2023 season

4-team CFP era

Latest update: 2026-10-18 5:01 am ET  |  Week 15 status: COMPLETE

20262025202420232022202120202019201820172016201520142013201220112010200920082007200620052004200320022001

2023 Harbin Rankings

FBS (national)All teamsPostseason scoresWhat if? (weighting sliders)

Power 4 conferences

ACCBig 12Big TenSECPac-12

Other FBS conferences

Conference USAMACMountain WestSun BeltAmerican (AAC)

Independents

FBS Independents

2023 season summary

Harbin's top four matches the real one exactly.

Among games between teams the committee ranked, Harbin put the loser above the winner in 5 of 38 (the committee: 4), and picked 3 of 11 postseason games (the committee: 5), behind the committee overall.

It flags 2 teams that look out of place: Toledo (odd-in) and Troy (odd-in).

Ineligible for the postseason, so ranked but left out of Harbin's field: James Madison (#15).

MeasureHarbinCFP committee (2023-12-03)
Results contradicted (fewer is better)5 of 38 (13%)4 of 38 (11%)
Postseason picks right3 of 11 (27%)5 of 11 (45%)
Playoff field matched4 of 4 (100%)n/a

Not eligible for the postseason

These teams rank in Harbin's top 25 but could not play in the postseason. They stay in the rankings, shaded pink, and are skipped for every field and cut line.

TeamHarbinW-LWhy
James Madison#1511-1FBS transition: James Madison (11-1) was in year two of the FCS-to-FBS transition and not eligible for the postseason rankings; only conditionally bowl-eligible and the NCAA denied its waiver.

Outliers

TeamKindHarbinCFP committeeAPFPIW-LWhy
ToledoHarbin has in, reference does not#21--#6111-2
  • Against Tulane (Harbin #26, the first spot outside the committee's top 25), Toledo beat 1 team with 9+ wins (Miami OH 11-2), so its beaten opponents earned 276.0 Level 1 points in all (258.0 for Tulane's), a Level 2 term of 19.2 to 18.2; Tulane beat 6 teams with losing records (East Carolina 2-10, Southern Miss 3-9, FAU 4-8, Tulsa 4-8 and 2 more) and FCS Nicholls, which add little to Level 2. This accounts for 1.26 points, more than the whole 0.93-point rating gap; other factors pull the other way.
  • Toledo's loss to 5-7 Illinois counts no more in Harbin than a loss to a top-10 team; it is unranked in the committee ranking.
TroyHarbin has in, reference does not#23--#4511-2
  • Tulane (Harbin #26, the first spot outside the committee's top 25) lost the AAC title game to SMU (11-2): it was its 13th game and earned 0 points, so its Level 1 average fell from 4.42 to 4.08 and SMU's 13 games went into its divisor with nothing earned; without that game Tulane would rate 22.59 instead of 20.72. This accounts for 1.86 points, more than the whole 0.66-point rating gap; other factors pull the other way.
  • Troy beat no team with 9+ wins, so its beaten opponents earned 270.0 Level 1 points in all (258.0 for Tulane's), a Level 2 term of 17.3 to 18.2; Tulane beat 6 teams with losing records (East Carolina 2-10, Southern Miss 3-9, FAU 4-8, Tulsa 4-8 and 2 more) and FCS Nicholls, which add little to Level 2. This accounts for 0.81 points, more than the whole 0.66-point rating gap; other factors pull the other way.

Biggest disagreements

TeamHarbinCFP committeeW-LWhy
Liberty#10#2313-0
  • Against Troy (Harbin #23, the spot the committee gives Liberty), Liberty beat 1 team with 9+ wins (New Mexico St 10-4), so its beaten opponents earned 335.5 Level 1 points in all (270.0 for Troy's), a Level 2 term of 21.0 to 17.3; Troy beat 3 teams with losing records (UL Monroe 2-10, Southern Miss 3-9 and Army 5-6) and FCS SF Austin, which add little to Level 2. This accounts for 4.15 of the 4.58-point rating gap (90%).
  • Liberty won all 13 of its games, against 2 of 13 for Troy (lost to Kansas St 8-4 and James Madison 11-1), and every loss is a game worth 0 points in the Level 1 average. This accounts for 0.76 of the 4.58-point rating gap (16%).
  • Liberty's Level 2 term (21.0) ranks #11 in FBS: Harbin rewards beating teams that won a lot.
Tennessee#34#218-4
  • Against Toledo (Harbin #21, the spot the committee gives Tennessee), Tennessee's beaten opponents earned 209.5 Level 1 points in all (276.0 for Toledo's), a Level 2 term of 14.4 to 19.2. Tennessee beat 4 teams with losing records (Vanderbilt 2-10, UConn 3-9, Virginia 3-9 and South Carolina 5-7) and FCS Austin Peay, which add little to Level 2; Toledo beat 1 team with 9+ wins (Miami OH 11-2). This accounts for 4.60 points, more than the whole 3.42-point rating gap; other factors pull the other way.
  • Tennessee lost 4 of its 12 games (Florida 5-7, Alabama 12-1, Missouri 10-2 and Georgia 12-1), and every loss is a game worth 0 points in the Level 1 average: 8 of its 12 games earn points, against 11 of 12 for Toledo (lost to Illinois 5-7). This accounts for 1.31 of the 3.42-point rating gap (38%).
  • Tennessee has no win over a Harbin top-25 team (best win: UTSA #50).
Oregon St#29#198-4
  • Against NC State (Harbin #19, the spot the committee gives Oregon St), Oregon St's beaten opponents earned 236.0 Level 1 points in all (265.0 for NC State's), a Level 2 term of 16.3 to 18.4. Oregon St beat 3 teams with losing records (Stanford 3-9, Colorado 4-8 and San Diego St 4-8) and FCS UC Davis, which add little to Level 2; NC State beat no team with 9+ wins. This accounts for 2.01 of the 2.67-point rating gap (75%).
  • Oregon St lost 4 of its 12 games (Washington St 5-7, Arizona 9-3, Washington 13-0 and Oregon 11-2), and every loss is a game worth 0 points in the Level 1 average: 8 of its 12 games earn points, against 9 of 12 for NC State (lost to Notre Dame 9-3, Louisville 10-3 and Duke 7-5). This accounts for 0.48 of the 2.67-point rating gap (18%).
Arizona#22#149-3
  • Against Ole Miss (Harbin #14, the spot the committee gives Arizona), Arizona's beaten opponents earned 247.0 Level 1 points in all (299.0 for Ole Miss's), a Level 2 term of 17.2 to 20.5. Arizona beat 5 teams with losing records (Arizona St 3-9, Stanford 3-9, UTEP 3-9, Colorado 4-8 and 1 more) and FCS N Arizona, which add little to Level 2; Ole Miss beat 2 teams with 9+ wins (Tulane 11-2 and LSU 9-3). This accounts for 3.59 of the 3.74-point rating gap (96%).

Compared with CFP committee of 2023-12-03. Written 2026-10-17 from games through 2023-12-03.