📈 Player strength, roster strength, and the record
20-0 game scoring starts with evaluated player-seasons, then accounts for the positions they occupy. A finished roster is more than a collection of the largest visible football statistics: the game uses its own player evaluation and a non-linear relationship between roster strength and the projected record.
The published full-roster weights add to 12.9. That total is useful for explaining a weighted average. It does not reveal every part of the game’s evaluation or give us a license to invent the exact win curve.
Keep the levels separate. A player has an evaluated contribution. The roster combines contributions. The result gives you a record and its associated labels. Confusing those levels is how “one extra point” accidentally becomes “one extra win.”
Where the extra position weight sits
Quarterback has the largest listed multiplier. Edge rusher and cornerback also count above the standard slot weight. The other nine roster positions count once each, so the total is 1.5 + 1.2 + 1.2 + 9 = 12.9.
For a practice QB value of 88, the weighted term is 88 × 1.5 = 132 before normalization. That is a contribution to a sum, not a new player rating of 132. Dividing the combined terms by the total slot weight keeps an all-88 practice roster at 88.
The actual QB destination visibly carries its premium indicator during the draft. For the legal roles behind these weights, see the full position map.
🧮 A transparent weighted-average example
For this arithmetic exercise, suppose all twelve players have an illustrative evaluation of 88. These are invented practice values attached to no real players. Their weighted mean is also 88, because every term starts at the same value.
Improve the QB by four
The weighted total grows by 4 × 1.5 = 6. Divide that change by 12.9 and the mean rises by about 0.465, from 88 to 88.465.
+0.465Illustrative change in the weighted mean.
Improve a standard slot by four
The weighted total grows by 4 × 1.0 = 4. The mean rises by about 0.310, from 88 to 88.310. The same player-value increase has a smaller weighted effect.
+0.310The same increase under a standard slot weight.
Same four-point gain, different weighted effect
This example explains the published weighting. It does not reproduce the complete game engine, account for every possible setting, or predict an exact record from a made-up roster.
Change the slot and inspect the arithmetic
Start with an otherwise equal practice roster. Change one player’s evaluation and see how the slot weight affects the mean. The calculation is deliberately limited to that one question.
Why a weak slot is easy to underestimate
Stay with the same educational roster at 88 throughout. Replace one standard-slot value of 88 with 60. The weighted total loses 28 points, and the mean falls by 28 ÷ 12.9, or about 2.171. It becomes approximately 85.829.
Now compare that with improving one standard slot from 88 to 99. That adds eleven weighted points, or about 0.853 to the mean. In this example, one large shortfall outweighs one modest improvement elsewhere. No invented claim about a real player is required to see the arithmetic.
A practical reading of that example
When reviewing a roster, inspect the selection you settled for before celebrating its most famous name. The size of a gap matters alongside the importance of the slot. A premium-position improvement can be valuable without making the other eleven selections irrelevant.
Do not confuse an unresolved position with a confirmed low rating, either. The game still needs a legal lineup. The draft rules explain completion and placement before any scoring interpretation begins.
🏁 Read the completed result in its own terms
In the captured run, the game reported 17–3, grade A−, and 1,168 points, with offense and defense both shown at 94. Those fields belong to different scales. A unit rating of 94 is not a 94% chance to win every match.
Record and points are two outputs of the same underlying roster evaluation, rather than independent votes on whether the lineup was good. The win projection is non-linear, so a point increase need not produce another displayed win. The one-QB comparison shows that happening in an observed pair.
The captured result is one completed run used to show the screen. It is not a claim about the typical player, a guaranteed strategy, or the chance that your next attempt will end the same way.
Record
The number of wins and losses assigned to the completed run.
Grade
A result label that helps describe where that record sits.
Points
The game’s displayed score. Compare it within the applicable mode and settings.
The published full-roster grade ladder
| Grade | Wins | How to read it |
|---|---|---|
| S+ | 20 | The perfect target: no losses. |
| A+ | 19 | One win short of perfection. |
| A | 18 | A strong result with two losses. |
| A− | 17 | Seventeen wins across the full run. |
| B+ | 16 | Four losses remain to examine. |
| B | 15 | Three quarters of the twenty-game target won. |
| B− | 14 | A winning record with room to improve. |
| C+ | 13 | Thirteen wins, seven losses. |
| C | 12 | Twelve wins across the completed season. |
| C− | 11 | One win above an even record. |
| D | 9–10 | Inspect the largest roster compromises. |
| F | 0–8 | Revisit card reading and legal roster construction. |
Review the decision you can explain
If you want to review the arithmetic, keep the inputs explicit: the value change, its destination weight and the 12.9 denominator. If you want to review the draft, keep the actual alternatives explicit instead. A worked mean and a real roster decision are useful for different reasons.
For the distinction between a season stat and an overall evaluation, read player ratings and era context. For the columns on the cards themselves, use the football-stat glossary. Then return to the 20-0 game with a clearer question than “why did the biggest names not win everything?”