valencia 5.5%, range to 20% | ⚠️ YELLOW FLAG |
| Centrist strangulation (Fajardo ~10%) | Fajardo terminal 9-10% | Fajardo 10% exactly | ✅ GREEN |
| Field consolidation (Others <5%) | Consolidated to main candidates | “Another” at 5% | ✅ GREEN |
Overall Assessment: 1 Red Flag, 2 Yellow Flags, 2 Green Flags
The essay’s strongest prediction (centrist strangulation) holds perfectly. The weakest prediction (unstable bipolar parity) is contradicted—markets show clear De la Espriella dominance, not parity.
What Changed Since January 16?
These markets opened 10 days ago (01/16/26). The key question: Has forecaster opinion shifted in those 10 days?
If markets have moved toward De la Espriella since opening: This suggests either:
- New polling data (Guarumo 01/26?) showing his strength
- Security events reinforcing his message
- Cepeda missteps or Petro approval drop
If markets opened with De la Espriella at 46%: This suggests forecasters always saw him as stronger than the essay’s parity prediction, even before recent data
Action item: Track whether De la Espriella’s 46% is rising, stable, or falling over the next 2 weeks. If rising, the rightward momentum is real. If falling, it was an overreaction to early data.
Updated Falsification Tests (Incorporating Market Data)
The prediction market divergence creates new testable predictions:
Test 1: De la Espriella’s 46% is real momentum vs. market overconfidence
- If real: Polling by Feb 15 will show De la Espriella at 24-28%, Cepeda at 28-32% (tightening but still separated)
- If overconfident: Polling by Feb 15 will show convergence to 32-35% each (essay’s parity prediction)
Test 2: Cepeda’s 31.5% is structural ceiling vs. temporary dip
- If structural: He remains 30-33% through March, can’t break 35% even with campaign efforts
- If temporary: He rebounds to 34-37% by Feb 15 as Left Rope re-consolidates
Test 3: Valencia’s 5.5% is primary momentum vs. terminal ceiling
- If momentum: She rises to 8-10% by Feb 15, wins primary convincingly March 8, consolidates >15% nationally
- If ceiling: She plateaus at 5-7%, primary is fragmented, can’t expand beyond institutional loyalists
Critical Question for Next Update (Feb 15)
Did the 14.5-point gap (De la Espriella 46%, Cepeda 31.5%) narrow or widen?
- Widening → Red flag for essay’s bipolar parity theory
- Stable → Yellow flag, suggests structural right-wing advantage
- Narrowing → Green flag, markets were overreacting to early data
This market snapshot suggests the essay may have underestimated right-wing consolidation speed and overestimated left-wing coalition strength, while correctly predicting centrist strangulation.
Core Claims from the Essay
Claim 1: The Noose (Bipolar Compression)
“The undecided collapse (62% → 5.7%) demonstrates active constraint suppression forcing binary alignment”
Claim 2: Institutional Collapse on the Right
“Traditional conservative infrastructure has decayed; establishment figures cannot consolidate support”
Claim 3: Centrist Strangulation
“The runoff system mathematically extracts viability from non-polar positions (Fajardo terminal at ~9-10%)”
Claim 4: Left Rope Functional but Non-Generative
“Pacto Histórico successfully consolidated but cannot expand beyond ~35% convinced left voters”
Claim 5: Terminal Attractor
“System converges to unstable bipolar equilibrium (Cepeda vs right-wing candidate, separated by 1-4 points)”
Falsification Schedule: Critical Dates & Thresholds
DATE 1: February 15, 2026 (Pre-Primary Snapshot)
Event Context: 3 weeks before Gran Consulta primary
Data Required: National polling from major firms (Invamer, Guarumo, AtlasIntel)
| Metric | Essay Prediction | Falsification Threshold | Interpretation if Threshold Met |
|---|---|---|---|
| Undecided % | Should continue compressing toward <10% | >22% stable or rising | Noose theory fails; voters retain meaningful choice space |
| Fajardo Support | Remains 8-11% range | >15% | Centrist strangulation fails; third-way viable |
| Valencia (Traditional Right) | Plateaus 7-9% | >12% | Institutional collapse overstated; Rope still functional |
| De la Espriella | Remains 16-20% | <10% or >28% | Either: right consolidated elsewhere OR outsider surging beyond volatility |
| Cepeda | Stable 32-36% | >40% | Left Rope is generative, not just consolidating |
Critical Test: If undecided remains >22% AND Fajardo breaks 15%, the compression mechanism is not active.
