SEVERE WX : Flood Warning View Alerts

Friday Feedback: Sports & Weather Forecasting

Forecasting a team's performance & record, leading scorers & other stats is similar to weather.

Posted: Dec. 26, 2018 2:09 PM
Updated: Dec. 26, 2018 6:13 PM

Hey Chad, in this day in age I see forecasts are not right sometimes.  This is everywhere and all over.  You would think this day in age with all the technology this wouldnt happen.  Why are forecasts just off sometimes?

John D.

_______________________________________________________________________________________________________

Thank you for the question John!

Weather forecasting is much like sports forecasting.  The equivalent of my job would be to forecast the outcome of every high school sporting event in the area based on math, research & analog of the school's past performace in the program over a period of decades & how the current talent ranks with past talent.  However, let's say that my forecast of every athletic event could cost people money, time or put them in danger based on what I say. 

That is basically what my job is.

Lets take a couple of examples here.

Who would have thought the Colts would be where they are now (at 9-6, just behind the 10-6 Texans in their division) after a 1-5 start to the season?  If this were weather, we'd probably be criticized for missing this forecast in thinking that the Colts would be near/at the bottom of the division.  That forecast would be based on the high likelihood of no playoff berth because in the past, analog shows & very high likelihood that they would not have a good season based on such a bad start.  This statistic permeates all teams in the NFL, according to league analog (history).

Mathematically, this should not have happened, but there is a missing piece here as to why they have won so many games as of late.  What numerical data is it?  What has changed?  Why?  If we could rank psychological effects, what has changed?  What has Frank Reich & all of the other coaches of the team done to skew this forecast?  If we made the forecast when they were 1-5 we would have missed it.  Data changed. 

Forecasting the very next game may be an easier forecast (sometimes a difficult one) because we know how each team performed in the previous game & we know the coaches' style in these situations.  Overall, forecasting closest to the next games should easier than forecasting for the tournament or games well down the road.  Games down the road do not accout for factors unseen.  For some teams, the math points for them to stay on top & for others, it looks as if they should be toward the bottom.  However, other factors contribute to an outcome that may surprise you.

Let's take Purdue men's basketball for example or really, a better example would be for me to forecast the entire Big Ten teams' season, statistics & records.

Here is their schedule from here on out to the end of the season.  We does not even account for the tournament.  Let's say we have to forecast the tournament brackets, too & we have to give the seeds.  We need to forecast how the crowd noise may affect the next team & how different players respond to noise.  Again, this is like short, medium & long range forecasting.  All eyes are on what I say & what I predict with Purdue.  Money, livelihoods, schools, safety, economic planning, etc. are all based on it. 

Dec. 29, vs. Belmont, 4:30 p.m., FS1

Jan. 3, Thursday, vs. Iowa, 7 p.m., BTN

Jan. 8, Tuesday, at Michigan State, 9 p.m., ESPN/ESPN2

Jan. 11, Friday, at Wisconsin, 9 p.m., FS1

Jan. 15, Tuesday vs. Rutgers, 7 p.m., BTN

Jan. 19, Saturday, vs. Indiana, 2 p.m., Fox

Jan. 23, Wednesday, at Ohio State, 7 p.m., BTN

Jan. 27, Sunday, vs. Michigan State, 1 p.m., CBS

Jan. 31, Thursday, at Penn State, 6:30 p.m., FS1

Feb. 3, Sunday, vs. Minnesota, noon, BTN

Feb. 9, Saturday, vs. Nebraska, 8:30 p.m., BTN

Feb. 12, Tuesday, at Maryland, 6:30 p.m., BTN

Feb. 16, Saturday, vs. Penn State, 4 p.m., BTN

Feb. 19, Tuesday, at Indiana, 7 p.m., ESPN/ESPN2

Feb. 23, Saturday, at Nebraska, 4 p.m., BTN

Feb. 27, Wednesday, vs. Illinois, 8:30 p.m., BTN

March 2, Saturday, vs. Ohio State, 2 p.m., ESPN/ESPN2

March 5, Tuesday, at Minnesota, 8 p.m., BTN

March 9, Saturday, at Northwestern, 2:30 p.m., BTN

Here at the current Purdue mens game statistics:

Player GP MIN PPG RPG APG SPG BPG TPG FG% FT% 3P%
Carsen Edwards 12 32.9 26.0 3.0 3.8 1.3 0.3 3.4 .442 .894 .398
Ryan Cline 12 33.0 13.9 3.0 3.5 0.8 0.2 1.3 .448 .875 .400
Evan Boudreaux 12 17.8 8.2 4.9 0.8 0.3 0.1 0.8 .444 .737 .273
Matt Haarms 12 18.3 7.3 4.4 1.0 0.1 1.4 1.8 .583 .583 .273
Nojel Eastern 12 27.3 5.3 4.4 1.9 1.5 0.5 1.3 .588 .250 .000
Grady Eifert 12 23.2 4.8 4.9 1.2 0.6 0.1 0.6 .500 .500 .440
Aaron Wheeler 12 13.7 4.7 3.4 0.3 0.3 0.7 0.8 .413 .778 .367
Trevion Williams 10 4.8 3.1 1.8 0.2 0.2 0.3 0.1 .565 .625 .000
Sasha Stefanovic 12 15.0 2.9 1.9 0.6 0.8 0.2 0.9 .324 .500 .345
Eric Hunter Jr. 12 13.3 2.9 1.5 1.8 0.2 0.1 0.4 .316 .857 .278
Tommy Luce 5 2.2 1.0 0.0 0.0 0.0 0.0 0.2 .500 1.000 1.000
Kyle King 5 2.0 0.0 0.8 0.2 0.0 0.0 0.0 .000 .000 .000
Totals 12 -- 79 36 15 6 4 12 .458 .739 .375

Here at the current season statistics:

Carsen Edwards 395 103 233 59 66 47 118 312 6 30 36 46 41 15 3
Ryan Cline 396 60 134 7 8 40 100 167 2 34 36 42 16 10 2
Evan Boudreaux 213 32 72 28 38 6 22 98 29 30 59 9 10 3 1
Matt Haarms 219 35 60 14 24 3 11 87 18 35 53 12 22 1 17
Nojel Eastern 328 30 51 3 12 0 3 63 20 33 53 23 16 18 6
Grady Eifert 278 21 42 4 8 11 25 57 27 32 59 14 7 7 1
Aaron Wheeler 164 19 46 7 9 11 30 56 8 33 41 3 10 4 8
Trevion Williams 48 13 23 5 8 0 0 31 6 12 18 2 1 2 3
Sasha Stefanovic 180 12 37 1 2 10 29 35 2 21 23 7 11 10 2
Eric Hunter Jr. 160 12 38 6 7 5 18 35 8 10 18 21 5 2 1
Tommy Luce 11 1 2 2 2 1 1 5 0 0 0 0 1 0 0
Kyle King 10 0 0 0 0 0 0 0 0 4 4 1 0 0 0
Totals -- 338 738 136 184 134 357 946 145 291 436 180 140 72 44

Gene Keady analog data 1980-2005:

1980–81 Gene Keady 23–10 10–8 4th NIT Semifinals
1981–82 Gene Keady 18–14 11–7 5th NIT Finals
1982–83 Gene Keady 21–9 11–7 2nd NCAA Second Round
1983–84 Gene Keady 22–7 15–3 1st NCAA Second Round
1984–85 Gene Keady 20–9 11–7 5th NCAA First Round
1985–86 Gene Keady 22–10 11–7 4th NCAA First Round
1986–87 Gene Keady 25–5 15–3 1st NCAA Second Round
1987–88 Gene Keady 29–4 16–2 1st NCAA Sweet Sixteen
1988–89 Gene Keady 15–16 8–10 6th
1989–90 Gene Keady 22–8 13–5 2nd NCAA Second Round
1990–91 Gene Keady 17–12 9–9 5th NCAA First Round
1991–92 Gene Keady 18–15 8–10 6th NIT Quarterfinals
1992–93 Gene Keady 18–10 9–9 5th NCAA First Round
1993–94 Gene Keady 29–5 14–4 1st NCAA Elite Eight
1994–95 Gene Keady 25–7 15–3 1st NCAA Second Round
1995–96 Gene Keady 7–23* 6–12* 1st NCAA Second Round
1996–97 Gene Keady 18–12 12–6 2nd NCAA Second Round
1997–98 Gene Keady 28–8 12–4 3rd NCAA Sweet Sixteen
1998–99 Gene Keady 21–13 7–9 7th NCAA Sweet Sixteen
1999–00 Gene Keady 24–10 12–4 3rd NCAA Elite Eight
2000–01 Gene Keady 17–15 6–10 8th NIT Quarterfinals
2001–02 Gene Keady 13–18 5–11 8th
2002–03 Gene Keady 19–11 10–6 3rd NCAA Second Round
2003–04 Gene Keady 17–14 7–9 7th NIT First Round
2004–05 Gene Keady 7–21 3–13 10th
Gene Keady: 493–270 256–169

Matt Painter analog data 2005-2018:

