My record last year

Total – SU: 18-14 ATS: 19-13 O/U: 20-12

Week 1 – SU: 10-6 ATS: 10-6 O/U: 11-5

Week 2 – SU: 8-8 ATS: 9-7 O/U: 9-7

Denver at Kansas City

Prediction: Broncos 25-22 (57.4%)

Pick: Broncos +3

Total: Over 42

Houston at Carolina

Prediction: Panthers 22-20 (56.8%)

Pick: Texans +3

Total: Over 40.5

Tampa Bay at New Orleans

Prediction: Saints 28-20 (71.7%)

Pick: Buccaneers +11.5

Total: Over 47

New England at Buffalo

Prediction: Patriots 25-22 (57.5%)

Pick: Patriots +0

Total: Over 45

Arizona at Chicago

Prediction: Bears 23-20 (56.8%)

Pick: Bears +1

Total:  Under 45

San Diego at Cincinnati

Prediction: Bengals 24-21 (58.9%)

Pick: Bengals -3

Total: Under 45.5 

St. Louis at Washington

Prediction: Rams 24-19 (64.5%)

Pick: Rams -3.5

Total: Over 41

Miami and Jacksonville

Prediction: Dolphins 22-19 (59.1%)

Pick: Jaguars +7

Total: Under 41.5

Tennessee at Cleveland

Prediction: Browns 23-19 (60.4%)

Pick: Browns +0

Total: Under 41.5

Detroit at Minnesota

Prediction: Vikings 22-21 (52.2%)

Pick: Lions +3

Total: Under 44

Atlanta at NY Giants

Prediction: Giants 25-23 (56.1%)

Pick: Giants +3

Total: Under 50.5

Baltimore at Oakland

Prediction: Ravens 22-19 (59.5%)

Pick: Raiders +7

Total: Under 43

Dallas at Philadelphia

Prediction: Eagles 27-24 (58.0%)

Pick: Cowboys +5

Total: Under 55

San Francisco at Pittsburgh

Prediction: Steelers 22-21 (50.8%)

Pick: 49ers +6.5

Total: Under 45

Seattle at Green Bay

Prediction: Packers 23-22 (51.6%)

Pick: Seahawks +4

Total: Under 48.5

NY Jets at Indianapolis

Prediction: Colts 24-19 (64.0%)

Pick: Jets +7.5

Total: Under 47

My record last year

Total – SU: 10-6 ATS: 10-6 O/U: 11-5

Week 1 – SU: 10-6 ATS: 10-6 O/U: 11-5

Pittsburgh at New England

Prediction: Patriots 28-23 (65.5%)

Pick: Steelers +7.5

Total: Under 52

Green Bay at Chicago

Prediction: Packers 25-23 (55.5%)

Pick: Bears +7

Total: Under 49

Kansas City at Houston

Prediction: Texans 22-20 (56.7%)

Pick: Texans +0

Total:  Over 41

Cleveland at NY Jets

Prediction: Jets 21-19 (55.7%)

Pick: Browns +3

Total: Under 40

Indianapolis at Buffalo

Prediction: Bills 23-22 (51.0%)

Pick: Bills +2.5

Total: Under 45

Miami at Washington

Prediction: Washington Football Team 23-22 (53.1%)

Pick: Washington Football Team +3.5

Total: Over 43.5

Carolina at Jacksonville

Prediction: Panthers 22-18 (61.3%)

Pick: Panthers -3

Total:  Under 41.5

Seattle at St. Louis

Prediction: Seahawks 22-18 (61.3%)

Pick: Rams +4.5

Total: Over 40.5

New Orleans at Arizona

Prediction: Saints 24-23 (53.9%)

Pick: Saints +1

Total: Under 48.5

Detroit at San Diego

Prediction: Chargers 23-22 (52.3%)

Pick: Lions +2.5

Total: Under 46

Tennessee at Tampa Bay

Prediction: Buccaneers 21-18 (57.2%)

Pick: Titans +3

Total: Under 41

Cincinnati at Oakland

Prediction: Bengals 22-19 (59.5%)

Pick: Bengals -3

Total: Under 43

Baltimore at Denver

Prediction: Broncos 27-21 (66.9%)

Pick: Broncos -4.5

Total: Under 49

NY Giants at Dallas

Prediction: Cowboys 25-21 (61.9%(

Pick: Giants +7

Total: Under 51

Philadelphia at Atlanta

Prediction: Eagles 25-24 (50.3%)

Pick: Falcons +2.5

Total: Under 55

 Minnesota at San Francisco

Prediction: 49ers 24-17 (70.0%)

