The Ultimate NCAA Football 09 Predictions: Penn State

Over the past few days I have posted the results from my NCAA Football 09 simulations. On Wednesday, a 1-season simulation was posted. The results were not the best for Penn State. A 7-6 record, including an Alamo Bowl game loss to Texas A&M. Not the best by a long shot. Of course we know that doing one simulation using theoretical players, play-calling, injuries, etc. can't tell us much. A larger data set could. Therefore I simulated the 2008 college football season 10 times using EA Sports NCAA Football 09. Today's post focuses solely on Penn State and the results for them during the 10 simulations. How should Penn State fair this season according to the video game results? The answer seems to be in line with what many "experts" and Nittany Lion fans believe.

Below is the final Top 25 after aggregating the 10 simulation results together. Overall the simulation seems to Penn State finishing ranked 23^rd in the nation at the conclusion of the 2008-2009 college football season. This seems reasonable. A ranking in that vicinity would most likely mean the team wins 9 games. There are two scenarios as to how PSU could finish there: go 8-4 and win their bowl game or go 9-3 and lose their bowl game. A 9-4 record usually slots a team between spots 16-25.

Final Rank	Team
1	Florida
2	Ohio State
3	Wisconsin
4	USC
5	Missouri
6	West Virginia
7	Oklahoma
8	Clemson
9	Arizona State
10	Tennessee
11	Florida State
12	Texas
13	Texas Tech
14	BYU
15	LSU
16	Boise State
17	Texas A&M
18	Georgia
19	South Carolina
20	Virginia Tech
21	Auburn
22	Nebraska
23	Penn State
24	Pittsburgh
25	UCF

In order to evaluate the teams I came up with a few different variables. The data set includes 10-sim. Average wins, losses, and rank, change in rank, adjusted rank, and the number a team appears in the Top 25, 10, 5, as well as any National Championships won. The following text is a copy of the explanation from my post yesterday. How did I determine average rank and if a team was over- or underrated? Glad you asked. You probably figured out how I got the average wins/losses per team. I, um, averaged their wins/losses over the 10 simulations. Easy enough. The number of appearances in the Top 25, 10, and 5 were simply summed across the 10 simulations as well. The other things took a little creativity.

The average rank is what it is. I took an average of every team's final ranking over the 10 sims. Unfortunately NCAA 09 does not rank any team past 25 at the end of the season (though they do rank them before it). So in order to take a non-ranked season into account I had to, crudely, apply a number to each team. For simplicity I based it on wins (note: I kept the simulations Top 25; those teams' ranks were not altered). Any team that receives an NR (not ranked) needed a value. I gave any team with 11 or 10 wins not in the Top 25 an arbitrary value of 30. Common sense will say that, ultimately, these teams just missed the Top 25 cut and would have been (most likely) ranked between 26 and 35. A rank of 30 fits in there nicely. This type of ranking percolated all the way down to teams finishing with 0 wins (I kept track of 83 of the teams in the game). A team outside of the Top 25 with 9 wins received a rank of 40. Teams with 8, 7, 6, 5, 4, and 3 wins garnered a 50, 60, 70, 80, 90, 100, respectively. The few teams that had awful seasons (2 wins or less) were ranked 110. I know that isn't an ideal way to rank teams out of the Top 25. But without the game doing it, how else can I do it? I think this way is fair.

The average ranks were used to create the adjusted rank. This is simply sorting the average ranks from lowest to highest. The lowest average rank (Florida, 6) was assigned adjusted rank #1 and the lowest average rank (Vanderbilt, 104) given #83. The lower part of the ranked teams is a bit useless since I am missing 37 teams in the study. However it is doubtful that any of those missing teams would ever crack the Top 50 (I used EAs preseason Top 75 and then a few sparse extras that turned up in any of the 10 simulations, that's how I got to 83 teams).

OK all that is left is how you determine how is over- and underrated. To do this I simply took a team's adjusted rank (i.e. Florida at #1) and subtracted it from EAs preseason rank (Florida is 5). In this case Florida would boast a -4. Any number below 0 indicates an underrated team. The larger the number, the more underrated. Overrated is simply the opposite. Georgia was EAs preseason #1. In this study their adjusted rank was 18. That is a +17. That's bad. The larger the positive number, the more overrated they were in the study.

So there ya go. That's how I did it. I feel like I am writing a scientific paper. Enough with methods. On to more Penn State results (resume new content)!

Here are the average wins/losses and rank for the final Top 25...

