ADVERTISEMENT

Simulating PSU's 2015 season 30,000 times .....

M

michnittlion

Well-Known Member
Sep 3, 2003
14,112
3,490
1
If one believes the per-game percentages ESPN showed on-air today, running a Monte Carlo simulation 30,000 times, the results. Take it for what it is worth:

Odds of 12 wins: 0.14%
Odds of 11+ wins: 2.20%
Odds of 10+ wins: 13.2%
Odds of 9+ wins: 36.8%
Odds of 8+ wins: 65.6%
Odds of 7+ wins: 86.6%
Odds of 6+ wins: 96.5%

Odds of 8 conference wins: 0.34%
Odds of 7+ conference wins: 5.05%
Odds of 6+ conference wins: 25.5%
Odds of 5+ conference wins: 58.2%
Odds of 4+ conference wins: 85.1%
Odds of 3+ conference wins: 96.9%
Odds of 2+ conference wins: 99.6%

Total wins: Average = 8.00, Standard Deviation = 1.36
Conference wins: Average = 4.71, Standard Deviation = 1.18
OOC wins: Average = 3.29, Standard Deviation = 0.67

CLl7JfUUcAAJ58J.jpg:large
 
Last edited:
Have a hard time thinking a trip to Northwestern is going to be easier then a trip to Philly.
 
BlueBand said:
This isn't your fathers Temple team. At least on paper. They should be much improved, lots of returning starters.
Click to expand...

NW showed last season what they are capable of. Temple is going to be a home game for us. I really don't see it being close to be honest.
 
I too have a hard time thinking Temple is going to be a real challenge. Apparently their defense is VERY strong. They are all have a great deal of experience. That being said their offense does not appear to be very good. So game one will be a REAL test for our OL. If we can black the Temple D we can have a really good year. If we struggle to block them we may be in of another long year. Vegas has us as a 4 point pick.
I EXPECT we win by 28-10 :)
 
So Temple and Michigan are essentially equal? I'm thinking the extra 2% is the home field advantage. I know Temple is now respectable but I just don't see how they are at the same level as Michigan. Two other surprises, Maryland and NW. I'm also getting the feeling Illinois may be a tougher win than anticipated.

We'll find out soon.
 
  • Like
Reactions: ChiTownLion
dawgduice said:
I too have a hard time thinking Temple is going to be a real challenge. Apparently their defense is VERY strong. They are all have a great deal of experience. That being said their offense does not appear to be very good. So game one will be a REAL test for our OL. If we can black the Temple D we can have a really good year. If we struggle to block them we may be in of another long year. Vegas has us as a 4 point pick.
I EXPECT we win by 28-10 :)
Click to expand...
Several Vegas betting sites have PSU 7 pts
 
PennStateNate said:
I think scUM will be tough. I hate Harbaugh, but he is a good coach. He will have them playing.

Temple maybe better, but they are still Temple.
Click to expand...

This is why many board members refuse to acknowledge or believe that the Temple game could be a real nail biter. It is ingrained in our thinking that Temple is clearly inferior to us. That's understandable because it has been true 95% of the time in our lives. This Temple squad is a pretty darn good team.
 
I think we win. and i think it is closer than any of us would like... and i think that will lead to doubts and concerns that should not be there. we are playing a team that showed us last year that they play serious defense, have an explosive offense, and we played them at home and with them hurting.. better make sure everyone gets that.. temple was down men and hurting. they will be healthy, they will be ready, and we damned well better be the same or we blow it up in game 1.

my guess is 23-13. and it will be close (16-13) for a lot longer than we'd prefer. and we'll get zero respect for it.. but if there is anyone here who doesn't understand that this is temple's game of the century, you better wake up. they are dropping all their eggs in this one basket and they have the hosses to pull it off if we aren't ready to play to that level.

there is a reason the line is 4... this is a winnable game for temple and it changes everything if they pull it off.

the owls will come out with everything in the bag. i hope we win 50-7. i doubt it.
 
  • Like
Reactions: BlueBand
N&B4PSU said:
I think we win. and i think it is closer than any of us would like... and i think that will lead to doubts and concerns that should not be there. we are playing a team that showed us last year that they play serious defense, have an explosive offense, and we played them at home and with them hurting.. better make sure everyone gets that.. temple was down men and hurting. they will be healthy, they will be ready, and we damned well better be the same or we blow it up in game 1.

my guess is 23-13. and it will be close (16-13) for a lot longer than we'd prefer. and we'll get zero respect for it.. but if there is anyone here who doesn't understand that this is temple's game of the century, you better wake up. they are dropping all their eggs in this one basket and they have the hosses to pull it off if we aren't ready to play to that level.

there is a reason the line is 4... this is a winnable game for temple and it changes everything if they pull it off.

the owls will come out with everything in the bag. i hope we win 50-7. i doubt it.
Click to expand...

