How to Make NFL Power Ratings: A Simple Points-Based System

You make NFL power ratings by assigning every team a number based on how many points better or worse than average they are, then adding home-field advantage to project a spread. That's the whole concept — the rest is just deciding how to calculate the number honestly.

Here's the part most guides skip: your ratings aren't meant to beat the market. They're meant to give you a number of your own so that when the Kansas City Chiefs are -7.5 and your sheet says -4, you have a specific disagreement to investigate instead of a vague gut feeling.

What a Power Rating Actually Is

A power rating is a single number representing a team's strength, expressed in points. If the Philadelphia Eagles carry a rating of +6 and the league-average team sits at 0, the Eagles should beat an average opponent by about six points on a neutral field.

That's it. No magic. Every sportsbook and every serious bettor maintains some version of this — the only difference is sophistication. Yours can be built in a spreadsheet in an afternoon.

The Two Ingredients

1. Team rating — points above or below average.
2. Home-field advantage — historically around 2 to 2.5 points in the NFL, lower than it was a decade ago.

Projected spread = (Home team rating − Away team rating) + home-field advantage.

Step 1: Start With Point Differential

The simplest honest foundation is adjusted point differential. Take each team's average points scored minus points allowed per game over a meaningful sample — say, the last 8 to 10 games.

If the Detroit Lions have outscored opponents by 9.4 points per game over that stretch, their raw rating is +9.4. A team getting outscored by 5 per game sits at −5.

Adjust for Opponents (Lightly)

Raw differential overvalues teams with soft schedules. A basic fix: for each game, compare the opponent's scoring margin that game to their season average. If the Lions beat a team by 14 that usually loses by 3, that performance is worth +11 of credit, not +14. Average those adjusted margins across the sample.

Don't over-engineer this. A rough strength-of-schedule adjustment beats none, and a perfect one doesn't exist.

Step 2: Recency Weighting Without Overreacting

More recent games should count more — teams change. Injuries, scheme adjustments, a rookie quarterback settling in. A simple weighting: multiply each game's adjusted margin by its recency rank (most recent game ×10, then ×9, down to ×1), then divide by the sum of weights (55 for a 10-game sample).

One warning: the biggest mistake new ratings-builders make is letting last Sunday dominate the sheet. One blowout shouldn't swing a rating by three points. If it does, your weights are too aggressive.

Step 3: Worked Example

Say your sheet produces these ratings:

• Kansas City Chiefs: +7.0
• Philadelphia Eagles: +5.5
• Home-field advantage: 2.0

Eagles hosting the Chiefs: (5.5 − 7.0) + 2.0 = Chiefs by 0.5. Your number says pick'em, essentially.

Now the market opens Chiefs -3. That's a 2.5-point gap between your number and the market's. This is the entire point of the exercise — not "my number is right," but "why does the market see three more points of value on Kansas City than I do?"

Possible answers: a Philly injury your rating hasn't absorbed yet, a matchup dynamic (pass rush vs. a shaky offensive line) that raw margins don't capture, or your sample is stale. Sometimes the answer is that the market is shading toward a public team, and your disagreement is real. You only get to ask the question because you built the number first.

Step 4: Set a Threshold for Action

Disagreement alone isn't a bet. Small gaps — a point or two — are noise, because your ratings carry that much error by themselves. Many bettors won't act unless their number differs from the spread by 3 or more points, and even then only after checking injuries, weather, and line movement.

Track every projection against the closing line and the actual result. After 50+ games, you'll know whether your ratings add information or just echo the market with extra steps. If your number consistently lands closer to the final margin than the spread does in a specific spot — divisional underdogs, say — that's a real edge worth pressing.

Where Most Homemade Ratings Go Wrong

• Overfitting to recent results. One Monday night blowout moves your rating more than it should.
• Ignoring quarterback changes. A backup QB can be worth 4 to 7 points. Your differential-based rating won't know that until several games pass — adjust manually.
• Treating the output as gospel. The market pools millions of dollars of information. Your spreadsheet is a sanity check, not a rival.
• Never updating home-field advantage. It's drifted downward over the years; check it against recent seasons.

How Da Vinci Bets Uses the Same Idea at Scale

What you've built by hand is a stripped-down version of what the Da Vinci Bets model does with far more inputs. Instead of one adjusted point differential, the model weighs efficiency metrics, situational splits, pace, and personnel data to produce its own projected number for every game — then compares that number to the market, exactly like your spreadsheet does.

The philosophy is identical: the value isn't in the rating itself, it's in the disagreement. When the model's projection and the posted spread diverge meaningfully, that's a flagged spot worth examining — not an automatic bet, and never a guaranteed winner. Your homemade ratings teach you the discipline; the model just does the arithmetic on more variables than a spreadsheet can hold.

The Real Payoff

Building NFL power ratings won't make you sharper than the sportsbooks overnight. What it does is replace "I like the Lions here" with "my number says Lions -2, the market says -5.5, and here's what might explain the gap." That shift — from opinion to structured disagreement — is what separates bettors who improve from bettors who just bet.

Unlock all picks, signals, and deep analysis. Go Premium for full access to our AI prediction engine.