• statsinthewild's avatarStats in the Wild

    I went the the Cubs game today against the Dodgers.  I got to see Clayton Kershaw pitch (Awesome) and watch the Cubs get destroyed (Expected).  It was 14-3 going into the 9th inning and some people were joking that a touchdown would get the Cubs right back in the game.  (The Cubs “rallied” for two in the 9th to ultimately lose 14-5.)

    But, this got me thinking.  14-3 does look like a football score, but 14-5 is clearly not.  On the train back to school, I came up with this question: If all you knew about a game was the score, which sport would you guess generated that score?

    To answer this question, I scraped the scores of (almost) all of the baseball and hockey games from 2000-2014, and I found a table of all NFL scores all-time. (I haven’t exhaustively scraped every single NHL and MLB game ever because…

    View original post 247 more words

  • I recently submitted by entry to this Major League baseball data visualization contest.  My entry is below (click on it and zoom in and out to see each graph):

    testBB

    Each graph displays games played, home runs, stolen bases, hits, and on base percentage per season for each player as well as their career on base percentage.  Hits are represented by the color of the bars (darker equals more hits), the width on the bottom of the bar is game played and the width of the top of the bar is home runs.  The dashed line is career on base percentage and the dot represented the on base percentage in a given season.  The diameter of the dot displays the number of stolen bases for a given season.

    I don’t think I have a chance to win because apparently you were supposed to visualize the top 10 hitter AND (not OR) the top ten pitchers.  Oh well.  I think my plots are still pretty cool.

    Cheers.

  • irenehelenowski's avatarOut of This Dimension

    It may tell us if what we predict is right.

    Like where Anton may spend the night.

    This is what Randy needs to know.

    So the area-under-the-curve (AUC) can’t be too low.

    Can you just handle all the suspense?

    And will you then give my trilogy a chance?

    If so, I do hope you give it more than 50 percent.

    If I do say so myself, it would be money well spent.

    Chart_AUC_2

    View original post

  • My grandmother passed away Wednesday night.  She was 93 years old.  She had 9 children, 14 grandchildren, and 10 great-grandchildren.

    Amazing.

    Here is an excerpt from her obituary:

    Pauline would eventually meet her future husband, Lieutenant Michael Toms, on a ship heading for Pearl Harbor, Hawaii, where they would both survive the infamous attack on December 7, 1941.

    Amazing.

    Also, here is a joke that she told at Thanksgiving or Christmas a few years ago:

    Two guys were sitting in a room in their nursing home with the door open.  Some women in the nursing home thought the men weren’t getting enough excitement and decided to run past the door naked.  The men saw the women streak by out of the corners of their eyes.  One of the men asked the other, “Did you see that? What were they wearing?”  The other man responded, “I don’t know, but it needs to be ironed.”

    Amazing.

    Screen Shot 2015-11-06 at 3.50.52 PM

    Pauline Evelyn Toms (1922-2015)

  • My record last year

    Total (weeks 1-9) – SU: 83-47 ATS: 66-62-4 O/U: 75-56-1

    Week 1 – SU: 10-6 ATS: 10-6 O/U: 11-5

    Week 2 – SU: 8-8 ATS: 9-7 O/U: 9-7

    Week 3 – SU: 10-6 ATS: 2-14 O/U: 7-8-1

    Week 4 – SU: 10-5 ATS: 9-6 O/U: 9-6

    Week 5 – SU: 7-7 ATS: 9-4-1 O/U: 6-8

    Week 6 – SU: 13-2 ATS: 9-4-1 O/U: 8-6

    Week 7 – SU: 10-4 ATS: 6-8 O/U: 10-4

    Week 8 – SU: 9-5 ATS: 7-5-2 O/U: 7-7

    Week 9 – SU: 8-5 ATS: 5-8 O/U: 8-5

    Cleveland at Cincinnati

    Prediction: Bengals 25-19

    Pick: Browns +13

    Total: Under 45.5

    Tennessee at New Orleans

    Prediction: Saints 28-20

    Pick: Saints -7.5

    Total: Over 48

    Miami at Buffalo

    Prediction: Bills 22-21

    Pick: Dolphins +3

    Total: Under 44

    Washington at New England

    Prediction: Patriots 29-21

    Pick: Washington Football Team +14

    Total: Under 52

    Jacksonville at NY Jets

    Prediction: Jets 23-17

    Pick: Jets -2.5

    Total: Under 42.5

    Oakland at Pittsburgh

    Prediction: Steelers 25-19

    Pick: Steelers -4.5

    Total: Under 48

    St. Louis at Minnesota

    Prediction: Vikings 21-18 (NAILED IT!)