Scorecard on February 16, 2026
Undecided (7.7%) — Predicted <10%, falsification at >22%. Solidly confirmed. The Noose is active.
Fajardo (7.6%) — Predicted 8-11%, falsification at >15%. Confirmed, actually slightly below your floor. Centrist strangulation is even more severe than modeled. He loses every second-round scenario by 7-10+ points.
Valencia Traditional Right (3.8% first round, 8.7% even with consulta win) — Predicted plateau 7-9%, falsification at >12%. Confirmed. Institutional collapse is real. The article itself draws the Óscar Iván Zuluaga parallel from 2022.
Cepeda (31.4%) — Predicted 32-36%, falsification at >40%. Right in range. Left Rope is consolidating but clearly bounded.
De la Espriella (32.1%) — Predicted 16-20%, falsification at >28%. This is your red flag. He didn’t just cross the threshold — he blew past it by 4+ points. This is the major structural miss.
What this means for the five claims:
Claims 1, 3, and 4 are performing well. Bipolar compression is confirmed, centrist strangulation is confirmed (possibly accelerating), and the left’s ceiling constraint is visible — Cepeda tops out around 34-36% in every second-round scenario while carrying 43.9% rejection, compared to De la Espriella’s 33.6%.
Claim 2 needs revision, not abandonment. The institutional right did collapse — Valencia, Dávila, Pinzón, Oviedo are all in single digits. But the essay modeled this as creating a fragmented vacuum. Instead, the vacuum was filled almost entirely by a single outsider candidate. The institutional Rope decayed, but a new coordination mechanism formed around De la Espriella much faster than your model anticipated. The right didn’t fragment — it leapfrogged its own institutions.
Claim 5 (terminal attractor, 1-4 point separation) is the most interesting. First round shows 0.7 points — tighter than predicted. But the second-round data reveals an asymmetry your model may have underweighted: De la Espriella is the only right-side candidate who beats Cepeda head-to-head (36.8% to 34.6%). Everyone else loses. This means the bipolar equilibrium isn’t symmetric — the right’s viability is concentrated in a single individual while the left’s is concentrated in a single individual, but the right’s consolidation ceiling appears higher (lower rejection, broader regional spread).
On your critical question about the Kalshi gap: The 14.5-point market spread (De la Espriella 46%, Cepeda 31.5%) looks more justified after this data, not less. The polling shows near-parity in first round, but the structural indicators favor De la Espriella: lower rejection floor, sole viable right-side candidate in runoffs, and Cepeda’s CNE exclusion from the Frente por la Vida consulta (which the article notes could suppress left-coalition turnout on March 8). The markets appear to be pricing in the second-round structural advantage, not just the first-round horserace. I’d call this yellow-to-green for market rationality.
Suggested model revision: The essay’s compression mechanism works. What it underspecified was the rate of right-wing re-coordination around non-institutional nodes. In your constraint taxonomy, what happened isn’t that the Rope failed and nothing replaced it — it’s that an outsider built a new Scaffold that rapidly hardened into a functional Rope, bypassing the party infrastructure entirely. De la Espriella’s movement (“Defensores de la Patria”) is doing what Petro’s coalition did in 2022 but from the right.
What to watch for next: The March 8 consulta turnout will be your next critical test. If the Gran Consulta draws less than 3 million votes (AtlasIntel estimates 2.8-3M), that confirms the institutional right is effectively dead and De la Espriella’s parallel structure is the real coordination mechanism. If Cepeda supporters stay home from the Frente por la Vida consulta (potentially under 1 million), that tests whether the left Rope can maintain cohesion when its institutional pathway is disrupted.
