Thursday, October 31, 2013

Florida Atlantic University Head Football Coach Resigns

The Sun-Sentinel reports that Florida Atlantic University head football coach Carl Pelini (and defensive coordinator Pete Rekstis) have resigned over allegations of illegal drug use. According to College Football Talk, this could result in Carl Pelini paying $500,000 in damages - if pursued by FAU.  While there is a great deal of speculation with regard to this unfortunate incident, let's take a look at the Florida Atlantic University Owls under now former head football coach Carl Pelini's who was hired at the end of the 2011 season and also take a look at the Owls during Howard Schnellenberger's last four seasons (2008-2011) as the Owls head coach. (I am stopping the analysis in the approximated last third of Schnellenberger's tenure at FAU since that is as far back as I have complete data to analyze team performance using the model at hand).

For a quick look at the Owls in terms of total team production ranking, offensive production ranking and defensive production ranking take a look at the graph below.

A more detailed analysis of the Florida Atlantic Owls follows below using the Complex Invasion College Football Production Model.

At the end of last weekend Florida Atlantic was 2-6 with their best win (win over highest currently ranked opponent) over #92 University of Alabama-Birmingham and their worst loss (loss to lowest currently ranked opponent) to #86 ranked Middle Tennessee State.  FAU is ranked as the #101 team in overall productivity (out of 125) with the #105 ranked offense and the #86 ranked defense.  The Owls have played against a strength of schedule (SOS) of 54.75 which was "tougher" than the average schedule to date for the "league".  A teams SOS is "tougher" if it is lower than between one and two standard deviations from the average "league" SOS, which is currently the case.

In former head coach Carl Pelini's first season at the helm of the FAU Owls, the Owls finished 3-9 and were the #110 ranked team in terms of overall productivity with the #112 ranked offense and the #80 ranked defense against an average SOS of 66.08.  The Owls best game was a 37-28 home win over #56 ranked Western Kentucky and their worst performance was a 34-37 loss to #112 South Alabama.
In Howard Schnellenberger's last season as head football coach the Owls finished 1-11.  Florida Atlantic was the #120 (out of 124) ranked NCAA FBS team in overall production, with the dead last ranked offense and the #102 ranked defense.  Florida Atlantic's best (and only) victory was a 38-35 victory over #117 ranked UAB and the Owls' worst performance was a 14-38 loss to #118 ranked Middle Tennessee State.  The Owls played against an easier SOS of 72.42 as compared to the "league" SOS of 63.24.

The Owls finished the regular season at 4-8.  The Owls were the #98 ranked team in terms of overall productivity; with the #109 ranked offense and the #64 ranked defense.  The Owls played against a SOS of 74.92, which is easier (i.e. between one and two standard deviations greater than the "league" average SOS of 63.10).

Florida Atlantic finished the season at 5-7.  FAU's best win was over #87 ranked Louisiana-Lafayette and of their worst loss was to #105 Wyoming all against a SOS that was easier than the average SOS for the "league".  In terms of overall production, Florida Atlantic University was the #85 ranked team with the #50 ranked offense and the #105 ranked defense.

The Florida Atlantic University Owls finished the regular season bowl eligible at 6-6 and won their bowl game over #63 ranked Central Michigan to finish with a winning record at 7-6.  The Owls were the #82 ranked team in overall productivity with the #48 ranked offense and the #103 ranked defense, while playing against an easier SOS as compared to the "league".  The Owls best performance was against #63 ranked Central Michigan and their worst loss was to #76 ranked Middle Tennessee State.