Matt Painter (Big Ten Conference) (2005–Present)
2005–06 Matt Painter 9–19 3–13 11th
2006–07 Matt Painter 22–12 9–7 4th NCAA Second Round
2007–08 Matt Painter 25–9 15–3 2nd NCAA Second Round
2008–09 Matt Painter 27–10 11–7 2nd NCAA Sweet Sixteen
2009–10 Matt Painter 29–6 14–4 1st NCAA Sweet Sixteen
2010–11 Matt Painter 26–8 14–4 2nd NCAA Third Round
2011–12 Matt Painter 22–13 10–8 6th NCAA Third Round
2012–13 Matt Painter 16–18 8–10 T-7th CBI Quarterfinals
2013–14 Matt Painter 15–17 5–13 12th
2014–15 Matt Painter 21–13 12–6 T-3rd NCAA First Round
2015–16 Matt Painter 26–8 12–6 T-3rd NCAA First Round
2016–17 Matt Painter 27–8 14–4 1st NCAA Sweet Sixteen
2017-18 Matt Painter 30-7 15-3 T-2nd NCAA Sweet Sixteen
Matt Painter: 295–148 142–88
Total: 1777–1025[8]

Here is a list of Purdue NCAA tournament data back to 1969 (not regular season or NIT data):

1969 Sweet Sixteen
Elite Eight
Final Four
National Championship Miami (OH)
Marquette
North Carolina
UCLA W 91–71
W 75–73
W 92–65
L 72–95
1977 First Round North Carolina L 66–69
1980 #6 First Round
Second Round
Sweet Sixteen
Elite Eight
Final Four
National 3rd Place Game #11 La Salle
#3 St. John's
#2 Indiana
#4 Duke
#8 UCLA
#5 Iowa W 90–82
W 87–72
W 76–69
W 68–60
L 62–67
W 75–58
1983 #5 First Round
Second Round #12 Robert Morris
#4 Arkansas W 55–53
L 68–78
1984 #3 Second Round #6 Memphis L 48–66
1985 #6 First Round #11 Auburn L 58–59
1986 #6 First Round #11 LSU L 87–94 2OT
1987 #3 First Round
Second Round #14 Northeastern
#6 Florida W 104–95
L 66–85
1988 #1 First Round
Second Round
Sweet Sixteen #16 Fairleigh Dickinson
#9 Memphis
#4 Kansas State W 94–79
W 100–73
L 70–73
1990 #2 First Round
Second Round #15 Northeast Louisiana
#10 Texas W 75–63
L 72–73
1991 #7 First Round #10 Temple L 63–80
1993 #9 First Round #8 Rhode Island L 68–74
1994 #1 First Round
Second Round
Sweet Sixteen
Elite Eight #16 UCF
#9 Alabama
#4 Kansas
#2 Duke W 98–67
W 83–73
W 83–78
L 60–69
1995 #3 First Round
Second Round #14 Green Bay
#6 Memphis W 49–48
L 73–75
1996 #1 First Round
Second Round #16 Western Carolina
#8 Georgia W 73–71*
L 69–76*
1997 #8 First Round
Second Round #9 Rhode Island
#1 Kansas W 83–76 OT
L 61–75
1998 #2 First Round
Second Round
Sweet Sixteen #15 Delaware
#10 Detroit
#3 Stanford W 95–56
W 80–65
L 59–67
1999 #10 First Round
Second Round
Sweet Sixteen #7 Texas
#2 Miami (FL)
#6 Temple W 58–54
W 73–63
L 55–77
2000 #6 First Round
Second Round
Sweet Sixteen
Elite Eight #11 Dayton
#3 Oklahoma
#10 Gonzaga
#8 Wisconsin W 62–61
W 66–62
W 75–66
L 60–64
2003 #9 First Round
Second Round #8 LSU
#1 Texas W 80–56
L 67–77
2007 #9 First Round
Second Round #8 Arizona
#1 Florida W 72–63
L 67–74
2008 #6 First Round
Second Round #11 Baylor
#3 Xavier W 90–79
L 78–85
2009 #5 First Round
Second Round
Sweet Sixteen #12 Northern Iowa
#4 Washington
#1 Connecticut W 61–56
W 76–74
L 60–72
2010 #4 First Round
Second Round
Sweet Sixteen #13 Siena
#5 Texas A&M
#1 Duke W 72–64
W 63–61 OT
L 57–70
2011 #3 Second Round
Third Round #14 Saint Peter's
#11 VCU W 65–43
L 76–94
2012 #10 Second Round
Third Round #7 Saint Mary's
#2 Kansas W 72–69
L 60–63
2015 #9 First Round #8 Cincinnati L 65–66OT
2016 #5 First Round #12 Little Rock L 83–852OT
2017 #4 First Round
Second Round
Sweet Sixteen #13 Vermont
#5 Iowa State
#1 Kansas W 80–70
W 80–76
L 66–98
2018 #2 First Round
Second Round
Sweet Sixteen #15 Cal State Fullerton
#10 Butler
#3 Texas Tech W 74–48
W 76–73
L 65–78

So let's take these statistics & look at likely outcomes by taking all of this data & running through a supercomputer over & over three times a day.