10-sim Average
Team	Wins	Losses	Rank
Florida	12	2	6
Ohio State	11	2	7
Wisconsin	11	2	7
USC	11	2	8
Missouri	12	2	9
West Virginia	12	1	9
Oklahoma	12	2	9
Clemson	12	2	11
Arizona State	11	3	15
Tennessee	10	3	15
Florida State	10	3	20
Texas	10	3	21
Texas Tech	10	3	22
BYU	10	3	26
LSU	9	4	27
Boise State	10	3	27
Texas A&M	10	4	27
Georgia	9	4	28
South Carolina	9	4	31
Virginia Tech	9	4	32
Auburn	9	4	34
Nebraska	9	4	34
Penn State	9	4	36
Pittsburgh	9	4	39
UCF	9	4	40

So there Penn State is, at an average rank of 36^th in the nation which is adjusted to 23^rd. Not too bad I guess. Except EA's ratings started Penn State at 13^th. Eep. Penn State was/is overrated according to the game. The chart below is an excerpt from the overrated chart created from the 10 simulations. Each of these teams was deemed as overrated according to their average starting rank and end rank.

Team	Avg. Wins	Avg. Losses	Avg. Rank	Δ Rank	Adj. Rank
Oregon State	5	7	82	39	75
Boston College	6	7	71	28	65
Iowa	4	8	91	22	81
Georgia	9	4	28	17	18
Illinois	7	6	51	15	33
Minnesota	5	8	83	12	76
Penn State	9	4	36	10	23
LSU	9	4	27	9	15
Oklahoma	12	2	9	5	7
Auburn	9	4	34	5	21

For Penn State to not be considered overrated they'd have to finish no lower than 17^th, if we give the the underrated/overrated change in rank a leeway of +/- 4. Penn State would have to win 10 games to meet exopectations this season. That is certainly attainable. I alluded to this at the beginning with their final rank, but how many wins did PSU average over the 10 simulations?

The chart below shows 10-sim. Average wins, losses, and rank, as well as their change in rank and adjusted final rank. Teams included are Penn State, the rest of the Big Ten, some of this years top ranked teams, and any teams that may be of interest to Penn State fans (Notre Dame, Oregon State, etc.).

	10-sim Avg.
Team	Ws	Ls	Rank	Δ Rank	Adj. Rank
Alabama	7	6	48	7	29
Auburn	9	4	34	5	21
Boston College	6	7	71	28	65
Florida	12	2	6	-4	1
Florida State	10	3	20	-18	11
Georgia	9	4	28	17	18
Illinois	7	6	51	15	33
Indiana	7	6	60	-15	50
Iowa	4	8	91	22	81
LSU	9	4	27	9	15
Michigan	8	5	49	3	30
Michigan State	8	5	51	0	34
Minnesota	5	8	83	12	76
Missouri	12	2	9	-2	5
Northwestern	6	7	73	-6	67
Notre Dame	7	5	55	-11	42
Ohio State	11	2	7	-1	2
Oklahoma	12	2	9	5	7
Oregon State	5	7	82	39	75
Penn State	9	4	36	10	23
Pittsburgh	9	4	39	-14	24
Purdue	7	5	55	-1	43
Rutgers	8	5	52	-18	37
Texas	10	3	21	2	12
USC	11	2	8	0	4
West Virginia	12	1	9	-2	6
Wisconsin	11	2	7	-6	3

So Penn State averaged 9 wins and 4 losses as predicted. They are expected to finish 23^rd, 1 spot ahead of state rival Pittsburgh (who was underrated). A record of 9-4 would place Penn State 3^rd in the Big Ten, behind Ohio State (11-2, ranked 2^nd) and Wisconsin (11-2, ranked 3^rd). With the distance between these three teams leads me to this conclusion: Ohio State wins the Big Ten and then the Rose Bowl, Wisconsin gets an at-large bid to a BCS Bowl and wins it, Penn State heads to the Capital One bowl and loses. I have Penn State losing the Bowl Game based on the fact that the 3^rd place team in the Big Ten would not be ranked so low prior to the bowl, especially with a record of 9-3. Penn State most likely would be ranked 14^th-17^th prior to the game. A loss drops them to 9-4 and a 23^rd rank. The chart below is the second set of data from the analysis. This chart includes the number of teams a team made the Top 25, 10, 5, and won National Championships. Again it only includes Penn State, the rest of the Big Ten, some of this years top ranked teams, and any teams that may be of interest to Penn State fans.

