Bill Radjewski (aka BlueSCar) | Software guy who dabbles in analytics | patreon.com/collegefootballd…
Cincinnati, OH
- Tweets 4,592
- Following 322
- Followers 11,063
- Likes 1,662
Pinned Tweet
The 2026 Preseason Edition: all the data, all 14 notebooks, both PDF guides.
$49 → buff.ly/mzK6vQW
Already own the 2025 pack? Check your email for a $25 returning-buyer upgrade.
1
2
603
Been working on a new metric. We'll call it the Third Base Index.
Ryan Day's TBI of 0.971 is closer to the maximum possible value of 1.0 than the next closest coach, Ed Orgeron at LSU (0.867).
Maybe I should've called this the Ryan Day Index (RDI).
Which P5 coaches inherited the best situation for success?
In honor of The Game this week, I seek to answer definitively, who was born on third base?
Ryan Day is the obvious answer. We also have two other candidates in contention.
210
1,172
683,990
Which P5 coaches inherited the best situation for success?
In honor of The Game this week, I seek to answer definitively, who was born on third base?
Ryan Day is the obvious answer. We also have two other candidates in contention.
223
1,034
1,307,213
Plotting college football’s DNA through Week 8.
Offense runs right. Defense climbs up.
• Top-right: Complete teams
• Top-left: Defensive anchors
• Bottom-right: Shootout specialists
• Bottom-left: Pain
Where’s your team? 👀
45
114
911
692,887
Are Ohio State football fans the most spoiled fandom in all of sports?
They've been on a run for 70+ years that's unprecedented. And they've taken it up a whole notch over the past 15 years.
The string of coaches going back to Woody is ridiculous.
42
65
842
113,313
Flip it to the defense 🔒
Here’s opponent-adjusted success rate allowed (through Week 4).
Oklahoma’s defense is bullying everyone: elite on standard downs and passing downs. If you’ve played the Sooners, chances are you’ve been smothered.
11
147
672
141,425
Time for one of my favorite annual charts!
This chart looks at which teams are the best at producing NFL talent by comparing draft picks with recruiting classes over relevant years.
No surprise to see the reigning National Champs as the top overperformer here.
28
166
604
237,478
This annual chart looks at how many draft picks a school has produced over the past 5 NFL drafts compared to recruiting class ratings from relevant years.
Michigan, Penn State, LSU some of the top overperformers at producing draft picks. Clemson and Texas A&M not so much.
34
141
515
160,132
Accounting for postgame win expectancies, Oklahoma State has been the "luckiest" team this year. Their 9 wins is 4.2 wins above their expected win total of 4.8 this season.
53
85
461
470,366
From where do players participating in 2023 College Football Playoff hail? Which of the four participants has the most "national" roster?
Here's the player hometown map for the 2023 CFP.
19
46
433
120,355
Top QBs in opponent-adjusted EPA/play through week 6.
Note that this includes all plays (passing and rushing).
19
110
431
104,718
Here are the "unluckiest" teams. Apparently, you can take the Scott Frost out of Nebraska but you can't take the Nebraska out of the Scott Frost. Or something.
They are once again at the top of this metric, coming in at 3.1 wins under their expected total of 8.1 wins.
25
82
362
296,329
Just for fun, decided to filter this down to just Power 4 + ND. Tends to shift the perspective quite a bit.
Still can't believe that Indiana is the most complete team at this point in the season but I'm here for it.
Plotting college football’s DNA through Week 8.
Offense runs right. Defense climbs up.
• Top-right: Complete teams
• Top-left: Defensive anchors
• Bottom-right: Shootout specialists
• Bottom-left: Pain
Where’s your team? 👀
14
45
332
92,122
Who’s really a dual threat?
Each QB’s impact on the ground (x) and through the air (y), using opponent-adjusted EPA.
The top-right quadrant is where the nightmares live — guys like Taylen Green and Gunner Stockton.
15
44
326
138,272
QB efficiency check 🔥
Here are the Top 15 QBs in opponent-adjusted EPA/dropback through Week 4.
USC’s Jayden Maiava is running away with the #1 spot right now. Huge gap between him and the rest of the field.
1
31
251
67,122
Opponent-adjusted QB EPA/play through week 13
Little movement at the top as this remains the Jayden Daniels show.
6
52
208
50,229
Which FBS teams do the best job of translating recruiting success into NFL draft picks?
This chart comparers average class ratings from 2014-2018 to picks from the past five draft classes.