Pick: 49ers +2.5

Total: Under 41

+1

I have now officially lived in Chicago-land for over a year, and I’m beginning my second academic year at Loyola University Chicago.  After one full academic year, I can say that I still absolutely love it:  I love the department, I love the students, I love Chicago.  I’ve also learned quite a bit in the past year both academic and non-academic.  So here it is.  My advice to someone and thoughts on the first year of the tenure track:

• Find out who can help you with administrative stuff.  Every university everywhere is going to have some arcane system of paper work for getting reimbursed or when you want to purchase something.  Then you have to send it to the right person.  And you’ll always send it to the wrong person.  I’ve worked at two state schools (Umass and UConn) and, while it’s better at a private school, you’ll never escape the paper work.  So find out who knows what they are doing and let them help you.  We have an amazing administrative assistant (I love you Agnes!) who helps me with all of my reimbursements and purchasing paperwork. I didn’t realize this for about 6 months, which is why (at least in my mind) I got nothing accomplished the first semester.  Paperwork is the worst.
• Advertise what you do for research to students.  No matter what you do, some student will be interested in it (well maybe not all areas of research, but most).  And the good ones will come ask if they can help.  Just because they are interested.  Let them help.  It’s a win-win for everyone as long as the student is good.  The real keys are (1) figuring out which students are good and (2) picking an appropriate project for the level the student is at.  These are skills that you won’t learn in grad school or at a post-doc.  You just sort of have to figure this out as you go along (like so many other things in academics.)
• Go to as many department/college/university events as you can.  Even if you think some of these events are corny (and some of them will be), just go.  Go and just meet people.  You never know who you’ll meet at these things.   Or who you’ll be introduced to.  In my day-to-day life in the department there are countless professors who I never get to interact with for whatever reason.  They may have a totally opposite schedule than me, they may be in an entirely different field (i.e. analysis, number theory, any of the other math fields), they may just be trying to avoid other people.  But this might be your only opportunity to meet the really interesting people in your department.  And everyone likes meeting really interesting people.  I met a collaborator who I am currently writing papers and a grant with at a joint math/anthropology department event called “bacon and booze”.  (Science seems to be based on booze.  Data and booze. Quote me on that: The two most important ingredients in science are data and booze.  Is there an event called “Data and Booze”.  If not, I’m starting one. #ramblingOver)
• Where ever you end up, find the local R/python/etc. users group and attend meetings.  For me, this is the Chicago R Users Group (CRUG or ChicagoRUG).  I’ve been to a handful of meetings and I’ve met a bunch of really interesting people from both academics and industry.  And I’ve learned a lot about R.  I’ve even been asked to present twice (#linesForTheCV).
• #lifeAdvice “I can’t do that, I’m not [blank]” This doesn’t really have to do with my job, but it’s something I learned in the last year.  I remember when I was a kid, I’d always think “I can’t do that, I’m not [blank]”.  The blank could be a baseball player or a musician or a skateboarder or a programmer or a business major.  But the thing is, no one IS anything.  If you want to be something, just do it even if you aren’t that thing.  For me this is most relevant in my art in the last year.  Since I’ve moved to Chicago I’ve been submitting my art to shows and I keep getting accepted.  I was even invited to do an entire show of my work, which was up for a weekend this past summer.  So I’m an artist because I say I’m an artist.  And whatever you want to do, you are that just because you say you are that (I mean don’t be delusional about this, you’re not the president of the USA.)  But if you want to be an author, don’t let the fact that you aren’t an author stop you from being an author.  If you want to be a musician, don’t let the fact that you can’t play music stop you.  If you want to be a statistician, don’t let the fact that you don’t have a statistics degree stop you from doing that.  The internet will teach you everything you need to know.  You just need to practice.  Everything is made up.  Be whatever you want.
• I had the opportunity of being on two search committees my first year.  I probably wouldn’t advise this if you had a choice, but I was grateful that I was able to participate on two search committees in my first year.  I was on a computer science search committee and a statistics search committee.  Being involved in such important decision making in my first year really made me feel like my department (and the CS department) valued my opinion.  I also cherished the opportunity to be involved in having input into the direction the department would move in in the next decade.  There are currently 4 tenured/tenure track statisticians in the department and three of them have been hired in the last two years.  This is an incredible opportunity for me to really have a lot of input in the direction that a statistics program will move in next few years.  (I hope I don’t screw it up.)  But I immediately have the chance to try and fix all of the things that I thought were wrong or not perfect about my undergraduate/graduate experience (which overall was excellent!)  Usually a new professor does not have these opportunities to change a program untiul many years into their academic careers.  I’m really excited to have this much influence at the beginning of my career.  (And again, I’ll try not to blow it.)
• Your are going to be tired on Friday night.  When I was a grad student, time was absolutely unlimited.  At least that’s how it felt.  Want to spend a week learning about web scraping?  No problem!  That dissertation can wait!  As a post-doc, I had a little bit less time as there was a project that always needed to be worked on, but I also didn’t have any teaching or service related responsibilities.  Now, as an assistant professor, I still have to do research, but just throw teaching two classes a semester and departmental service (like two search committees!), and you can see that time is not on only no longer unlimited, it’s almost non-existent.  (I’m writing this in a rare free moment during the semester).  Mike Lopez (a.k.a @statsbylopez) had this to say about being tired: “My major issue was getting enough rest, especially the first semester. If I had to do things over again, I would’ve recognized that you can’t keep up the same research pace you had when you were in grad school or a post-doc. As a grad student or a post doc, you have the time nearly every day to devote to major projects.  That time just doesn’t exist when you also have to teach and advise. Time for faculty should be a zero-sum game; roughly fix the hours, and let things fall where they fall. To interrupt sleep, social life, or family responsibilities is a major mistake that many make in their first year (me being one of them).”
• Finally, students have lots of personal problems and sometimes you might need to help.  Remember what it was like being 20?  It’s not easy.  Girl problems, boy problems, family problems, money problems, school problems, life problems.  When you are 18/19/20/21/22, you’ve got these problems.  And sometimes you might be the person that a student feels comfortable talking to about it.  Or, because a student either misses class or assignments, you may end up  hearing about their troubles.  I was not prepared for this, and I’m still not exactly sure how to deal with it.  I guess my best advice is to just sit and listen.  Sometimes that’s what people need.  Just someone to listen.  Other times students tell me about problems that they are going through, and I went through the exact same thing.  Without going into details here, I think it’s really helpful to know that someone went or is going through what you are going through.  I know it made me feel better when I was in college, and I hope that I can do that for someone else.