Team	Top 25	Top 10	Top 5	National Champs.
Alabama	4	1	0	0
Auburn	6	0	0	0
Boston College	1	0	0	0
Florida	10	9	6	3
Florida State	8	2	2	0
Georgia	7	2	2	0
Illinois	3	1	0	0
Indiana	1	0	0	0
Iowa	0	0	0	0
LSU	6	4	1	0
Michigan	3	1	0	0
Michigan State	2	0	0	0
Minnesota	0	0	0	0
Missouri	10	6	3	1
Northwestern	0	0	0	0
Notre Dame	2	0	0	0
Ohio State	10	9	4	1
Oklahoma	9	7	6	0
Oregon State	0	0	0	0
Penn State	4	2	0	0
Pittsburgh	5	0	0	0
Purdue	2	0	0	0
Rutgers	1	0	0	0
Texas	7	6	2	0
USC	10	5	4	1
West Virginia	9	8	5	1
Wisconsin	10	8	5	0

 

Wow, Penn State only finished in the Top 25 on four occasions. That's not to promising. However, two of the four were Top 10 finishes. So it seems to say that if Penn State has a good year it could turn out great. Odds are, according to this analysis, that PSU has only a 40% chance of finishing the year in the Top 25 and 20% chance of being in the Top 10. Not too great. But obviously don't take these simulations as 100% truth. There are plenty of drawbacks of doing this. It was just too fun to pass up.

I was too lazy to accumulate and compile every Bowl game result for every team but I did summarize the BCS games: National Championship, Rose, Orange, Sugar, and Fiesta. Here is a look at them, simulation by simulation…

Sim 1			Sim 2
BCS National Championship			BCS National Championship
USC	Missouri		Clemson	Missouri
41	20		31	37
Rose Bowl			Rose Bowl
Michigan	Washington		Penn State	USC
38	34		14	21
Orange Bowl			Orange Bowl
Clemson	West Virginia		Oklahoma	West Virginia
42	20		35	42
Sugar Bowl			Sugar Bowl
South Carolina	Ohio State		LSU	Boise State
7	42		41	34
Fiesta Bowl			Fiesta Bowl
Florida State	Texas Tech		Wisconsin	Florida
26	7		27	31

 

Sim 3			Sim 4
BCS National Championship			BCS National Championship
West Virginia	Tennessee		Florida	West Virginia
14	40		20	31
Rose Bowl			Rose Bowl
Ohio State	USC		Ohio State	USC
42	28		23	40
Orange Bowl			Orange Bowl
Clemson	Wisconsin		Florida State	Illinois
29	52		21	17
Sugar Bowl			Sugar Bowl
Hawai'i	LSU		Georgia	Arizona State
13	35		28	24
Fiesta Bowl			Fiesta Bowl
Oklahoma	Notre Dame		Texas A&M	BYU
24	17		49	27

 

Sim 5			Sim 6
BCS National Championship			BCS National Championship
Clemson	Oklahoma		Arizona State	Florida
48	34		13	55
Rose Bowl			Rose Bowl
Wisconsin	USC		Ohio State	San Diego State
24	14		13	17
Orange Bowl			Orange Bowl
Ohio State	West Virginia		Georgia Tech	West Virginia
24	34		24	21
Sugar Bowl			Sugar Bowl
Georgia	Oregon		Boise State	Wisconsin
45	28		33	37
Fiesta Bowl			Fiesta Bowl
Missouri	Florida State		Oklahoma	South Carolina
32	28		45	21

 

Sim 7			Sim 8
BCS National Championship			BCS National Championship
BYU	Clemson		Texas	Ohio State
38	45		14	45
Rose Bowl			Rose Bowl
Ohio State	Arizona State		Boise State	USC
35	0		14	21
Orange Bowl			Orange Bowl
Missouri	West Virginia		Florida State	Pittsburgh
			59	21
Sugar Bowl			Sugar Bowl
Florida	Wisconsin		Florida	Wisconsin
53	34		48	7
Fiesta Bowl			Fiesta Bowl
Oklahoma	Tennessee		Oklahoma	Tennessee
48	10		30	16

 

Sim 9			Sim 10
BCS National Championship			BCS National Championship
Florida	West Virginia		Wisconsin	Florida
31	22		30	44
Rose Bowl			Rose Bowl
Wisconsin	USC		Penn State	USC
42	16		16	42
Orange Bowl			Orange Bowl
Clemson	South Carolina		Miami	Pittsburgh
24	20		33	22
Sugar Bowl			Sugar Bowl
Texas	Northern Illinois		Oklahoma	LSU
42	14		21	31
Fiesta Bowl			Fiesta Bowl
Oklahoma	Ohio State		Missouri	Oregon
46	28		48	17

 

Hey, we made the Rose Bowl twice! Unfortunately they were both losses. At least the computer thinks there is a chance for a Big Ten Championship (10%) or highly successful season.

