SciWard vs Quadratic Funding — core advantage ⚙️

Quadratic funding (QF) is excellent for one narrow problem:

“Which public goods have many supporters, and how should a matching pool amplify small donations?”

SciWard (SciWard) is broader:

“Which scientific/software contributions are objectively valuable, dependent on what, reusable by whom, and how should rewards flow through the contribution graph?”

So SciWard is not just “better QF.” It solves a different, larger allocation problem.

---

1. QF measures popularity; SciWard can measure merit

Quadratic funding rewards projects with many distinct contributors. This is useful, but it can confuse popularity with importance.

Example:

Case	QF likely outcome	SciWard likely outcome
Popular educational app	Strong funding	Strong funding if useful
Deep theorem used by few experts	Weak funding	Potentially strong funding if foundational
Invisible infrastructure library	Often underfunded	Can be rewarded through dependency impact
Flashy project with many fans	Over-rewarded risk	Checked against actual contribution value

For science, this matters a lot. Many major discoveries are not immediately popular.

---

2. SciWard can reward dependencies, not only final projects

QF usually funds the visible project receiving donations.

SciWard can fund the whole dependency chain:

Mathematical theory
   ↓
Algorithm
   ↓
Open-source library
   ↓
Scientific tool
   ↓
Published result
   ↓
Social/economic benefit

In SciWard, the upstream contributors can receive credit because the system can model causal and dependency relations.

That is a major advantage over QF.

---

3. SciWard is better for expert domains

QF works best when ordinary donors can judge value.

But in advanced science, donors usually cannot reliably evaluate:

• whether a theorem is correct;
• whether a proof technique is novel;
• whether a software library is foundational;
• whether a paper is derivative or original;
• whether a contribution is useful but obscure.

SciWard can use AI-assisted review, expert signals, citation/dependency graphs, reproducibility checks, and post-publication evaluation.

So SciWard is more suitable for fields where truth and utility are hard to judge by popularity.

---

4. QF is vulnerable to sybil games; SciWard can use richer anti-gaming signals

QF has a known weakness: if matching funds depend on the number of donors, fake or coordinated identities can distort allocation.

SciWard can still be gamed, but it has more possible defense layers:

Attack	QF vulnerability	SciWard defense direction
Fake donors	High risk	identity, reputation, graph analysis
Popularity campaigns	High risk	merit-weighted review
Low-quality viral projects	High risk	outcome/dependency evaluation
Citation rings	Possible	graph anomaly detection
Expert capture	Possible	transparent competing evaluations

SciWard’s advantage is that it does not rely on one scalar signal like number of donors.

---

5. SciWard can be post-publication and dynamic

QF usually funds based on current donor support.

SciWard can update rewards as evidence changes:

Contribution made → reviewed → reused → cited → depended on → replicated → rewarded more

This is closer to how science actually works. A discovery may look minor at first and become foundational later.

QF is more like a funding snapshot.
SciWard can be a living reputation-and-reward system.

---

6. SciWard can include negative evaluation

QF mainly asks: “Who supports this?”

SciWard can also ask:

• Is the work wrong?
• Is it plagiarized?
• Is it non-reproducible?
• Is it redundant?
• Is it harmful hype?
• Is the dependency claim fake?

That matters in science funding. Popular but false work should not be amplified.

---

7. SciWard is better for non-donor value

QF depends on people donating.

But many valuable scientific contributions help people who never donate:

• poor students;
• future researchers;
• open-source users;
• governments;
• hospitals;
• anonymous downstream users;
• AI systems using the knowledge.

SciWard can estimate value from usage, dependencies, citations, expert review, and downstream impact, not only donations.

---

Summary table

Dimension	Quadratic Funding	SciWard
Main signal	Number and size of donations	Merit, dependencies, impact, review
Best for	Popular public goods	Science, FOSS, deep research
Rewards hidden foundations	Weakly	Strongly
Handles expert knowledge	Poorly	Potentially well
Updates over time	Limited	Natural
Sybil resistance	Difficult	More signal layers possible
Measures popularity	Strong	Optional
Measures objective contribution	Weak	Central goal
Suitable for obscure but important science	Often no	Yes