This is the fourth NCAA FBS head football coaching change to date (that I know).  Here is the analysis of the previous coaching changes this season using the NCAA FBS Production Model.
Miami of Ohio and Don Treadwell
UConn and Paul Pasqualoni
USC and Lane Kiffen

Wednesday, October 30, 2013

Wyoming Changes Defensive Coordinator

ESPN reports that the University of Wyoming Cowboys have replaced defensive coordinator Chris Tormey with Jamar Cain.  This season the Wyoming Cowboys are a team of contrasts, with a very productive offense and a rather weak defense using the Complex Invasion College Football Production Model.  If Wyoming's defense was better this team could be a BCS buster - their offense is that good.  Yet, losses to Texas State (better than many think), my alma mater - Colorado State and last week to San Jose State seem to be reasons for the change.  The next four games - only the Hawaii game seems to be a likely win as Fresno State, Boise State and Utah State all match up well with Wyoming.

Here is a graph of Wyoming's total rank, offensive rank and defensive rank using the Complex Invasion College Football Production Model.

Here is Wyoming's regular season schedule, with their results and weekly total, offensive and defensive rankings for each week.  Note that Wyoming's best defensive rank was during their bye week on October 5th.

Date Venue PF PA Result Total Rank Off Rank Def Rank
Nebraska 8/31/2013 Away 34 37 LOSS ----- ----- -----
Idaho 9/7/2013 Home 42 10 WON 16 5 78
Northern Colo. 9/14/2013 Home 35 7 WON 21 5 77
Air Force 9/21/2013 Away 56 23 WON 13 5 61
Texas St. 9/28/2013 Away 21 42 LOSS 36 7 94


30 13 59
New Mexico 10/12/2013 Home 38 31 WON 26 12 70
Colorado St. 10/19/2013 Home 22 52 LOSS 26 5 93
San Jose St. 10/26/2013 Away 44 51 LOSS 53 14 111


Fresno St. 11/9/2013 Home

Boise St. 11/16/2013 Away

Hawaii 11/23/2013 Home

Utah St. 11/30/2013 Away

Here is yesterday's analysis of UAB's replacement of their defensive coordinator earlier this month.

Tuesday, October 29, 2013

UAB Changes Defensive Coordinator

UAB changes defensive coordinator earlier this month.  Let's take a look at UAB's defensive rank for this season up to last weekend using the Complex Invasion College Football Production Model.  I am starting with the second week of the season, as this gives enough data for the regression results to be fairly representative of the season.  The first week was good as well, but just to be on the safe side, I will start with the second week.  Here is UAB's regular season schedule for 2013, with the results to date including their week-by-week defensive ranking, including their defensive rank overall during their two bye weeks this season.


Opponent Date Venue Def Rank PF PA Result
Troy 8/31/2013 Away
31 34 LOST
LSU 9/7/2013 Away 95 17 56 LOST


Northwestern St. 9/21/2013 Home 70 52 28 WON
Vanderbilt 9/28/2013 Away 83 24 52 LOST
Fla. Atlantic 10/5/2013 Home 103 23 37 LOST
FIU 10/12/2013 Away 100 27 24 WON


UTSA 10/26/2013 Away 93 31 52 LOST
Middle Tenn. 11/2/2013 Home

Marshall 11/9/2013 Away

East Carolina 11/16/2013 Away

Rice 11/21/2013 Home

Southern Miss. 11/30/2013 Home

UAB's defense has been in the lower half of the "league" for the entire season, with their best showing in the "league" with their defeat of FCS Northwestern State.  While since the firing of defensive coordinator Reggie Johnson, after the Florida Atlantic loss on Oct. 5th, has resulted in an improvement in defensive ranking, (from 103 out of 125 to 93 out of 125) this is still a rather low production defense.

Monday, October 28, 2013

2013 NCAA FBS Top 25 Ranking for Week 9

Here is the latest Complex Invasion College Football Production Model Top 25 ranking with Baylor still the top team in terms of production.  Other notes are that Florida State has moved ahead of Louisville and that Alabama is making a surge - now the number 5 team overall.  Likewise, Ohio State's outstanding performance over the weekend has allowed them to finally leap over Wisconsin.

These NCAA FBS Top 25 rankings are based on regression analysis using the data posted at College Football Statistics based on the Complex Invasion College Football Production Model.