A lot of variance and not too many duplicates overall. There were a lot of different teams taking their game to the big bowls. Good for them. What are the chance of Arizona State, Northern Illinois, Michigan, Hawai'i, and Notre Dame doing that this year? Probably not too high. But, hey, as last year proved, there is a reason for the cliché anything can happen.

For Penn State I chronicled all 10 simulations (for better or worse). The results are below, including Bowl Games (if we made one)...

Penn State	Sim 1	Result	Penn State	Sim 2	Result
vs Coastal Carolina	35-7	W	vs Coastal Carolina	37-0	W
vs Oregon State	27-14	W	vs Oregon State	48-21	W
at Syracuse	10-13	L	at Syracuse	34-13	W
Temple	60-7	W	Temple	37-10	W
Illinois	35-31	W	Illinois	37-34 OT	W
at Purdue	17-14	W	at Purdue	52-21	W
at Wisconsin	20-27 OT	L	at Wisconsin	42-28	W
Michigan	30-51	L	Michigan	31-14	W
at Ohio State	13-30	L	at Ohio State	31-24	W
BYE	-----	-----	BYE	-----	-----
at Iowa	39-16	W	at Iowa	36-0	W
Indiana	31-35	L	Indiana	34-23	W
Michigan State	41-30	W	Michigan State	58-35	W
Bowl Game	Alamo	Result	Bowl Game	Rose	Result
vs Texas A&M	24-37	L	vs USC	14-21	L
Final Record	Wins	Losses	Final Record	Wins	Losses
	7	6		12	1

 

Penn State	Sim 3	Result	Penn State	Sim 4	Result
vs Coastal Carolina	62-7	W	vs Coastal Carolina	62-6	W
vs Oregon State	27-6	W	vs Oregon State	39-3	W
at Syracuse	37-10	W	at Syracuse	41-16	W
Temple	51-20	W	Temple	41-7	W
Illinois	21-31	L	Illinois	17-21	L
at Purdue	35-28	W	at Purdue	45-6	W
at Wisconsin	20-27	L	at Wisconsin	38-34	W
Michigan	41-14	W	Michigan	22-20	W
at Ohio State	42-39	W	at Ohio State	17-35	L
BYE	-----	-----	BYE	-----	-----
at Iowa	38-13	W	at Iowa	48-6	W
Indiana	33-35	L	Indiana	31-22	W
Michigan State	17-30	L	Michigan State	26-27	L
Bowl Game	Insight	Result	Bowl Game	Outback	Result
vs Oklahoma State	33-28	W	vs Ole Miss	29-21	W
Final Record	Wins	Losses	Final Record	Wins	Losses
	9	4		10	3

 

Penn State	Sim 5	Result	Penn State	Sim 6	Result
vs Coastal Carolina	52-3	W	vs Coastal Carolina	56-10	W
vs Oregon State	34-28	W	vs Oregon State	29-27	W
at Syracuse	17-10	W	at Syracuse	34-17	W
Temple	38-7	W	Temple	43-17	W
Illinois	24-6	W	Illinois	38-16	W
at Purdue	28-21	W	at Purdue	30-27	W
at Wisconsin	7-35	L	at Wisconsin	35-45	L
Michigan	24-28	L	Michigan	21-28	L
at Ohio State	14-38	L	at Ohio State	24-34	L
BYE	-----	-----	BYE	-----	-----
at Iowa	48-17	W	at Iowa	54-7	W
Indiana	41-10	W	Indiana	41-40	W
Michigan State	38-12	W	Michigan State	27-17	W
Bowl Game	Outback	Result	Bowl Game	Outback	Result
vs Auburn	31-34	L	vs Ole Miss	16-37	L
Final Record	Wins	Losses	Final Record	Wins	Losses
	9	4		9	4

 

Penn State	Sim 7	Result	Penn State	Sim 8	Result
vs Coastal Carolina	52-6	W	vs Coastal Carolina	63-7	W
vs Oregon State	31-17	W	vs Oregon State	17-10	W
at Syracuse	43-14	W	at Syracuse	52-10	W
Temple	21-9	W	Temple	49-13	W
Illinois	41-17	W	Illinois	38-7	W
at Purdue	28-44	L	at Purdue	25-31	L
at Wisconsin	17-48	L	at Wisconsin	17-49	L
Michigan	17-38	L	Michigan	14-24	L
at Ohio State	17-49	L	at Ohio State	14-56	L
BYE	-----	-----	BYE	-----	-----
at Iowa	49-14	W	at Iowa	52-17	W
Indiana	45-35	W	Indiana	45-7	W
Michigan State	42-16	W	Michigan State	24-30 OT	L
Bowl Game	Outback	Result	Bowl Game	Liberty	Result
vs South Carolina	38-55	L	vs UCF	27-26	W
Final Record	Wins	Losses	Final Record	Wins	Losses
	8	5		8	5

