The PacketStorm CFB Ranking System is a transparent, interactive résumé-style tool that ranks teams by accumulating points from game results, where the points depend on the opponent’s tier (derived from a user-supplied Starting Order) and the margin of victory or defeat. It remains a flexible strength-of-schedule and results system rather than a pure predictive power rating.
Evaluation of the PacketStorm System
The core logic is straightforward and sensible:
- You supply a Starting Order (seeds). Lower seed = stronger team. These seeds determine each opponent’s tier.
- An editable Point Matrix assigns points for wins and losses according to the opponent’s tier and margin of victory/defeat. Higher-tier opponents produce larger rewards for wins and smaller penalties (or even small positive credit) for losses.
- Tiers are defined by maximum seed cutoffs. The system uses five tiers. The matrix is fully customizable.
- Game data includes score and margin; the system can (and should) use those margins when awarding points.
- Final ranking is the sum of points earned across included games. You can include/exclude individual games, add games manually, and export/import settings.
Strengths
Explicitly rewards quality of competition and performance against it. Highly transparent and interactive. Incorporates game margins so a dominant win is valued differently from a narrow one. Matches the direction of recent College Football Playoff adjustments that emphasize record strength and quality wins while reducing the damage of competitive losses to strong teams.
Limitations
Completely dependent on the quality of the Starting Order—a biased or outdated seed list propagates through every tier and point calculation. Discrete tiers can still create cliff effects (beating the last team in one tier versus the first team in the next). Point values remain user-chosen and require careful calibration against historical outcomes. Home/away adjustments are not automatic and must be added manually if desired.
Best Practices for Starting Order, Tiers, and Point Values
1. Setting the Starting Order (Seeds)
The Starting Order is the foundation. Begin with a consensus of strong predictive systems (SP+, FPI, FEI, or calibrated Elo). Blend in recent results and roster continuity. Update gradually rather than creating a pure feedback loop from the previous week’s PacketStorm output. Keep the order continuous and realistic.
2. Setting Tiers
The system uses five tiers. Recommended cutoffs for 136–138 FBS teams:
- Tier 1 (Elite): Seeds 1–15
- Tier 2 (High Contender): 16–35
- Tier 3 (Solid / Average Power): 36–70
- Tier 4 (Below Average): 71–105
- Tier 5 (Weak / Lower G6 & FCS): 106+
Adjust the Max seed values so each tier remains meaningful and relatively stable in size. Treat road games against the same seed as roughly one half-tier harder if desired.
3. Setting Point Values with Margin of Victory
The Point Matrix should satisfy these principles:
- Beating a stronger opponent is worth substantially more than beating a weaker one.
- Losing to a stronger opponent is only mildly damaging (or slightly positive if competitive).
- Losing to a weaker opponent is heavily penalized.
- Narrow wins earn less than dominant wins; narrow losses cost less than blowout losses.
- Extreme margins are capped so garbage-time scoring does not distort rankings.
Recommended Point Matrix (Win / Loss by Margin Band) — Five Tiers
| Opponent Tier | Win 1–8 pts | Win 9–16 pts | Win >16 pts | Loss 1–8 pts | Loss 9–16 pts | Loss >16 pts |
| Tier 1 (Elite, 1–15) | +14 | +18 | +22 | +2 | –1 | –4 |
| Tier 2 (High Contender) | +10 | +13 | +16 | –1 | –3 | –6 |
| Tier 3 (Solid / Avg Power) | +6 | +8 | +10 | –5 | –7 | –10 |
| Tier 4 (Below Average) | +3 | +4 | +5 | –9 | –12 | –15 |
| Tier 5 (Weak) | +1 | +1.5 | +2 | –13 | –16 | –20 |
Design rationale
A dominant win (>16) over a Tier 1 team is worth roughly 10–20× a narrow win over a Tier 5 team. A close loss (1–8) to a Tier 1 opponent can still award a small positive or near-zero score, reflecting the “coupon for playing a tough schedule.” A blowout loss (>16) to a Tier 5 opponent carries a severe penalty. The three margin bands provide meaningful differentiation while the system’s built-in margin data makes these distinctions practical to implement.
Calibration tips
Run the matrix on prior seasons and compare the resulting rankings against actual playoff performance, betting markets, and predictive systems such as SP+ or FPI. Avoid making any single game worth more than roughly 15–20 % of a typical top team’s season total. Optional refinements include modest home/away multipliers (±10–15 %), a slight late-season weighting, or a soft cap that reduces additional points beyond a 28- or 35-point margin.
Recommended Workflow
- Build or import a high-quality Starting Order from predictive systems plus current form.
- Define the five tiers with clear Max seed cutoffs.
- Populate the Point Matrix using the three margin bands shown above (or a tuned variant), taking advantage of the margin data already present in the Games tab.
- Score every game, review outliers, and recalculate.
- Treat the output as a strong résumé + strength-of-schedule ranking and cross-check it against pure predictive ratings for a complete picture.
- Export settings so the same philosophy can be applied consistently week to week.
This margin-aware, five-tier version of the PacketStorm framework better distinguishes dominant performances from close ones while still heavily rewarding quality of competition. When the Starting Order is disciplined and the Point Matrix is calibrated, the rankings become both fair and informative.