DATE 2: March 8, 2026 – Original (Gran Consulta Primary Results)
Event Context: Right-wing primary determines institutional coordination capacity
Data Required: Primary vote totals and immediate post-primary national polling
| Metric | Essay Prediction | Falsification Threshold | Interpretation if Threshold Met |
|---|---|---|---|
| Primary Winner Margin | Close race (<5 points) or fragmented field | Valencia wins by >15 points | Strong institutional coordination; Rope functional |
| Primary Turnout | Low (<2 million votes) showing weak engagement | >3 million votes | Right-wing base energized and organized |
| Primary Winner’s Post-Primary National Poll | Winner consolidates to 18-22% nationally | Winner jumps to >26% | Successful Rope coordination creates momentum |
| Runner-up Dropout Speed | Slow or contested | All losing candidates endorse winner within 48 hours | Disciplined institutional behavior; not collapsed |
| De la Espriella Performance if NOT Primary Winner | Remains in race as independent | Drops out and endorses winner | Institutional pressure still functions |
Critical Test: If Valencia wins >15 point margin + consolidates to >26% nationally + all losers endorse within 48 hours, institutional collapse thesis is falsified.
DATE 2 REVISED: March 8, 2026 (Gran Consulta Primary + Congressional Elections)
Revision Note (Feb 16, 2026): Original thresholds assumed Cepeda would participate in the Frente por la Vida consulta. The CNE blocked his participation in early February. This fundamentally changes what March 8 measures. The left consulta is no longer a test of Cepeda’s coalition strength — it’s a test of whether the left Rope holds cohesion when its institutional pathway is disrupted. Additionally, concurrent congressional elections provide an independent track for testing bipolar compression at the legislative level.
Event Context: Right-wing primary + left consulta (without Cepeda) + congressional elections Data Required: Primary vote totals, consulta turnout, Senate results by party, immediate post-primary national polling
| Metric | Essay Prediction | Falsification Threshold | Interpretation if Threshold Met |
|---|---|---|---|
| Gran Consulta Turnout | Low (2.8-3M per AtlasIntel estimate) | >4 million votes | Right-wing base energized beyond polling; institutional Rope still functional |
| Gran Consulta Winner Margin | Fragmented field, close race | Valencia wins by >15 points with >3M turnout | Institutional coordination capacity intact; collapse thesis weakened |
| Frente por la Vida Turnout (without Cepeda) | Suppressed, potentially <1 million | >2.5 million | Left Rope maintains cohesion independent of Cepeda; coalition deeper than one candidate |
| Frente por la Vida Turnout (without Cepeda) | Suppressed, potentially <1 million | <500,000 | Left Rope is Cepeda-dependent; structurally fragile if he faces legal/institutional obstacles |
| Senate: Pacto Histórico + Centro Democrático Combined Valid Vote Share | >40% combined (bipolar compression at legislative level) | <30% combined | Compression is presidential-race-specific, not structural; centrist parties viable at legislative level |
| Senate: Centrist Parties (Alianza Verde, Liberal, Partido de la U) Combined | Squeezed below 2022 levels | Combined >30% of valid votes | Centrist strangulation is NOT operating at legislative level; Noose is race-specific, not systemic |
| Post-Primary De la Espriella Polling | Stable or rising (28-34%) | Drops below 24% | Gran Consulta winner absorbing his support; institutional right reconsolidating |
| Post-Primary Cepeda Polling | Stable (30-35%) | Drops below 27% | CNE disruption damaged left coalition; Rope fraying |
Critical Tests:
- If Gran Consulta >4M votes AND winner consolidates to >26% nationally within one week, institutional collapse thesis requires significant revision.
- If Frente por la Vida <500,000 votes, the left Rope is revealed as Cepeda-dependent rather than structurally robust — this changes the runoff calculus because it means institutional disruptions (CNE rulings, legal challenges) become existential threats to the left coalition, not just tactical setbacks.
- If Senate results show Pacto Histórico + Centro Democrático >40% combined while centrist parties are squeezed, this independently confirms the Noose is systemic, operating across institutional levels. This would be the strongest structural confirmation available at this date.
