QualQuant Signals :: Delphi Focus Groups™ | NY Senate - Campaign Success Framework + Predictions

🗽 New York Senate Race · Campaign Success Framework + Predictions

Example : Based on 200 voters participating in Delphi Focus Group. Hybrid QualQuant Survey

Where You Stand March 2026

Alexandra Chen (YOU)

James Morrison (Opponent)

YOU'RE AHEAD BY +3 POINTS

📌 WHAT THIS MEANS: Your Candidate Potential Score is 71. Your opponent is 68. You're winning — but it's close. 3 points is 15,000 voters. The predictions below will tell you exactly how to grow that lead.

Your Biggest Advantage 🔥 Gold Mine Found

Message #1 VCS: 89

"I'm a diabetic and my insulin costs have tripled. I'm choosing between food and medicine."

47% of voters agree 9.2/10 passion

82% say it's urgent 78% will vote for you

🎯 WHY THIS MATTERS: This one message could win you 19,000 votes. Lead with it.

🎯

Your Campaign Success Framework

Do this → Capture these voters → Here's why it works

"I'm a diabetic and my insulin costs have tripled. I'm choosing between food and medicine."

Reach (how many) 47%

Passion (how strongly) 9.2/10

Urgency (unmet need) 82%

Conviction (will vote) 78%

VOTER CAPTURE SCORE: 89 (GOLD MINE)

DO THIS: Run as PRIMARY message on TV + YouTube

GET THIS: +3.8% votes (19,000 voters)

WHY IT WORKS: 47% reach = 47,000 voters care. 78% conviction = they'll vote for you.

✓ STATISTICAL CONFIDENCE: 92% (based on 200 voters, margin ±0.4%)

"I'm self-employed and my family health insurance is $1,800/month. It's killing my business."

Reach 38%

Passion 9.4/10

Urgency 76%

Conviction 71%

VOTER CAPTURE SCORE: 84 (GOLD MINE)

DO THIS: Target small business owners on radio + digital

GET THIS: +2.9% votes (14,500 voters)

WHY IT WORKS: 9.4 passion means this group will TURNOUT for you.

✓ STATISTICAL CONFIDENCE: 89%

"My daughter's school can't afford books. I'm spending grocery money on supplies."

Reach 42%

Passion 8.4/10

Urgency 72%

Conviction 68%

VCS: 76 (QUICK WIN)

DO THIS: Mention in education-focused events + mail

GET THIS: +2.4% votes (12,000 voters)

WHY IT WORKS: 42% reach = broad appeal, but lower passion.

✓ STATISTICAL CONFIDENCE: 86%

🔮

The Prediction Engine

4 Delphi Algos that enhances your win chances with 92% confidence (explained simply)

👋 Hi! Let's learn some data science together. It's not scary. It's just finding patterns in what voters told us. Think of it like this:

📊 200 voters

🔍 Find patterns

🎯 Predict votes

📊

Prediction #1: What Drives Your Vote

ALGORITHM: Multiple Linear Regression

🧪 HOW IT WORKS: Regression finds the relationship between things. Like: "If it rains more, do more people bring umbrellas?" Here, we find: "If voters trust you more, do they vote for you more?"

Your Image (CPS) 40% impact

Message #1 (Healthcare) 35% impact

Message #2 (Small Business) 18% impact

Party ID 15% impact

🧠 KINDERGARTEN: Your image matters 40%. Your top message matters 35%. That means if you fix your message, you can move 35% of the vote. In a close race, that's everything.

R² = 0.78 (78% accurate) · p < 0.001 · 200 voters

🔄

Prediction #2: Who Will Switch

ALGORITHM: Logistic Regression

🧪 HOW IT WORKS: Logistic regression predicts YES/NO questions. Like: "Will this voter switch to you?" It learns from patterns of voters who already switched.

Undecided voters switching TO you +12%

Opposition voters switching FROM opponent +8%

🧠 KINDERGARTEN: 20% of voters are undecided. Our algorithm predicts 12% of them will come to you if you run Message #1. Another 8% of opponent voters are ready to leave him. Target them with Message #2.

Why it's accurate: The algorithm found that voters who selected Message #1 were 3.2x more likely to vote for you.

🌲

Prediction #3: Message Performance by Voter Type

ALGORITHM: Random Forest Classification

🧪 HOW IT WORKS: Random Forest is like asking 1000 tiny trees to vote. Each tree looks at different voter traits (age, party, issues) and predicts which message they'll like. Then all trees vote.

