Massey-Peabody
@MasseyPeabody
Followers
17K
Following
249
Media
107
Statuses
3K
Analytics, predictions and occasional preaching. Mostly football-related. Part of @UnabatedSports (https://t.co/TnPHlnCBLx)
Austin / Philly / Vegas
Joined September 2012
In a just world, GMs who hate analytics would have to renounce all assets accumulated through analytics.
0
21
70
There's a downloadable Excel file in that post, too. Here's the way-too-big PNG version.
3
8
50
MP ratings by conference. Messy figure, but still a sobering look at how the other P5 conferences stack up vs. SEC.
5
17
50
Our latest power rankings for the playoff teams:. 1, BAL, 12.7.2, NO, 10.1.3, KC, 8.4.4, NE, 6.6.5, SF, 5.4.8, MIN, 3.5.9, HOU, 3.1.10, PHI, 2.2.11, TEN, 2.2.13, SEA, 1.5.14, BUF, 1.4.16, GB, 0.6.
7
10
41
Our Top 25 teams and their chances to make the playoff. World ain't fair, man. We'd make Auburn a 4-point fave vs. OU on neutral field, but they have less than 1/2 the Sooners' chance of making the playoff (12% v 28%).
4
16
42
0
2
38
PSA: @RufusPeabody & @bcmassey both loving on The Ringer NFL Show this season. We each listen to only one NFL pod, and arrived there independently. Shout out to @bykevinclark & @robertmays.
1
1
39
1/ One of our favorite analyses to run this time of year is a look back at each team's performances, game by game. These are game-grades we give based on that day's "fundamentals" (our model's inputs - YPA, etc.). Opponent-adjusted, of course.
5
1
37
Outstanding. Everyone, repeat after Bill: Priors are good for us, priors are good for us, priors are good for us. .
As mentioned, I’ve made a couple of ratings tweaks. Preseason projections are currently 85% of the ratings, and at the end of the year they’ll still be around 15%.
2
12
32
If you don't understand that a model will be wrong, inevitably, you don't really believe in models.
1
8
32
New CFB rankings:.1 ALA.2 LOU.3 OHST.4 MICH.5 LSU.6 CLEM.7 WASH.8 AUB.9 FSU.10 OU. So, dropped MICH 2, CLEM 1, WASH 1. Can't shake LSU.
3
18
26
For your bowl confidence pools: Don't spend all your winnings in one place.
1
5
23
More important announcement: Your owner is still Dan Snyder.
The Washington Redskins have announced Ron Rivera as the team’s new Head Coach. 🔗:
0
4
24
Here's our top 25, top to bottom. Gold diamonds are projected wins. Green means easier schedule, red means harder.
1
4
19
Our most likely title matchup for P5 conferences. Inspired by @38Godfrey on new #PAPN pod musing that OHST-WISC is the most certain. We felt that way last year, big time, but not this. SEC: ALA-UGA 39%.ACC: CLEM-MIA 34%.PAC12: USC-WASH 33%.BIG12: OU-UT 26%.BIG10: OHST-WISC 24%.
4
8
19
One last preseason analysis: Predicted reg season win total for our top 25 teams. We also show historical wins for a P5 team with the same preseason MP rating. The difference indicates 2018 SOS, measured in terms of incremental wins (green) or losses (red) due to their schedule.
2
7
19
We think we're down to 7 teams. Absurdly early, we get it. Uncertainty, we're all about it. Still:The playoff teams are coming from these 7.
5
25
16
We have OU Defense now up to #13. Last 3 years, they reliably disappointed vs our pre-season expectation. This year the opposite. Is this what a real DC looks like? @Ian_A_Boyd @BlatantHomerism
0
5
18
Models quite different from the 5.5-point market line on this title game. MP: ALA -1.3.S&P+: ALA -1.0 .FPI: ALA -0.3. @RufusPeabody @SethWalder @SBN_BillC.
5
8
18
A look at how teams playing this weekend did over the course of the season, according to our game grades.
5
4
18
Top 5 looks a lot like other folks', though we're even bigger on the Vols than most folks: Bama, LSU, Clemson, FSU, Tennessee.