New Diagnostic (not in original matrix): Congressional results as independent compression test. The essay’s constraint logic implies bipolar compression should manifest wherever Colombian voters face aggregation choices, not just in the presidential race. If the Senate shows the same pattern — two polar parties dominating, center squeezed — that’s evidence the Noose is a property of the political system, not an artifact of the specific presidential candidate matchup.
DATE 3: April 1, 2026 (Post-Primary Consolidation)
Event Context: 3 weeks after primary; time for right to consolidate or fragment
Data Required: National polling, runoff scenarios, security issue salience
| Metric | Essay Prediction | Falsification Threshold | Interpretation if Threshold Met |
|---|---|---|---|
| Undecided % | Compressed to <8% | >18% | Compression stalled; Noose not tightening |
| Combined Right Support (Primary winner + De la Espriella if still running) | 24-28% total | >35% combined | Right has more structural support than essay estimated |
| Fajardo Support | Terminal at 8-10% | >13% | Centrist space expanded post-primary |
| Cepeda Ceiling | Hard ceiling at 35-37% | >39% | Left expanding beyond base |
| Security Issue Salience (% voters ranking it #1) | Remains >60% | <45% | Issue landscape shifting; security crisis easing |
Critical Test: If undecided >18% AND Fajardo >13% AND combined right >35%, the Terminal Attractor prediction fails.
DATE 4: May 15, 2026 (Two Weeks Before First Round)
Event Context: Final consolidation before May 31 first round vote
Data Required: Final polling averages, runoff matchup scenarios, voter certainty metrics
| Metric | Essay Prediction | Falsification Threshold | Interpretation if Threshold Met |
|---|---|---|---|
| Top Two Candidates Margin | Within 1-4 percentage points | Separation >8 points | Not converging to bipolar equilibrium; clear frontrunner |
| Undecided % | Compressed to <5% | >12% | Noose failed to extract choice; voters retain flexibility |
| Fajardo + Other Non-Polar Candidates | Combined <15% | Combined >22% | Non-polar space survived; not strangled |
| Cepeda vs Right-Winner Runoff Poll | Within margin of error (±3 points) | Either candidate leading by >7 points | Not unstable equilibrium; clear favorite |
| Voter Certainty (“Definitely voting for X”) | <70% for both frontrunners | >82% for frontrunner | High certainty = stable preference, not forced alignment |
Critical Test: If clear frontrunner (>8 point lead) + low certainty (<70%) exists, it suggests preference formation, not constraint extraction.
DATE 5: May 31, 2026 (First Round Results)
Event Context: Actual vote totals vs. final polls
Data Required: Electoral authority official results, geographic/demographic breakdowns
| Metric | Essay Prediction | Falsification Threshold | Interpretation if Threshold Met |
|---|---|---|---|
| Polling Error Magnitude | Standard ±2-3 points | Systematic error >6 points in one direction | Polls missed hidden dynamics; constraint model incomplete |
| Fajardo Performance vs Polls | Within ±2 points of final polls | Outperforms polls by >5 points | Late-breaking centrist consolidation; Noose loosened |
| Top Two Separation | 1-4 percentage points | Separation >7 points OR <1 point | Either: not bipolar OR hyper-compressed beyond prediction |
| Third Place Candidate % | <12% | >18% | Non-polar space survived to election day |
| Geographic Concentration | Cepeda wins cities, Right wins rural (polarized map) | Mixed geographic pattern | Not pure polarization; issue-based voting persists |
Critical Test: If Fajardo outperforms by >5 points OR third place gets >18%, the centrist strangulation mechanism failed.