11
55
217
🔥 QB Efficiency Leaders (through Week 6) 🔥
Top 15 QBs in opponent-adjusted EPA per dropback.
USC’s Jayden Maiava still owns the top spot, but Illinois’ Luke Altmyer and Ohio State’s Julian Sayin are making noise. Air Force’s Liam Szarka also quietly cracking the top 5 👀
9
35
222
48,705
Snatching defeat from the jaws of victory.
Based on in-game performance, Auburn would be expected to have nearly 6 wins so far, but they only have 3.
18
31
205
78,788
💨 QB Efficiency Leaders (through Week 11) 💨
Here are the Top 15 QBs in opponent-adjusted EPA per dropback, the ones generating the most value every time they throw.
🔥 Ohio State’s Julian Sayin jumps to the top spot
⚡ USC’s Jayden Maiava right on his heels
7
44
201
31,678
Indiana football, from forgotten to feared.
Curt Cignetti’s squad has reached its highest Elo rating ever after this past weekend's emphatic victory over Oregon.
The Hoosiers are officially elite.
7
14
193
16,527
I will just say this. My primary mission has always been about accessibility. I am not a "vendor". I am one dude with a family and a real job apart from this.
If you take issue with what I work tirelessly to provide for free, I welcome you to find a new "vendor".
(cont)
11
6
176
43,538
Plotting college football’s DNA through Week 9 (P4 + ND) 🧬
Offense runs →
Defense climbs ↑
• Top-right: Balanced juggernauts
• Top-left: Defensive grinders
• Bottom-right: Track-meet offenses
• Bottom-left: Chaos & suffering
Who’s evolving, and who’s regressing? 👀
9
43
177
119,227
Not the strongest year for QBs so far in the B1G. Bit of a drop-off after JJ McCarthy.
Rank here refers to national rank.
12
36
169
62,075
NFL draft picks plotted against overall rankings as recruits. As expected, a lot of density along the y-axis here.
8
42
167
🔥 QB efficiency leaders (through Week 5) 🔥
Here are the Top 15 QBs in opponent-adjusted EPA per dropback.
USC’s Jayden Maiava is still setting the pace — but Kansas’ Jalon Daniels, Bama’s Ty Simpson, and even Vandy’s Diego Pavia aren’t far behind.
8
27
169
98,840
Marcus Freeman has taken ND to Elo heights not seen since the days of Ara Parshegian. Will he be able to pay it off and end the Irish's national title drought?
5
11
163
35,752
Nebraska has placed 1st in this metric in 3 of the past 6 years: 2018, 2021, and 2023.
I count 15 losses since 2018 in which Nebraska had >50% win expectancy based on postgame statistics.
Here are the "unluckiest" teams. Apparently, you can take the Scott Frost out of Nebraska but you can't take the Nebraska out of the Scott Frost. Or something.
They are once again at the top of this metric, coming in at 3.1 wins under their expected total of 8.1 wins.
9
23
155
88,535
Kicker PAAR rankings update: Alabama's Will Reichard expands a wide lead on the season after going 4/4, nailing FGs of 34, 38, 48, and 48 yards against Mississippi State this past Saturday.
UNLV's Jose Pizano shoots up to #2 after making 48 and 52 yarders against Hawai'i.
2
27
162
96,830
🏃♂️ RB Efficiency Leaders (through Week 11) 🏃♂️
Here are the Top 15 RBs in opponent-adjusted EPA per carry, measuring who adds the most value every time they touch the ball.
🔥 Michigan’s Justice Haynes holds onto #1
🌲 Oregon’s Davison and Whittington aren’t far behind
7
27
144
40,167
It's the time of season to bust out opponent-adjusted rushing charts!
This chart plots offensive adjusted line yards against adjusted highlight yards. Ashton Jeanty breaks this chart.
8
32
139
34,659
Updated opponent-adjusted rushing metrics through week 13.
LSU continues to break this chart.
3
22
122
27,371
We are finally at the point in the season where we can begin generating some decent-ish opponent-adjusted metrics.
Here are the top RBs in rushing efficiency thus far by weighted EPA/rush.
Ashton Jeanty is just unreal.
4
18
123
25,196
Opponent-adjusted QB EPA/dropback through Week 13
SMU's Kevin Jennings jumps up to #3 behind Will Howard and Jaxson Dart. Kurtis Rourke takes a small tumble after the loss to Ohio State.