You might very well be correct with this prediction. But I'm of the belief that PSU and many of the unknowns people mention (OL, special teams, etc.) land with the better-than-forecast outcome. Our defense could be special. I don't think it will be a high scoring game but I'll take out defense to shut Temple down every day when it counts.

PSU 28-17.
 
lobo lion said:
This is why many board members refuse to acknowledge or believe that the Temple game could be a real nail biter. It is ingrained in our thinking that Temple is clearly inferior to us. That's understandable because it has been true 95% of the time in our lives. This Temple squad is a pretty darn good team.
Click to expand...

Temple may be better, but we will also be better and talent wise, it won't be a matchup.

I know being a Philly team, there is alot of closet Temple fans here, but Temple should never be considered a tough game for us and that is nothing against Temple, it just means we should be looking to get better.
 
A factor that plays into many college games is experience. Even award winning seniors often had bumps in the road their first year playing.

PSU was the second youngest team last year and we saw problems related to youth. Perhaps not understanding blocking routines as well across the offense. Perhaps just being smaller and not quite as competitive, without years of intense weight room work.

Fortunately lots of PSU players now have last year's experience as a step up. Yet, PSU will likely still be playing additional players in our scheme for the first year this season.

One advantage Temple may have is a ton of players with multiple years in the weight room and on the field. That may minimize errors and make them near their peak in development. TBD

Games can turn on one or a few plays. Errors, penalties and missed assignments can play factors.

JoePa was known in the early years as waiting to play players rather than to put them in "too early." Expect that times have changed and CJF is more willing to play his stable of stud recruits earlier. Plus from a depth standpoint, it may still be required.

So one thing to watch will be the possible trade-off between PSU trying to get newer higher potential players on the field this year and the real possibility that errors, penalties or missed assignments are part of the early learning process for many players.

This thread has pointed to lots of good points to consider. Makes for more interesting games, having heard differing inputs from many knowledgable posters.
 
So their graphic shows us 53%+ chance of beating 10 of our 12 opponents, yet the odds of 10+ wins is only 13% ?
I'm not a statistics guy, but that seems... wrong. :confused:
 
Sviatoslav said:
So their graphic shows us 53%+ chance of beating 10 of our 12 opponents, yet the odds of 10+ wins is only 13% ?
I'm not a statistics guy, but that seems... wrong. :confused:
Click to expand...

It's a bit counterintuitive at first but if you think about it for a minute it makes sense. Suppose you had exactly 51% chance of beating each of the 12 teams on your schedule. Would you then expect to go 12-0? No, you'd expect to go 6-6 or maybe 7-5. Maybe if you got really lucky or unlucky you'd get up to 8 or 9 or down to 3 or 4 wins. But it would be very unlikely to go 12-0 with a slightly better than coin toss odds of winning each game.
 
The Temple % is a head scratcher. My guess is the computer views that as an away game --- but being located in Philly it won't be like an away game. The crowd will probably be majority PSU.
 
  • Like
Reactions: 94LionsFan
Sviatoslav said:
So their graphic shows us 53%+ chance of beating 10 of our 12 opponents, yet the odds of 10+ wins is only 13% ?
I'm not a statistics guy, but that seems... wrong. :confused:
Click to expand...
It's about 50%, so a coin flip. Would you bet me $100 that one toss of a coin would come up heads? Maybe, because you'd have a 50% shot. Would you bet me $100 that you could toss a coin 12 times and all 12 would be heads? Probably not, because you know that while each individual toss is mostly even, the odds of getting the same result over a longer term is much less likely.
 
PSUSignore said:
It's about 50%, so a coin flip. Would you bet me $100 that one toss of a coin would come up heads? Maybe, because you'd have a 50% shot. Would you bet me $100 that you could toss a coin 12 times and all 12 would be heads? Probably not, because you know that while each individual toss is mostly even, the odds of getting the same result over a longer term is much less likely.
Click to expand...

Yeah I get it. Just at first glance 13% seemed ridiculous.
 
tboyer said:
The Temple % is a head scratcher. My guess is the computer views that as an away game --- but being located in Philly it won't be like an away game. The crowd will probably be majority PSU.
Click to expand...

I have attended many of the Temple/PSU games in Philadelphia. The crowds have generally been about 70-80% PSU and close to sellouts. I hope Franklin convinces the kids this Temple team is the real deal. If he can do that and the team is focused we win "comfortably". If "we" look at them as, well, Temple we could be in trouble. The the "screeching" Owl they use !!!!
 
BBHorn86 said:
You might very well be correct with this prediction. But I'm of the belief that PSU and many of the unknowns people mention (OL, special teams, etc.) land with the better-than-forecast outcome. Our defense could be special. I don't think it will be a high scoring game but I'll take out defense to shut Temple down every day when it counts.

