Sun 9/20 · 7:20 PM CDT · final grade · injuries: 7 out · 4 on IR · 1 questionable
SPREAD · Colts +6
no lean
MARKET+6-111MODEL+6-111
cover 53% · market 51%
ML · Colts
no lean
MARKET+230MODEL+230
win 30% · market 29%
TOTAL · over 46
no lean
MARKET46+101MODEL45.1points
over 49% · market 48%
the letter is how far we sit from the market, not how sure we are
THE QUARTERBACKS
Indianapolis Colts · Daniel Jones — started
Rates +0.04 expected points a dropback (18th of the 32 starters this week), over his last 802 dropbacks.
This season: 19 of 31, 166 yards, 1 TD, 1 INT in 1 game.
Kansas City Chiefs · Patrick Mahomes — started
Rates +0.12 expected points a dropback (7th of the 32 starters this week), over his last 818 dropbacks.
That is 0.06 above whoever threw the Chiefs' last 10 games, so their recent numbers undersell this offense.
This season: 15 of 27, 184 yards, 2 TD, 1 INT in 1 game.
On rating Patrick Mahomes is 0.08 a dropback better than Daniel Jones, about 3 points over a game's dropbacks.
What it means for the grade: nothing yet. No letter reads this rating. Tested on 2021-25, it made the all-factors model better but no better than today's letter for betting, and the moneyline market weighs a quarterback about as far as past results justify.
SPREAD · Colts +6
SPREAD: Graded, no lean — the number is where we'd put it.
THE PICTURE — what the number already weighssized against each other · measured against the line — a stat the market already prices can point opposite its usual read
Power ratings give the Chiefs an 80% win chance here.100%
Market: Colts +6 at -111 (pinnacle, quoted 6:45 PM CDT) · the full picture prices it -111 · divergence 1 point — band none.
MONEYLINE · Colts
ML: Graded, no lean — the number is where we'd put it.
THE PICTURE — what the number already weighs
Power ratings give the Chiefs an 80% win chance here.77%
The Chiefs keep drives on schedule on 39% of snaps.23%
Market: Colts at +230 (pinnacle, quoted 6:45 PM CDT) · the full picture prices it +230 · divergence 1 point — band none.
TOTAL · over 46
TOTAL: Graded, no lean — the number is where we'd put it.
The closing total on NFL games predicts the outcome about as well as a coin flip — nobody reliably out-guesses it, and we won't pretend to. Here's what the number already weighs.
THE PICTURE — what the number already weighs
The Chiefs go for it on 4th-and-short 79% of the time this season.100%
Market: over 46 at +101 (pinnacle, quoted 6:45 PM CDT) · the full picture prices it +103 · model expects ~45.1 points · divergence 1 point — band none.
INJURY REPORT
the wire as of 7:15 PM CDT · polled every five minutes inside the last ninety before kickoff · shown whether or not the number weighs it — the closing line already carries the report, so the model reads none of this (audition 2026-09-01)
Indianapolis Colts · 6 out · 2 on IR · 1 questionable
Riley LeonardQBoutcoach's decisionout: played 0% of the time (n 131)Sun 9/20 6:00 PM CDT
Ashton DulinWRoutankleout · did not practice: played 0% of the time (n 487)Sun 9/20 5:55 PM CDT
Austin AjiakeLBoutcoach's decisionout · full practice: played 0% of the time (n 99)Sun 9/20 5:55 PM CDT
DJ GiddensRBoutkneeout · full practice: played 0% of the time (n 99)Sun 9/20 6:03 PM CDT
Dalton TuckerGoutcoach's decisionout: played 0% of the time (n 687)Sun 9/20 5:55 PM CDT
George Gumbs Jr.LBoutcoach's decisionout: played 0% of the time (n 527)Sun 9/20 5:55 PM CDT
Sean McKeonTEquestionableundisclosedquestionable: played 67% of the time (n 325)Tue 9/8 3:42 PM CDT
2 on injured reserve — season-long, already in the line · show
Carson TowtTEinjured reserveundisclosedwill not playTue 9/8 3:42 PM CDT
D.J. MontgomeryWRinjured reserveabdomenwill not playMon 8/31 12:16 PM CDT