0
7
15
Our final bracket goes 1) Bama, 2) Ohio State, 3) Oregon, 4) TCU. Easiest yet - straight off the top of our rankings. http://t.co/zMHVbXvW4O.
0
18
2
This afternoon we're realizing WVU is our sleeper pick for the year. We have them winning 8.4 with significant upside. Esp in a down Big 12.
2
3
14
1/ Time of year to play closer attention to the sims. With 3 weeks remaining this is how we see everybody's prospects. .
1
2
15
Probability of making the Super Bowl. BAL 61.KC 25.NE 7.HOU 4.BUF 2.TEN 1. SF 51.NO 22.GB 12.PHI 7.SEA 4.MIN 3.
2
9
16
My panel this Friday at MIT. Can you believe they didn't put me on the top row with these jokers?! Sheesh. ;) #SSAC17
3
1
15
Good time for a PSA: If you're running a sim, one of the most important things you need to worry about it is whether it is wrong the right amount. Noise is a critical tuning parameter. History tells you how often we're wrong - pay attention.
0
3
15
(Best thing about baseball is you can pay virtually no attention to it while legitimately watching it).
0
0
15
Talked w/ @bencfalk on @WMoneyball this morn. Warming up on NBA analytics and debriefing the Hinkie era. At 32:18:
0
1
14
Finally digging into our preseason numbers, warm from the oven. Will get some stuff up over next couple of days.
3
1
14
We are going to post our pre-season numbers shortly. They aren't fully baked yet so we will continue to refine, but want to get something out. We are in the middle of some changes that are taking a bit longer than expect. This version is very close to last year's approach.
0
0
13
Expected win totals for our Top 25, along with our way of looking at strength of schedule.
1
5
12
These are awesome. And, wow, Baker! Hmm. FWIW, we ran across the same "clean pocket" result on a consulting project a few years ago. Surprised us at the time, which is always a pleasure.
0
7
13
Most likely Final 4 in our latest CFP sim (unordered): UGA, tOSU, ORE, TEN. Probability=17%.
3
3
13
What we'd expect the lines to be. Has to count for something. ALA -2.5 v. OHST.ALA -6.0 v. CLEM.ALA -6.5 v. UGA.ALA -7.0 v. OU.
7
8
13
1/ If we can get the intern moving this week we'll have a proper CFB sim. Until then a few musings. .
1
0
13
Ravens take over the top spot in our power rankings, bumping the Pats out for the first time since week 1.
1
1
12
This week's NFL #s are up. Top 5 go: .BAL 11.97.NO 8.52.KC 7.19.NE 7.03.SF 4.81. Full rankings at
1
2
12
Don't think Longhorn fans are ready to be +7 in Ames next weekend. Doubt the books put it there but we will, sadly, disagree.
1
1
11
Three things people are getting wrong about the college football season:
1
8
12
Bowls start tomorrow. Our advice for your confidence pools:
3
0
11
Biggest edge of the season for us on Arizona vs. Dallas tonight. We make ARI a 4-point fave. No, David Johnson is not worth 6 points.
3
2
10
Just when you thought you'd seen every kind of bad beat.
Thanks @DavidPurdum and @BFawkesESPN for representing my viewpoint accurately.
2
0
11
Throwing shade at Mack Brown in the nerdiest of ways. Thank you, @SBN_BillC! Excited about the book.
0
2
10
Would you buy 95%, Eric? Here's what you get running Massey-Peabody's current ratings through @UnabatedSports's sim.
Uhhhh, according to @espn and their FPI, the @Eagles have a 99% chance to make the playoffs. Really? Not sure @WMoneyball would give it those extreme tail-odds!.
3
1
9
This fits with the "broken leg" exception for algorithms. But how can we objectively quantify "the QB ain't right" factor so it can be dropped in as a check on an algo? Sincere question, as it's really important. @PFF_Eric @PFF_Moo @bburkeESPN @RufusPeabody @jeffma @WMoneyball.
This is why you gotta watch the #film on Brees, can't just box score scout
5
2
10
Something surprising we learned from the model/sim last night: UT can take a 3rd conf loss to TCU this weekend and still *very likely* make the Big 12 title game by winning out (v Baylor and Kansas). Longhorn fans don't know this. Thought it was a bug at first. Good model moment.