DATE 6: June 21, 2026 (Runoff Day)
Event Context: Final binary choice test
Data Required: Runoff results, Fajardo voter behavior, turnout patterns
| Metric | Essay Prediction | Falsification Threshold | Interpretation if Threshold Met |
|---|---|---|---|
| Winning Margin | 50-52% (narrow victory per “unstable equilibrium”) | Victor wins with >55% | Stable coalition built; not unstable |
| Fajardo Voter Behavior (if he’s eliminated) | ~55-60% break for one candidate, ~40-45% for other | >75% break to one candidate | Clear centrist preference exists; not forced binary |
| Turnout vs First Round | Similar or slightly higher | >8% higher turnout | Runoff energized voters; not exhaustion from forced choice |
| Geographic Polarization Index | High (Gini coefficient >.45 for vote distribution) | Low (<.35) | Not pure polarization; diverse coalition |
| Post-Election Approval for Winner (1 week after) | Immediate polarization (approval ≈ vote share) | Honeymoon effect (approval >vote share +5 points) | Voters accepting outcome; not viewing as existential |
Critical Test: If winner gets >55% + Fajardo voters break >75% to one side + geographic diversity, the unstable bipolar equilibrium prediction is falsified.
Compound Falsification Conditions
STRONG FALSIFICATION (Theory Fundamentally Wrong)
Requires THREE OR MORE of the following:
- Undecided remains >20% through April 1
- Fajardo exceeds 15% in any poll after March 8
- Primary winner consolidates right to >35% combined support by April 1
- First round has clear winner by >8 points
- Runoff winner exceeds 55% with diverse geographic coalition
- Cepeda breaks 40% in national polling before May 15
Interpretation: The constraint topology is not determinative; traditional electoral dynamics (campaigning, issues, coalitions) dominate outcomes.
PARTIAL FALSIFICATION (Theory Incomplete)
Requires TWO of the following:
- Valencia consolidates institutional right (wins primary >15 points + reaches >26% nationally)
- Security issue salience drops below 45% of voters by April 1
- Undecided stabilizes at 15-18% rather than compressing to <8%
- Geographic voting patterns show issue-based diversity rather than pure polarization
Interpretation: Constraints exist but interact with variables the model didn’t weight sufficiently (candidate quality, issue dynamics, external shocks).
CONFIRMATION (Theory Holds)
Requires ALL of the following:
- Undecided compresses to <8% by April 1, <5% by May 15
- Fajardo remains in 8-11% range through May 31
- First round produces top two within 1-4 points
- Runoff winner margins between 50-53%
- Post-election approval ≈ vote share (no honeymoon; immediate polarization)
- Cepeda never exceeds 38% in polling (hard ceiling holds)
Interpretation: The constraint system is dominant; Mountains (term limit) + Nooses (runoff math) + Rope failures (institutional decay) produce the predicted bipolar compression.
Data Collection Protocol
Required Data Sources (Monitor These):
- Polling firms: Invamer, Guarumo/EcoAnalítica, AtlasIntel, Cifras y Conceptos, CNC
- Security metrics: Government territorial control data, homicide rates, forced displacement statistics
- Primary results: Registraduría Nacional official vote counts
- Runoff scenarios: Track all major polls’ head-to-head matchups
- Issue salience: Regular tracking of “most important issue” polling
Key Variables to Track Weekly (Feb 1 – May 31):
- Undecided %
- Fajardo %
- Cepeda %
- Right-wing candidate(s) % [Valencia, De la Espriella, or consolidated winner]
- Security issue salience
- Petro approval rating
- Voter certainty levels
- Runoff scenario polling (all combinations)
Geographic Indicators (Monitor After March 8):
- Urban/Rural split in candidate support
- Regional concentration (Antioquia, Bogotá, Caribbean coast, Pacific coast, eastern regions)
- Gini coefficient for geographic vote distribution
Using This Matrix
Green Flag (Theory Confirmed): Data consistently meets Essay Predictions column
Yellow Flag (Theory Under Stress): Data approaches but doesn’t exceed Falsification Thresholds
Red Flag (Theory Falsified): Data meets Falsification Thresholds on multiple metrics
Track cumulative Red Flags:
- 0-1 Red Flags: Theory robust
- 2-3 Red Flags: Theory needs revision
- 4+ Red Flags: Theory fundamentally wrong
The advantage of this matrix is it forces specific, dated, quantitative predictions that can’t be retrofitted. The essay makes bold structural claims—this matrix makes them testable.
cafebedouin.org 