Kansas City Chiefs · 1 out · 2 on IR
Omarr Norman-LottDToutknee - acl (surgery)out: played 0% of the time (n 523)Mon 8/31 12:45 PM CDT
2 on injured reserve — season-long, already in the line · show
Cooper McDonaldLBinjured reservekneewill not playWed 9/16 7:38 PM CDT
Jimmy HolidayWRinjured reserveundisclosedwill not playMon 8/31 12:40 PM CDT
PLAYER PROPS
projected from play-by-play — team volume x player share x efficiency, every piece shrunk; receivers' yards and catches from the receiver read, a model of their targets, air yards and who is throwing — against the latest captured line · the model's % is corrected against a replay of every 2025 game · a lean needs 8 points of leftover after the vig-free price, except a tight end's touchdown, which the touchdown read prices and leans on whenever its chance beats the raw price · FAIR is the chance once the model gets only the say it has earned against the book this season, market by market (none on backs' carries and rushing yards, QB passing yards and attempts or touchdowns; a tenth to a sixth on the rest; on the receiver read about a quarter on yards and two-fifths on catches) — the return per $1 is priced on it, so a lean can show a loss · leans rank by how often leans of their kind (market and side) have hit this season, then by the model's return, receivers' yards and catches half on their own FAIR chance and half on their return · STRONGEST marks the top firm ones: a big sample behind the share, several books on the line, and no sign the book is pricing a role we have not seen · the lane's own record: 1778-1923 on 3701 settled leans, -0.4% at the price · every priced projection: model said over 42%, over hit 42% (n 7,638)
built 6:25 PM CDT · lines as of 6:20 PM CDT
Travis KelceTEreceiving yardsprojects 52line 41.5over 62% vs book 50% · fair 52%lean over -112 betonlineag-2% per $1strongest1 game this season, the rest last season's role; IND pass D +62 yds/g vs league: dropbacks x1.051, yds/att x1.018, catch rate x1.011; 7.1 targets x 7.75 yds/target (shrunk, 103 tgt)103 targets behind the share, receiving yards overs have hit 62% of 53 so far (receivers 58% of 36), 6 books on the line, 12 points of leftover after the 2025 correction, fair chance 52%, -2% per $1 at the price
Keenan AllenWRreceptionsprojects 1.2line 3.5over 9% vs book 51% · fair 45%lean under +100 bovada+11% per $1strongest1 game this season, the rest last season's role; KC pass D -31 yds/g vs league: dropbacks x0.975, yds/att x0.991, catch rate x0.993; KC run D -16 yds/g vs league: carries x0.993, yds/carry x0.993; 2.0 targets x 65% catch rate (shrunk, 108 tgt)108 targets behind the share, receptions unders have hit 57% of 46 so far (receivers 53% of 38), 6 books on the line, 42 points of leftover after the 2025 correction, fair chance 55%, +11% per $1 at the price
Rashee RiceWRreceptionsprojects 2.2line 4.5over 12% vs book 53% · fair 47%lean under +106 draftkings+10% per $1strongest1 game this season, the rest last season's role; IND pass D +62 yds/g vs league: dropbacks x1.051, yds/att x1.018, catch rate x1.011; 3.4 targets x 71% catch rate (shrunk, 109 tgt)109 targets behind the share, receptions unders have hit 57% of 46 so far (receivers 53% of 38), 5 books on the line, 40 points of leftover after the 2025 correction, fair chance 53%, +10% per $1 at the price
Travis KelceTEreceptionsprojects 4.6line 3.5over 71% vs book 56% · fair 58%lean over -147 betonlineag-3% per $11 game this season, the rest last season's role; IND pass D +62 yds/g vs league: dropbacks x1.051, yds/att x1.018, catch rate x1.011; 7.1 targets x 70% catch rate (shrunk, 103 tgt)same read as the receiving yards lean above — one bet, not two