Let's analyze all of this analog data, too.

Then, let's:

Shows the score of each game.  Also, let's show each players GP MIN PPG RPG APG SPG BPG TPG FG% FT% 3P% for each game now-January 15, then after that, just show each player's expected raw contributions & points scored & the score of each game.

Tournament time, since it is so far out, let's just analog & give an idea of if there will be NCAA birth, NIT birth, what seed they will be & why & show what will be the biggest factors to determine this.  Let's show who will be the leading player on the team at that point.

Let's also show how far they MIGHT make it in the tournament & why. 

There WILL be factors in games we will not see.  Let's say a star player develops a stiff wrist before the game because they slam their wrist in a door & instead of scoring the analog data/season data average of 22, they score 10 & that would not be based on the other team's defense, but a sore wrist. 

What if the THIRD leading scorer on the team sprains their ankle 4 minutes into the game, but they are the LEADING shot blocker.  The other team is shorter, so with this player, our team should dominate, but this happened.  The forecast I came up with an hour before the game may be off then.  The computer model for the game did not know this would happen.

This is exactly what I do every single day on-air & on-line.  The difference is that if I am off some, it affects people much more.  Now, if I am off on the Big Ten or Purdue or high school sports, it may affect them monetarily.  More people will come to games if certain sports teams are winning in the area or region.

However, there is much more pressure in weather than sports.  So, if I would have forecasted Purdue having a much better or worse record than the present, as a forecaster, I would know it.  I would hear about it & I would feel bad about it.  This is the career of a meteorologist.  You have really good hits & you are very right for a period, then you have a bust & you hear about it & you hang your head low.

So as I work on the weather forecast now in the short term, this would be my job equating it to sports.  I have to forecast these games.

I also have to show who the leading scorer is & how many points they will have & why, what the half-time score will be, FG % of the team, etc., etc.  This would all be based on mathematical data of each team, computer model projections of each team & how each team performs based on analog in these certain enviroments.

I could take forecasts from other sources, like ESPN, Gold & Black, etc, much like you could pull an NWS forecast for our area.  However, what fun would that be?  The funnest part of this job is putting a forecast together from scratch! 

Saturday, December 29

12:00 PM High Point at #13 Ohio State Big Ten Network

1:00 PM UMBC at Penn State Big Ten Network

2:00 PM Northern Illinois at #8 Michigan State Big Ten Network

2:00 PM Southwest Minnesota State at Nebraska

3:00 PM Florida Atlantic at Illinois

3:00 PM Maine at Rutgers

4:30 PM Belmont at Purdue FOX Sports 1

5:30 PM #15 Wisconsin at Western Kentucky CBS Sports Network

6:00 PM Radford at Maryland

8:00 PM Bryant at #24 Iowa

Sunday, December 30

12:00 PM Binghamton at #2 Michigan Big Ten Network

4:00 PM Mt. St. Mary's at Minnesota

5:00 PM Columbia at Northwestern

Wednesday, January 2

6:30 PM Nebraska at Maryland Big Ten Network

8:30 PM Northwestern at #8 Michigan State

Article Comments

West Lafayette
Broken Clouds
76° wxIcon
Hi: 74° Lo: 57°
Feels Like: 76°
Kokomo
Broken Clouds
76° wxIcon
Hi: 76° Lo: 57°
Feels Like: 76°
Rensselaer
Overcast
70° wxIcon
Hi: 73° Lo: 54°
Feels Like: 70°
Fowler
Overcast
70° wxIcon
Hi: 73° Lo: 57°
Feels Like: 70°
Williamsport
Overcast
77° wxIcon
Hi: 74° Lo: 59°
Feels Like: 79°
Crawfordsville
Clear
71° wxIcon
Hi: 76° Lo: 60°
Feels Like: 71°
Frankfort
Broken Clouds
73° wxIcon
Hi: 74° Lo: 58°
Feels Like: 73°
Delphi
Broken Clouds
76° wxIcon
Hi: 74° Lo: 57°
Feels Like: 76°
Monticello
Broken Clouds
76° wxIcon
Hi: 74° Lo: 55°
Feels Like: 76°
Logansport
Broken Clouds
76° wxIcon
Hi: 72° Lo: 54°
Feels Like: 76°
Some showers
WLFI Radar
WLFI Temps
WLFI Planner

Community Events