Rank Team
1 Baylor
2 Oregon
3 Florida State
4 Louisville
5 Alabama
6 Ohio State
7 Wisconsin
8 Missouri
10 Miami (Florida)
11 Houston
12 Michigan State
13 Cincinnati
14 Arizona State
15 Oregon State
16 Texas A&M
17 Northern Illinois
18 Oklahoma State
19 Ball State
20 Utah State
21 Michigan
22 East Carolina
23 UCF
24 Auburn
25 Arizona

Previous Top 25 Ranks for 2013
2013 NCAA FBS Top 25 Ranking for Week 8
2013 NCAA FBS Top 25 Ranking for Week 7
2013 NCAA FBS Top 25 Ranking for Week 6
2013 NCAA FBS Top 25 Ranking for Week 5
2013 NCAA FBS Top 25 Ranking for Week 4
2013 NCAA FBS Top 25 Ranking for Week 3
2013 NCAA FBS Top 25 Ranking for Week 2

Wednesday, October 23, 2013

Martin Biron Retires

The USA Today reports that Martin Biron has retired from the NHL.  So I thought that I would take a look back at his NHL career using our NHL goalie performance measure called Wins Above Average (WAA).  Here is Martin Biron's goals against average (GAA) and his WAA over his career.


Notice that in terms of GAA, his numbers are fairly similar, but in terms of WAA his numbers move around much more.  The Pearson correlation between GAA and WAA is -0.323, indicating that GAA and WAA weakly move in opposite directions.  One of the concepts that we are interested in is how do NHL goalies perform from year-to-year, or how consistent are NHL goalies. Earlier this month I reported that NHL goalies remain "consistent" only about 25% of the time covering the 1997-98 to 2012-13 NHL regular seasons.  To help with that, here is where Martin Biron ranked each season in terms of Wins Above Average (remember that lower numbers are higher or better ranking).

As you can tell from the graph above, Biron was rather inconsistent in terms of his overall WAA, especially over the end of his career.

Tuesday, October 22, 2013

MLB Payroll & Performance from 1988 to 2013

Last week I wrote about the 2013 MLB regular season's team payroll and performance relationship and found that their is zero statistically significant amount of common variation between team payroll and team performance in the regular season.  Problem with this is that it takes into account only one season (hence the use of team total payroll) and one season is a limited sample size.  That I will correct.

In analyzing the relationship between payroll and performance, first let's think about what MLB regular season performance looks like during a season.  As you know, winning percent has an average of 0.500 for each season - thus winning percent as a variable is stationary, meaning that the average of all the teams winning percent over time does not change.  Here is the graph for winning percent.  As you can see, the average of winning percent remains the same over time.

Again, I use relative payroll since relative payroll is a stationary variable in that the mean is one for each season.  Here is graph of relative payroll from 1988 to 2013.  Yes, those spikes are the New York Yankees (smaller ones on the right are the Phillies, Dodgers and Red Sox's). 


When looking at a longer time period (1988-2011), I find that relative payroll (a team's total payroll divided by the MLB season average payroll for all teams) is positive and statistically significantly related to team performance.  I use relative payroll and winning percentage since both variables are stationary and thus the amount of variation that is in common for both will give a better picture as to how well payroll and performance are related.

So, here are the updated results using data from 1988 to 2013 for the regular MLB season.  I find that relative payroll is positive and statistically significant.  The "explanatory power" of relative payroll for the 1988-2011 time period was 17.6%.  Now adding two more seasons I find that it has fallen to 16.4%, indicating that relative payroll and team regular season performance is weakening.

What if instead of using relative payroll I used total payroll?  Since total payroll have been increasing, then this will give weaker statistical results compared to using relative payroll, since total payroll is non-stationary.  As you can see below the variable is increasing from left to right.

Running the regression reveals that the amount of variation that is in common between total payroll and team regular season performance is now 5.9%.