Kenneth Walker IIIRBreceptionsprojects 1.7line 3.5over 15% vs book 42% · fair 38%lean under -156 draftkings+2% per $11 game this season, the rest last season's role; IND pass D +62 yds/g vs league: dropbacks x1.051, yds/att x1.018, catch rate x1.011; 2.4 targets x 74% catch rate (shrunk, 45 tgt)thin sample behind the share (45 tgt)
Josh DownsWRreceptionsprojects 3.1line 3.5over 43% vs book 59% · fair 57%lean under +130 betmgm-2% per $11 game this season, the rest last season's role; KC pass D -31 yds/g vs league: dropbacks x0.975, yds/att x0.991, catch rate x0.993; KC run D -16 yds/g vs league: carries x0.993, yds/carry x0.993; 5.1 targets x 65% catch rate (shrunk, 89 tgt)
Tyler WarrenTEreceptionsprojects 3.9line 4.5over 41% vs book 53% · fair 52%lean under +110 betmgm+1% per $11 game this season, the rest last season's role; KC pass D -31 yds/g vs league: dropbacks x0.975, yds/att x0.991, catch rate x0.993; KC run D -16 yds/g vs league: carries x0.993, yds/carry x0.993; 6.4 targets x 66% catch rate (shrunk, 101 tgt)
Xavier WorthyWRreceptionsprojects 2.7line 3.5over 34% vs book 46% · fair 44%lean under -125 bovada+0% per $11 game this season, the rest last season's role; IND pass D +62 yds/g vs league: dropbacks x1.051, yds/att x1.018, catch rate x1.011; 4.6 targets x 63% catch rate (shrunk, 87 tgt)
Tyquan ThorntonWRreceptionsprojects 0.6line 1.5over 28% vs book 39% · fair 37%lean under -167 draftkings+0% per $11 game this season, the rest last season's role; IND pass D +62 yds/g vs league: dropbacks x1.051, yds/att x1.018, catch rate x1.011; 1.2 targets x 57% catch rate (shrunk, 46 tgt)thin sample behind the share (46 tgt)
Jonathan TaylorRBrushing yardsprojects 89line 79.5over 67% vs book 50% · fair 50%lean over -114 betonlineag-7% per $11 game this season, the rest last season's role; KC pass D -31 yds/g vs league: dropbacks x0.975, yds/att x0.991, catch rate x0.993; KC run D -16 yds/g vs league: carries x0.993, yds/carry x0.993; 19.9 carries x 4.74 yds/carry (shrunk, 299 carries)
38 more priced projections with less leftover
17 projections without a posted line (Brashard Smith, Chris Oladokun, Daniel Jones, Emmett Johnson, Jalen Royals, Mo Alie-Cox)
ANYTIME TD
from the day's BEST ANYTIME TD on the sheet: the model's touchdown leans ranked by its chance the player scores — the top five on a day of two or more games, the one best on a one-game day. A lean means the model's chance beats the price's by 8 points or more, not that it expects a touchdown. They stay out of the list above and off STRONGEST, and the return per $1 uses the book's own chance, since the model has not earned a say on touchdowns · this list: 14-56 on 70 picks, -28.6% at the price (luck alone swings that ±17%) · the model said 36%, they scored 20% · every touchdown lean: 45-236 on 281 touchdown leans, +27.2% at the price (luck alone swings that ±21%) · the model said 27%, they scored 16%
Patrick MahomesQBanytime TDprojects 0.61 expected TDsscores 38% vs book 21% · fair 21%lean yes +370 betrivers-1% per $11 game this season, the rest last season's role; IND pass D +62 yds/g vs league: dropbacks x1.051, yds/att x1.018, catch rate x1.011; 3.03 team TDs x 20% share (TD share 26%, red-zone touch share 14%)no. 4 of the day's best anytime TD
SITUATIONAL CONDITIONSdetected for this game · not a grade input — anything that carries weight appears in THE PICTURE
none detected for this game
Honest row: grades cap at the band — extreme disagreement is as often our error as the market's. Totals carry no lean this season by ruling, not by this game's number. Graded against the line at grade time; the model was validated against closing lines. Weather from forecast; regrade if conditions flip. Injury wire as of 7:15 PM CDT; a game-time decision can flip after the last poll.