 

Penn State	Sim 9	Result	Penn State	Sim 10	Result
vs Coastal Carolina	37-10	W	vs Coastal Carolina	35-3	W
vs Oregon State	28-15	W	vs Oregon State	61-14	W
at Syracuse	28-7	W	at Syracuse	24-6	W
Temple	21-24 OT	L	Temple	24-3	W
Illinois	41-20	W	Illinois	38-21	W
at Purdue	19-21	L	at Purdue	31-13	W
at Wisconsin	9-28	L	at Wisconsin	31-42	L
Michigan	24-25	L	Michigan	24-6	W
at Ohio State	14-31	L	at Ohio State	31-54	L
BYE	-----	-----	BYE	-----	-----
at Iowa	38-35	W	at Iowa	35-10	W
Indiana	31-34	L	Indiana	27-26	W
Michigan State	40-33	W	Michigan State	49-14	W
Bowl Game	Insight	Result	Bowl Game	Rose	Result
vs Arizona	16-24	L	vs USC	16-42	L
Final Record	Wins	Losses	Final Record	Wins	Losses
	6	7		10	3

 

Simulations 2 and 10 are the best. Number 2 resulted in a Big Ten championship and Rose Bowl berth and 10 took us there as well, but not Big Ten title. Sim. number 9 was dreadful. A losing record? What is this, 2004? Highly unlikely, though based on this there is a 10% chance. God, I hope not. You may be going through the simulations to see how fared against each team or what happened in each Bowl Game. No need, I have you covered.

10-sim Average	Wins	Losses
	9	4
Bowl Game	Appearances	Wins	Losses
Outback	4	1	3
Insight	2	1	1
Rose	2	0	2
Alamo	1	0	1
Liberty	1	1	0
vs. Opponents	Wins	Losses	Win %
vs. Coastal Carolina	10	0	100%
vs. Oregon State	10	0	100%
at Syracuse	9	1	90%
Temple	9	1	90%
Illinois	8	2	80%
at Purdue	7	3	70%
at Wisconsin	2	8	20%
Michigan	4	6	40%
at Ohio State	2	8	20%
at Iowa	10	0	100%
Indiana	7	3	70%
Michigan State	7	3	70%

 

Well that's not a very good bowl record (3-7). We made the Outback Bowl 4 times. I'm not sure how the fan base would react if that were to happen this year. I think most are expecting a Capital One berth (as the Big Ten's #2 or 3 team). Sadly, we didn't make that game once. We also ended up at some pretty crappy bowls (Insight, Liberty).

Looking at our record versus each of the opponents on are schedule this year could be hazardous to your health. While there was at least 1 expected outcome (100% chance of beating Coastal Carolina) there were also a few surprising (100% chance of beating Oregon State, Iowa) as well as disturbing ones. There is a 10% chance of losing to Syracuse and Temple? I hope not. I do like the 80% success rate against overrated Illinois. However the heart of the schedule, the gauntlet in successive weeks at Wisconsin, versus Michigan, and at Ohio State was disheartening. Against the Big Ten favorites, Ohio State and Wisconsin, we won 4 of 20 games (2 versus each team). In reality, I don't dispute this. I think Ohio State and Wisconsin are better than us. I actually would place our chances at taking down Ohio State in Columbus at less than 10%. However I may bump up our chances at Wisconsin, but not by much. Maybe to 30%. The Michigan results hurt much more. Does EA take into account how much we suck against them? Maybe. Only 4 wins against a mediocre team in Beaver Stadium? I hope not. Our win % for that game should be closer to 60%. Elsewhere, the Iowa result surprised me just because we were a perfect 10 for 10. We are better but winning every time was surprising. The Indiana and Michigan State games are at home and could be the last 2 of JoePa's career. I think our chances of winning the Indiana game are ~90%. Michigan State probably is right where it should be.

I definitely call this simulation analysis a success. I think Penn State finished right around where they should, on average. The 9-4 record seems acceptable for now. The game seems to agree that October 11^th-25^th will suck for us. Hopefully we will win 1 of the 3.

I hope you enjoyed this little project. I'd like to do the same for Madden and will attempt to next week, prior to the start of the season. I'm not promising anything but I may get to it. Thanks for reading.