3
25
126
16,640
Updated list of "luckiest" teams based on postgame win expectancies through bowls and CFP Semis.
Washington would be expected to have 3 or 4 losses based on how they've played in each game.
FWIW Michigan would be #33 with just under one expected loss (0.87).
18
23
120
87,039
QB opponent-adjusted EPA/play rankings through week 10
Jayden Daniels jumps to the top by a razor-thin margin over Kaidon Salter.
Includes all plays (pass + rush)
7
29
118
63,746
UConn's comeback win over Temple tops the CFBD Excitement Index at this point in the season.
UConn also had a 96-yard fumble recovery TD as time expired as Temple was attempted to punch in a go-ahead TD of their own from the 1-yard line.
Here's the rest of the Top 10.
3
14
113
16,783
With the additions of UCLA and USC in 2024, the Big Ten's geographic center will move about 220 miles to the southwest from its current location outside Warsaw, Indiana to just outside Peoria, Illinois.
6
25
109
Who’s winning the battle in the trenches? 💪
This chart shows opponent-adjusted line yards per play, a measure of how well teams control the line of scrimmage.
⬆️ = Strong run defense
➡️ = Powerful run blocking
Indiana stands out as the most consistent up front.
2
20
114
63,113
Field goal kickers don’t always get the spotlight but these guys are adding real points above expectation 💥
This chart ranks kickers by Points Added Above Replacement (PAAR) through Week 11.
🏆 Oklahoma’s Tate Sandell leads the nation
🌈 Hawai‘i’s Kansei Matsuzawa right behind
2
13
109
17,937
Opponent-adjusted team offensive success rates through week 5.
Which teams stay ahead of schedule and which can get themselves out of a bind?
6
21
106
58,785
Getting Started with CFB Analytics
zpr.io/xhTgDct8AYgJ
3
16
102
Opponent-adjusted QB EPA/dropback through week 10
Dead-heat at the top between Will Howard, Kurtis Rourke, and Jaxson Dart.
4
22
104
29,725
Of course this chart would not be complete without its companion, the Third Base Index table.
It's long past time to update this chart for the new crop of head coaches.
Ryan Day, you are no longer the third-basest of coaches. That distinction belongs to Kalen DeBoer.
Sherrone Moore, you find yourself just barely on 3rd.
19
16
104
55,703
⏱️ First half vs second half ⏱️
Net success rate by half (offense – defense allowed) through Week 8:
📈 X = 1st half
📈 Y = 2nd half
▪️ Wire-to-wire winners dominate both halves
▪️ Fast starters who fade late live in the bottom right
▪️ 2nd-half surges climb the top left
2
17
109
15,714
Opponent-adjusted QB EPA/play through week 7
Dillon Gabriel jumps JJ McCarthy and Caleb Williams for the top spot this week.
Includes all plays (passing + rushing)
5
27
102
32,507
Opponent-adjusted success rates through Week 12 (P4+ND). A clean look at which teams are driving efficiency on offense, defense, or both. Chart ⤵️
6
24
114
13,135
QB opponent-adjusted EPA/play through week 12
The top three spots in this metric remain unchanged from last week, with Jayden Daniels leading the way.
4
23
97
16,818
💪 Domination in the Trenches (through Week 6)
📈 X = Opponent-adjusted line yards gained (offensive push)
📉 Y = Opponent-adjusted line yards allowed (defensive resistance)
Top-right = true Trench Kings 🏆
Auburn, Indiana, and FSU are bullying opponents up front.
8
17
104
53,286
JJ McCarthy cumulative mean EPA/dropback
Having closely tracked Michigan QBs for a number of years, this is the type of chart I always dreamed of for a Michigan QB.
Still a lot of season left, but very consistently at that high 1.0 EPA/dropback mark.
6
18
97
31,523
Transfer portal data can now be queried via the API and exported via the website. This data will be updated nightly.
Website - collegefootballdata.com/expo…
API docs - api.collegefootballdata.com/…
Python docs - github.com/CFBD/cfbd-python/…
1
14
98
As someone who spent their HS and college years in the aughts, it's weird not having USC as the dominant force they were in the Carroll years.
5
6
99
41,373
Teams that stand out as overachievers?
- Michigan
- Florida
- Washington
- Utah
- Iowa
Would be remiss not to also point out Temple.
3
17
89
It's long past time to update this chart for the new crop of head coaches.