Cheers.

The 2015 NFL season is almost here.  That means it’s time for my NFL preview  (2014 NFL Preview, 2013 NFL Season Preview)!  Cheers.

Ranks

Retro is based only on games played in the 2014-2015 season and heavily weighs strength of schedule.

Prosp is based on 4 years of data weighted for recency.  It’s based on expected points.

Both Retro and Prosp are framed in terms of the probability of defeating and average team.

Team Retro Prosp DEN 67 77 NE 71 77 GB 73 72 SEA 45 70 SF 66 69 PHI 59 68 NO 40 63 CAR 48 60 HOU 53 56 DAL 49 55 DET 64 54 CIN 54 53 BUF 53 52 PIT 56 51 IND 62 49 KC 59 49 ATL 42 48 CHI 34 48 BAL 62 48 WAS 37 47 SD 53 46 MIA 47 43 NYG 41 43 ARI 61 43 MIN 51 42 NYJ 35 39 STL 41 38 CLE 46 36 OAK 38 28 TEN 28 27 TB 25 27 JAC 38 21

Team WinDivison MakePlayoffs MakeSuperBowl WinSuperBowl ARI 0.20 2.20 0.10 0.02 ATL 5.70 22.22 1.40 0.36 BAL 45.60 64.62 6.90 2.40 BUF 4.28 18.66 1.50 0.52 CAR 10.76 34.56 2.72 1.30 CHI 3.22 9.50 0.64 0.14 CIN 37.68 56.50 6.40 2.28 CLE 1.36 2.94 0.10 0.00 DAL 24.78 41.04 3.94 1.56 DEN 96.52 99.38 38.24 23.60 DET 10.32 29.16 2.18 0.78 GB 85.98 92.52 19.74 10.80 HOU 43.46 62.52 8.10 3.00 IND 54.20 70.70 8.84 3.58 JAC 0.02 0.04 0.00 0.00 KC 1.36 24.02 1.20 0.46 MIA 9.10 32.66 2.48 0.74 MIN 0.48 1.84 0.04 0.00 NE 85.22 95.26 21.14 13.34 NO 83.46 92.72 20.62 9.08 NYG 4.48 8.88 0.64 0.08 NYJ 1.40 6.62 0.50 0.16 OAK 0.00 0.06 0.00 0.00 PHI 67.02 76.36 11.26 5.52 PIT 15.36 28.76 2.62 0.94 SD 2.12 32.58 1.68 0.60 SEA 62.90 94.14 22.38 11.54 SF 36.86 87.04 13.96 7.06 STL 0.04 0.70 0.02 0.00 TB 0.08 0.80 0.02 0.00 TEN 2.32 4.68 0.30 0.00 WAS 3.72 6.32 0.34 0.14