SEGMENT	MSG 1	MSG 2	MSG 3
Loyal	12%	8%	15%
Undecided	68%	22%	10%
Opposition (Available)	18%	52%	30%
Not Available	5%	8%	12%

🧠 KINDERGARTEN: Message #1 works BEST on undecided voters (68% will respond). Message #2 works BEST on opposition voters ready to switch (52% will respond). Put the right message in front of the right people.

Algorithm Accuracy: 89% cross-validated · 200 decision trees

💰

Prediction #4: Optimal Budget Allocation

ALGORITHM: Linear Programming Optimization

🧪 HOW IT WORKS: This algorithm solves a puzzle: "Given $1 million, how should I spend it to get the MOST votes?" It weighs each message's impact and finds the perfect mix.

Message #1 (VCS 89) 45%

Message #2 (VCS 84) 30%

Message #3 (VCS 76) 15%

Other / GOTV 10%

🧠 KINDERGARTEN: Put 45% of your money on Message #1. It gives you the most votes per dollar. 30% on Message #2. 15% on Message #3. This mix maximizes your vote gain.

Optimization Result: This mix yields 9.1% total vote lift

🧩 How These Algorithms Work Together (The Magic)

Regression

Finds WHAT matters

Logistic

Finds WHO switches

Random Forest

Finds WHAT works for WHO

Optimization

Finds WHERE to spend

Together, they create a complete picture: What to say, to whom, and how much to spend.

9.1%

Predicted Vote Gain

92%

Confidence Level

45.5K

Voters Gained

🔮 THE BOTTOM LINE: If you follow this playbook, you have a 92% chance of gaining 9.1% more votes. In a close race, that's the difference between losing and winning.

📊 Regression

Found that your image + Message #1 explain 78% of why people vote for you.

🔄 Logistic

Identified 12% of undecided voters who will switch to you.

🌲 Random Forest

Matched each message to the right voter segment with 89% accuracy.

💰 Optimization

Calculated the exact budget mix to maximize your votes.

🗺️ The Voter Battlefield Map

Different voters need different messages. Here's exactly who to target and what to say.

✓ LOYAL

30% of voters (60,000 people)

They already love you. Don't waste money persuading them. Just get them to vote.

What to say: "Join us. Be part of the win. Bring a friend."

Email Text Door knocking

🟡 UNDECIDED

20% of voters (40,000 people)

These decide elections. They need your best message: Healthcare.

What to say: Message #1: "I'll cap your healthcare at $50."

TV YouTube Local News

🟠 OPPOSITION (AVAILABLE)

15% of voters (30,000 people)

They lean opponent but have doubts. Exploit his weaknesses.

What to say: "Even his own party disagrees with him on taxes."

Facebook Mail Radio

🔴 NOT AVAILABLE

35% of voters (70,000 people)

They hate you. Don't waste money. Just plant seeds.

What to say: "Did you know he voted against school funding 3 times?"

Podcasts Streaming

📈 What You'll Gain (The Bottom Line)

Run these three messages → Get these votes. Predicted by our algorithms.

3.8%

from Message #1

19,000 voters

2.9%

from Message #2

14,500 voters

2.4%

from Message #3

12,000 voters

TOTAL PROJECTED LIFT

Based on 4 algorithms · 92% confidence

9.1% (45,500 votes)

✅ KINDERGARTEN SUMMARY: Our regression algorithm found that your image + messages explain 78% of votes. Our random forest matched messages to segments. Our optimizer found the perfect budget mix. The result: 9.1% more votes. In a close race, that's the win.

📋 YOUR WINNING PLAYBOOK (One Page)

Follow this exactly. The algorithms have spoken.

🎯 DO THIS:

• Run Message #1 on TV + YouTube (45% of budget)
• Target small business owners with Message #2 (30%)
• Use Message #3 in education mailers (15%)
• Mobilize loyal voters with texts/email (10%)

✅ GET THIS:

• +3.8% from Message #1 (19,000 votes)
• +2.9% from Message #2 (14,500 votes)
• +2.4% from Message #3 (12,000 votes)
• TOTAL: +9.1% (45,500 votes)

🔮 WHY THIS WORKS (The Algorithms): Regression (78% accuracy) + Logistic (92% confidence) + Random Forest (89% accuracy) + Optimization (perfect mix). This isn't guessing. It's science.

Want This For Your Campaign?

One Delphi Focus Group. 100+ voters. 24 hours. Exact messages + 4 algorithms + projected vote lift.

No obligation. Just a 15-minute demo showing what our algorithms would find for YOUR race.