Ryan Day, you are no longer the third-basest of coaches. That distinction belongs to Kalen DeBoer.
Sherrone Moore, you find yourself just barely on 3rd.
12
11
97
78,880
Ashton Jeanty retains the top spot among RBs in opponent-adjusted EPA/rush. A wide gulf separates him and Minnesota's Darius Taylor at #2.
4
28
92
23,351
Appalachian State has been the most exciting team so far based on the CFBD Excitement Index, which is calculated from in-game win probability swings.
The Mountaineers are averaging an Excitement Index of 7.9 per game.
5
22
89
50,507
🚨 Rushing breakdown 🚨
This chart plots opponent-adjusted rushing metrics through Week 5:
📈 X = Line yards (OL push)
📈 Y = Highlight yards (RB breakaways)
▪️ Top-right = Steamrollers + home runs (Arkansas, USC, Michigan, FSU, Mizzou)
▪️ Bottom-left = stuck in quicksand
7
19
93
31,065
QB passing efficiency by opponent-adjusted EPA/dropback through week 4
(minimum 50 dropbacks)
4
15
87
23,137
Nebraska was the elite of the elite during the Devaney and Osborne years. Will Rhule be the one to end their 25+ year journey through the wilderness?
12
11
93
24,252
Who are the top dual-threat QBs through week 5?
Jalen Milroe is in a class of his own in this graphic, which plots WEPA/dropback against WEPA/rush.
These metrics are adjusted for opponent strength and include all QBs that have met the necessary rush and pass thresholds.
3
16
83
32,798
Another fun annual draft chart - a map of the hometowns of 2024 NFL Draft picks.
Only includes picks from the continental US. Larger logo = higher pick (scales by day 1 vs day 2 vs day 3 picks).
7
12
86
19,689
Opponent-adjusted QB EPA/dropback
OSU's Will Howard takes the top spot this week. This sets up a Top 10 QB matchup with Drew Allar, assuming the latter is good to go this weekend.
3
26
86
21,314
Reminder: it’s still early. Opponent adjustments sharpen as the season rolls on, so these will shift.
But for now? USC’s offense is cooking 🍳 and Oklahoma’s defense is suffocating 😤
3
2
85
18,771
Oh definitely not retired! Just taking a break from whatever it is this platform is calling itself these days.
Here ya go!
@CFBNumbers do you think you can create one for this years draft? I think Bill retired
9
22
79
39,030
Where were the 2021 blue chips from and where did they commit?
Higher rated players stack on top of lower rated in this map.
5
26
79
The recent story of Tennessee had been a slow decline from the early Fulmer years into futility. Josh Heupel seems to have gotten them back on the right track.
5
5
83
15,198
Working on a new adjusted line yards metric that takes line yards and adjusts for opponent and down type (standard vs passing).
Miami and Michigan have been the most dominant in the trenches on both sides of the ball in this metric. (Garbage time has been excluded)
3
17
74
48,524
📦 The CFBD Starter Pack is officially live!
Whether you're building models, dashboards, or just exploring data, this is your launchpad.
What’s inside + how to get it 👇
#cfb #cfbd #sportsanalytics
3
12
79
46,168
Some useful links as the season is upon us.
Advanced Box Scores: collegefootballdata.com/boxs…
Advanced Team Stats: collegefootballdata.com/team…
CSV Data Export Tool: collegefootballdata.com/expo…
CFBD API docs: api.collegefootballdata.com/…
Python docs: github.com/CFBD/cfbd-python
(cont)
3
8
76
Opponent-adjusted EPA per dropback through week 7
No change in the top 3. Cam Ward shoots back up to #5 and Will Howard bumps up a couple spots to #4.
3
12
77
19,624
Texas A&M's 74-72 thriller win over LSU in 7OT back in 2018 currently holds the record for the highest Excitement Index I have recorded.
This is the game that prompted the NCAA to nerf OT rules. Can't be having more exciting football to fit in between those ads now can we.
6
11
76
31,524
Opponent-adjusted RB EPA/rush
UCF's RJ Harvey surges ahead of Quinshon Judkins and Kalel Mullings for the #3 spot.
5
19
71
17,631
Opponent-adjusted QB EPA/play through week 8
JJ McCarthy retakes the top spot and Jayden Daniels makes a push.
Includes all plays (passing + rushing)
2
17
69
12,294
Which FBS running back was most productive in the passing game this season? That would be Kenneth Gainwell of Memphis and it's not particularly close.