    Pick: Vikings -2

    Total: Under 40

    Green Bay at Carolina

    Prediction: Packers 24-23

    Pick: Panthers +2.5

    Total: Over 46

    NY Giants at Tampa Bay

    Prediction: Giants 22-21

    Pick: Buccaneers +2.5

    Total: Under 49.5

    Atlanta at San Francisco

    Prediction: 49ers 23-21

    Pick: 49ers +7

    Total: Under 44.5

    Denver at Indianapolis

    Prediction: Broncos 25-23

    Pick: Colts +5.5

    Total: Over 45

    Philadelphia at Dallas

    Prediction: Cowboys 25-23

    Pick: Cowboys +3

    Total: Over 44

    Chicago at San Diego

    Prediction: Chargers 25-21

    Pick: Chargers -4

    Total: Under 49.5

  • My record last year

    Total (weeks 1-8) – SU: 75-42 ATS: 61-54-4 O/U: 67-51-1

    Week 1 – SU: 10-6 ATS: 10-6 O/U: 11-5

    Week 2 – SU: 8-8 ATS: 9-7 O/U: 9-7

    Week 3 – SU: 10-6 ATS: 2-14 O/U: 7-8-1

    Week 4 – SU: 10-5 ATS: 9-6 O/U: 9-6

    Week 5 – SU: 7-7 ATS: 9-4-1 O/U: 6-8

    Week 6 – SU: 13-2 ATS: 9-4-1 O/U: 8-6

    Week 7 – SU: 10-4 ATS: 6-8 O/U: 10-4

    Week 8 – SU: 9-5 ATS: 7-5-2 O/U: 7-7

    Miami at New England

    Prediction: Patriots 29-22

    Pick: Dolphins +8

    Total: Under 51.5

    Detroit at Kansas City

    Prediction: Chiefs 24-21

    Pick: Lions +4

    Total: Under 45.5

    Minnesota at Chicago

    Prediction: Bears 22-21

    Pick: Bears EVEN

    Total: Over 42.5

    Tampa Bay at Atlanta

    Prediction: Falcons 26-20

    Pick: Buccaneers +7

    Total: Under 48.5 

    NY Giants at New Orleans

    Prediction: Saints 27-23

    Pick: Saints -3 PUSH

    Total: Over 49.5

    San Francisco at St. Louis

    Prediction: 49ers 20-19

    Pick: 49ers +8

    Total: Under 39.5

    Arizona at Cleveland

    Prediction: Cardinals 22-21

    Pick: Browns +6

    Total: Under 46

    Cincinnati at Pittsburgh

    Prediction: Steelers 23-22

    Pick: Bengals +1.5

    Total: Under 48.5

    San Diego at Baltimore

    Prediction: Ravens 24-22

    Pick: Chargers +3 PUSH

    Total: Under 50.5

    Tennessee at Houston

    Prediction: Texans 24-19

    Pick: Texans -4

    Total: Under 43.5

    NY Jets at Oakland

    Prediction: Jets 21-20

    Pick: Raiders +3

    Total: Under 44.5

    Seattle at Dallas

    Prediction: Seahawks 22-20

    Pick: Cowboys +6

    Total: Over 41

    Green Bay at Denver

    Prediction: Broncos 25-23

    Pick: Broncos +2.5

    Total: Over 45.5

    Indianapolis at Carolina

    Prediction: Panthers 24-22

    Pick:Colts +7

    Total: Under 46.5

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

  • JSE's avatarQuomodocumque

    A while ago I wrote a little Python code that used career data from Baseball-Reference Play Index (the best $36/year a number-loving baseball fan can spend) to answer the question:  given a pair of teams, which player contributed the most to both teams?  My metric for this is

    (WAR for team 1 * WAR for team 2)

    in order to privilege players who balanced their contributions to both teams.

    So who was the greatest Royal/Met?  In retrospect, this should have been obvious.  How many of the top 5 can you guess?