1
1
11
.@SethBurn is telling us we're wrong - *now* i know football is back!.
@MasseyPeabody That is very high on SF/NO. Slightly low on KC/Bal.
1
1
9
@SethBurn @RufusPeabody @bcmassey @sImPLyDiGGsAGAi Hypothesis: People who believe in Full Kelly have never made a living as a bettor.
5
2
9
A more complete rundown of our BIG10 title matchup probabilities:.OHST-WISC 24%.PSU-WISC 13%.MSU-WISC 10%.IOWA-OHST 7.5%.NWU-OHST 6%. @SBN_BillC.
0
4
10
A visual take on the CFB playoff picture at this point. Trying to capture the range of possibilities.
2
5
10
The Steelers, Bucs, and Raiders blew out their opponents, but games grade out fundamentally much closer. Our grades:.
1
2
9
NFL ratings scheduled to be up tomorrow morning, but they're ready, so we may sneak them out today ;).
1
0
10
Playoff probabilities through week 7:.ALA, 75.CLEM, 74,.OU, 70.OHST, 65.LSU, 45.WISC, 22.PSU, 17.UGA, 14.AUB, 4.UF, 3.ORE, 2.
4
4
10
Could watch for hours.
1
4
10
We see all your Clemson #1's, and we get it. But we'll give you them and 7 points and still take Bama. It's not close.
2
1
9
Great new handbook out for sports analytics: Highly rec'd.
1
2
10
Really it's too much to ask a guy to watch his teams play defense against Oklahoma State and Kansas City on consecutive days.
0
2
8
The stinker in our numbers is ND 50% to make the playoff, which is the 3rd highest number (tho practically same as Mich & OU). We have them only the #12 team in the country. Would be a ~19-point dog vs. Alabama.
1
0
8
This week's Massey-Peabody NFL picks are up. Biggest edges we've seen all season!.
1
0
9
The key question/stat: What is the chance all 4 playoff teams come from the "Big 6" of Bama, Clemson, Ohio State, UGA, OU and Notre Dame? *Feels* like 90%, right? We make it 40%.
2
3
8
Pre-season CFB data are up. Our Top 5: UGA, ALA CLEM, OHST, AUB. Second 5: ND, WASH, OU, PSU, MICH.Full Top 25 here:
1
2
8
Based on first-draft numbers we do want to push back against "just write Bama, Georgia and the Buckeyes into the playoff" narrative. Yes, they're head and shoulders above everyone else. And yes we agree with the market that it's about 4/1 one of them wins the title. But. .
1
1
9
*Thoroughly* enjoyed. Highly recommend for all analytics types.
Got half an hour with the incomparable @DeanO_Lytics this morning. Surprisingly, talked football. And a few other things. Highly rec'd (@ 53:06):
0
5
8
The variance always amazes me, and is a good reminder that most anything could happen. Nevermind losing games you should win and winning games you should lose. Also fun to squint hard and try to suss out any meaningful trends.
4
1
9
Are there *any* adults at Baylor? PRETTY sure regents should just turn the whole thing over to Jo & Chip. The ultimate #FixerUpper.
Kim Mulkey’s dismissive attitude toward a serious issue is proof some Baylor people still don’t get it. COLUMN:.
0
2
9
Our brothers from a different mother. Guessing our numbers will be a shade lower b/c we tend to (appropriately!) have more variance in our sims than even these guys, but they go about things very similarly.
Fun read. Great pre-pre-season, high-level look at CFB 2019. FPI + @SethWalder is a very effective combination.
0
0
9
NFL futures are also up: We have NE, PIT, JAC and LAC (!) as AFC division winners, PHI, MIN, NO and LAR (barely over SEA) in the NFC. Pats & Vikes in Atlanta.
5
1
9
So you think you want to do basketball analytics.
This year.Things I promised my advisor I would do over break: numerous.Things I actually have done: taught myself how to make this animation
1
2
8
If @bburkeESPN's talking, you should be listening. Dude's outstanding. Caught up with him yesterday morning on @WMoneyball: (at 53:42).
0
5
9
NFL Week 9 picks are up. Looking to bounce back from a rough Week 8.
0
0
7