2
23
69
It's never just been about the stats.
It's about helping people build their first model.
About unlocking access to some really cool tools and capabilities.
The CFBD Starter Pack drops Saturday.
It's not just a product but a launchpad.
#cfb #sportsanalytics #datascience
2
4
73
4,058
SMU has one of the more interesting histories in college football. Many questioned their addition to the new ACC this past offseason, yet they find themselves the class of the conference and in the CFP hunt at this juncture.
1
9
72
6,255
If Ryan Day runs off to Texas A&M, then I don't think I have any choice but to rename this to the Ryan Day Index.
Been working on a new metric. We'll call it the Third Base Index.
Ryan Day's TBI of 0.971 is closer to the maximum possible value of 1.0 than the next closest coach, Ed Orgeron at LSU (0.867).
Maybe I should've called this the Ryan Day Index (RDI).
4
7
68
14,743
🚨 Chart season update 🚨
Here’s how every offense stacks up on opponent-adjusted success rates (through Week 4).
📈 X = standard downs
📈 Y = passing downs
USC is surgical right now, carving up defenses every which way.
1
12
71
14,293
3rd downs are where games are won and lost. 📊
Here’s a look at how offenses are performing so far: plotting average distance to go on 3rd down vs. 3rd down conversion success rates.
IU, USC, Utah, and Mizzou have been the best at both staying ahead of schedule and converting.
6
12
68
10,161
Updated QB opponent-adjusted EPA/play through championship week
Includes all plays (pass + rush)
14
65
20,038
Where do players who will be participating in this year's #CFBPlayoff hail from?
(Only the contiguous US is pictured)
3
13
69
If you were a HS recruit in the 2013 to 2016 classes what were your chances of being drafted if you were rated...
5 stars: 62%
4 stars: 23%
3 stars: 7.5%
2 stars: 3%
1 star: 2%
3
23
64
Lastly to the community and all of my supporters, I appreciate you all. You all are why I keep doing what I do.
I will not put another person's work on blast, either subtly or directly. We need more building up of each other in this community.
1
62
3,990
this isn’t a game preview
this is a starter pack
feels like college football season my friends...
👀📊🏈
2
4
67
4,972
Not that anyone asked, but a full season's worth of college football play data takes up ~200MB on disk. A full season's worth of college basketball data takes up ~700MB. (This includes database table(s) and associated indexes and everything.)
Yeesh.
7
2
64
20,515
Opponent-adjusted Defensive Success Rates
Nine elite defenses in the CFP and then there's Clemson, Arizona State, and Boise State.
2
13
59
58,261
Shared the offensive chart a little bit ago, now for opponent-adjusted defensive success rates through week 5.
Which defenses are prone to busts on 3rd and long? Which leak yards on standard downs?
17
58
17,268
FG Kicker Points Added Above Replacement (PAAR) through week 6.
Michigan's Dominic Zvada increases his lead at the top, adding over 10 PAAR on the season. He has less than half the attempts (8) than SMU's Collin Rogers at #2 (17).
2
13
58
21,644
Domination in the Trenches aka Adjusted Line Yards through week 12
Top-right corner is where you want to be.
LSU's OL is completely dominating on the ground. Their DL... not so much.
3
7
54
24,210
For those of you hearing about us for the first time, we highly recommend you check out our live API documentation.
Not only can you see what all is available, but you can also test every endpoint out using live data, all thanks to tooling from @swagger.
api.collegefootballdata.com
4
13
57
Updated map based on some (much appreciated) corrections people have pointed out. Will update again if any more are found.
5
9
56
16,267
Opponent adjusted QB EPA/play through week 9
Bit of a drop for Caleb Williams. Jayden Daniels makes a push for the top spot and is just behind JJ McCarthy.
Includes all plays (passing + rushing)
2
15
51
9,536
Just dropped a huge new feature on the website in the form of advanced team stats pages. If you are familiar with the Advanced Box Score feature, these are very similar but broken down at the team level.
Check them out!
collegefootballdata.com/team…
2
9
52
I honestly have no idea how popular the blog is as I don't have any metrics on it whatsoever. But if I were to start vlogging or streaming similar type content on building data models and getting into the weeds technically, would there be any interest in that?
14
50
The team with the lowest average Excitement Index at this point? Michigan. The Wolverines average a mere Excitement Index of 2.6 per game with no game cracking above 4.1.
7
7
50
31,486