    View original post 57 more words

  • My record last year

    Total (weeks 1-7) – SU: 68-37 ATS: 54-49-2 O/U: 60-44-1

    Week 1 – SU: 10-6 ATS: 10-6 O/U: 11-5

    Week 2 – SU: 8-8 ATS: 9-7 O/U: 9-7

    Week 3 – SU: 10-6 ATS: 2-14 O/U: 7-8-1

    Week 4 – SU: 10-5 ATS: 9-6 O/U: 9-6

    Week 5 – SU: 7-7 ATS: 9-4-1 O/U: 6-8

    Week 6 – SU: 13-2 ATS: 9-4-1 O/U: 8-6

    Week 7 – SU: 10-4 ATS: 6-8 O/U: 10-4

    Seattle at San Francisco

    Prediction: Seahawks 21-19

    Pick: 49ers +6.5

    Total: Under 42.5

    Buffalo at Jacksonville

    Prediction: Bills 21-19

    Pick: Jaguars +4

    Total: Under 41

    Cleveland at St. Louis

    Prediction: Rams 22-19

    Pick: Browns +6

    Total: Under 42

    Pittsburgh at Kansas City

    Prediction: Chiefs 22-21

    Pick: Chiefs +2

    Total: Under 44.5

    Houston at Miami

    Prediction: Dolphins 24-22

    Pick: Texans +4.5

    Total: Over 44.5

    New York Jets at New England

    Prediction: Patriots 27-20

    Pick: Jets +9

    Total: Under 48

    Minnesota at Detroit

    Prediction: Lions 24-21

    Pick: Lions +2.5

    Total: Over 44.5

    Atlanta at Tennessee

    Prediction: Falcons 24-22

    Pick: Titans +5

    Total: Under 47

    Tampa Bay at Washington

    Prediction: Washington Football Teams 23-19

    Pick: Washington Football Team -3.5

    Total: Under 43

    New Orleans at Indianapolis

    Prediction: Colts 26-25

    Pick: Saints +4.5

    Total: Under 52

    Oakland at San Diego

    Prediction: Chargers 26-19

    Pick: Chargers -4

    Total: Under 47

    Dallas at New York Giants

    Prediction: Giants 23-22

    Pick: Cowboys +3.5

    Total: Over 45

    Philadelphia at Carolina

    Prediction: Panthers 24-23

    Pick: Eagles +3

    Total: Under 46

    Baltimore at Arizona

    Prediction: Cardinals 22-21

    Pick: Ravens +9

    Total: Under 48

  • Through 6 weeks of the NFL season, there are currently five (FIVE!?!?!) undefeated teams in the NFL: The Patriots, Bengals, Broncos, Panthers, and Packers.   This seems like a lot.

    So, as suggested by someone on twitter, I ran a simulation to see how rare this was.  I actually ran 2 quick simulations, the first assuming that each game was a coin flip and the second using the point spread as an implied win probability.

    First Simulation

    Let’s assume that each game is a coin flip.  A winner and loser is chosen for each game using the actual schedule, and we count how many teams are undefeated after 6 weeks.  I simulated the first 6 weeks of the season 100000 times.  How often should we expect to see 5 undefeated teams if all games are a coin flip?  Basically never.

    In just about 50% of the simulations there were 0 undefeated teams.  There were 1, 2, and 3 undefeated teams left in about 37.5%, 11%, and 1.8% of simulations.  In another 147 out of the 100000 simulations, there were 4 undefeated teams, which leaves only 10 of those 100000 seasons were there 5 undefeated teams (and never more than 5).  So if games were truly coin flips, we’d basically never see 5 undefeated teams after 6 weeks given this current schedule.  But games aren’t coin flips.  So we can do a bit better.

    Second Simulation

    Using the point spreads for the first six weeks of the 2015 NFL season, implied win probabilities were calculated for each game.  Game were then simulated based on these implied probabilities from the point spread.  I only ran 10,000 simulations here, but the results are largely the same.  It’s very rare to see 5 undefeated teams.

    In 10,000 simulations, there were 0 undefeated teams 6 weeks into the season about 40% of the time.  There were 1, 2, or 3 undefeated teams about 40% , 16%, and 3.2%, respectively.  In 26 simulations there were there 4 undefeated teams and finally, in only 4 simulations were there 5 undefeated teams.  Even when we account for the individual game win probabilities, it’s still incredibly unlikely to see 5 undefeated teams after 6 weeks.

    Finally, here’s a table for the number of times each individual team was undefeated in 10,000 simulated seasons based on implied win probabilities based on the spread.

    Screen Shot 2015-10-20 at 12.22.51 PM

    Cheers.