Projected Records

Team – (Median wins) expected wins

AFC East

New England – (11-5) 11.31

Miami – (8-8) 8.06

Buffalo – (7-9) 7.20

NY Jets (6-10) 6.24

AFC North

Baltimore (9-7) 9.31

Cincinnati – (9-7) 9.00

Pittsburgh – (8-8) 7.80

Cleveland – (5-11) 5.35

AFC South

Indianapolis – (10-6) 9.53

Houston (9-7) 9.18

Tennessee (6-10) 5.54

Jacksonville (3-13) 2.73

AFC West

Denver – (12-4) 12.16

San Diego – (8-8) 8.10

Kansas City – (8-8) 7.70

Oakland – (3-13) 3.10

NFC East

Philadelphia (10-6) 10.25

Dallas (9-7) 8.74

NY Giants (7-9) 6.58

Washington (6-10) 6.29

NFC North

Green Bay (12-4) 11.54

Detroit (9-7) 8.38

Chicago (7-9) 7.03

Minnesota (6-10) 5.49

NFC South

New Orleans (12-4) 11.70

Carolina (9-7) 8.64

Atlanta (8-8) 7.99

Tampa Bay (5-11) 4.88

NFC West

Seattle (12-4) 12.05

San Francisco (11-5) 11.21

Arizona (6-10) 5.81

St. Louis (5-11) 5.09

Projected Playoffs

AFC

1. Denver

2. New England

3. Indianapolis

4. Baltimore

5. Houston

6. Cincinnati

NFC

1. Seattle

2. New Orleans

3. Green Bay

4. Philadelphia

5. San Francisco

6. Dallas

Projected Wild Card Round

AFC

Indianapolis beats Cincinnati  24-22

Baltimore beats Houston 23-20

NFC

Green Bay beats Dallas 27-22

Philadelphia beats San Francisco 23-22

Projected Divisional Round

AFC

Denver beats Baltimore 27-21

New England beats Indianapolis 29-23

NFC

Seattle beats Philadelphia 26-21

New Orleans beats Green Bay 27-26

Projected Conference Round

AFC

Denver beats New England 29-25

NFC

Seattle beats New Orleans 26-22

Super Bowl

Denver beats Seattle 25-23

Season Long Bets

Win Totals

Arizona Under 8.5 -105

Buffalo Under 8.5 -130

Denver Over 10.5 Even

Jacksonville Under 5.5 -115

Minnesota Under 7.5 +200

New Orleans Over 8.5 -145

New York Jets Under 7.5 -140

Oakland Under 6 -115

San Francisco Over 6.5 Even

St. Louis Under 8 -165

Seattle Over 11 -135

Win Division

Philadelphia +125

Green Bay -250

New Orleans +220

San Francisco +2000

New England -140

Baltimore +150

Houston +550

Denver -160

Win AFC

Cincinnati +2000

Denver +400

Houston +2200

Win NFC

New Orleans +2200

Crazy Long Shot Super Bowl Match-Up

New Orleans vs Houston +50000

I got back from #JSM2015 in Seattle yesterday.  While I was there I compiled this list of links of interesting things (talks, R packages, etc.) that I took away from JSM2015.  There are a ton of slides that I would love to add to this list (i.e. the rest of the talks from the session that @styatsbylopez organized), so if you have a link to anything like that, please send it my way and I’ll add it.

Cheers.

Interesting talks at JSM:

• “Automatic Forecasting at Scale” Sean Taylor (@seanjtaylor)
• “Relax, I’m a Data Scientist” Jenny Bryan (@JennyBryan)
• “Interactive graphics for high-dimensional genetic data” Karl Broman (@kwbroman)
• “How Industry Views Data Science Education in Statistics Departments“, Chris Volinsky (@statpumpkin )
• “intRo: Statistical Analysis Software for Teaching“, Erix Hare and Andee Kaplan
• “Data Wrangling for the Lahman“, Ben Baumer (@baumerben)
• “Recent Advances in Interactive Graphics for Data Analysis“, Carson Sievert

Interesting talks at JSM about Sports (with some R packages): 

• “Building an NCAA men’s basketball predictive model and quantifying its success” My slides. (@statsinthewild)
• “Refs – They’re Just Like Us! Adversarial and Social Pressures in the NFL“, Mike Lopez (@statsbylopez)
• The full working paper about referees behavior under pressure, Mike Lopez
• “Acquiring, Visualizing, and Modeling MLB Umpire Strike/Ball Decisions with PITCHf/x Data“, Carson Sievert (@cpsievert)
• Pitch RX, Carson Sievert
• The Deuce package in R for tennis data, Stephanie Kovalchik
• Paper: “A Machine Learning Strategy for Predicting March Madness Winners“, Jordan Gumm, Andrew Barrett, and Gongzhu Hu

Shameless self-promotion:

• openWAR package
• opeWAR paper

Interesting links from JSM:

• ASA curriculum guidelines
• Information about Datafest
• HTML Widgets by @Rstudio
• Man Games Lost
• “U.S. Gun Deaths in 2013” (@Periscopic)
• “Statistical Model Predicts Play-calling of NFL Teams with High Accuracy” http://www.scientificcomputing.com