PSU 28-17.
Click to expand...

temple is picked to finish behind cincy and ucf, both pretty fair FB teams. so all i can say is we played a very banged up team that returns every starter on defense and 8 starters on offense (AND both their kickers). they will be healthy, they will be ready. we damned well better be, too.

i'm going to be interested in how the home field aspect plays out.. i have the sense from conversations with some philly folks that their fans are going to be there... they smell history making. if they show, it will be an added boost to their already sky high 'tude.

at home, hack went 12-26 with 2 picks against a busted up D. their qb threw 4 picks trying to do it all by himself... this year, he'll have help. sticking with my prediction.
 
N&B4PSU said:
temple is picked to finish behind cincy and ucf, both pretty fair FB teams. so all i can say is we played a very banged up team that returns every starter on defense and 8 starters on offense (AND both their kickers). they will be healthy, they will be ready. we damned well better be, too.

i'm going to be interested in how the home field aspect plays out.. i have the sense from conversations with some philly folks that their fans are going to be there... they smell history making. if they show, it will be an added boost to their already sky high 'tude.

at home, hack went 12-26 with 2 picks against a busted up D. their qb threw 4 picks trying to do it all by himself... this year, he'll have help. sticking with my prediction.
Click to expand...

Emotion is huge - I get it - but it also can be dangerous especially if you are TOO hyped. Temple would be better to approach the game as if it were any other opponent and not attempt to be more than they are. I'd be more concerned about a team with quiet confidence and a level head than one jacked up on emotion and big talk. Those things can lead to a fast start much like PSU vs. Ohio in the opener back in 2012 but once the emotions are spent execution and talent take over.

I think we are both expecting PSU to win I just think it will only be close for the first half.
 
BBHorn86 said:
Emotion is huge - I get it - but it also can be dangerous especially if you are TOO hyped. Temple would be better to approach the game as if it were any other opponent and not attempt to be more than they are. I'd be more concerned about a team with quiet confidence and a level head than one jacked up on emotion and big talk. Those things can lead to a fast start much like PSU vs. Ohio in the opener back in 2012 but once the emotions are spent execution and talent take over.

I think we are both expecting PSU to win I just think it will only be close for the first half.
Click to expand...

Absolutely agree with you. but I'm not suggesting temple will come out foaming at the mouth.. they have a pretty good coach and a lot of experienced players determined to end the penn state curse. I hope we get this done early and win comfortably.. would be an ideal way to start the season. Don't think it happens, but i'd prefer it :)
 
michnittlion said:
If one believes the per-game percentages ESPN showed on-air today, running a Monte Carlo simulation 30,000 times, the results. Take it for what it is worth:

Odds of 12 wins: 0.14%
Odds of 11+ wins: 2.20%
Odds of 10+ wins: 13.2%
Odds of 9+ wins: 36.8%
Odds of 8+ wins: 65.6%
Odds of 7+ wins: 86.6%
Odds of 6+ wins: 96.5%

Odds of 8 conference wins: 0.34%
Odds of 7+ conference wins: 5.05%
Odds of 6+ conference wins: 25.5%
Odds of 5+ conference wins: 58.2%
Odds of 4+ conference wins: 85.1%
Odds of 3+ conference wins: 96.9%
Odds of 2+ conference wins: 99.6%

Total wins: Average = 8.00, Standard Deviation = 1.36
Conference wins: Average = 4.71, Standard Deviation = 1.18
OOC wins: Average = 3.29, Standard Deviation = 0.67

CLl7JfUUcAAJ58J.jpg:large
Click to expand...

Instead of using the percentages from ESPN to do 30,000 simulations and come up with an estimate of the number of wins, you could have just used the percentages directly without doing simulations. The percentages could be used to compute a probability distribution. Your 30,000 simulations estimate that probability distribution. Of course, the usefulness of either depends on how good ESPNs percentages are in the first place.

ETA: In fact, for curiosity sake I did it and here are the two results side by side. Hopefully these line up correctly and are readable.

Your 30 K sims Actual Prob
Prob of 12 wins 0.14% 0.14%
Prob of 11+ wins 2.20% 2.29%
Prob of 10+ wins 13.2% 12.9%
Prob of 9+ wins 36.8% 36.7%
Prob of 8+ wins 65.6% 65.5%
Prob of 7+ wins 86.6% 86.7%
Prob of 6+ wins 96.5% 96.4%
 
Last edited:
  • Like
Reactions: michnittlion and N&B4PSU
Temple has a flavour for Penn State, and they want to make good to their fan base ( what fan base ? ) they can play big time football. Ain't gonna happen. Not this year...

N i t t a n y A m e r i c a
 
Op2 said:
Instead of using the percentages from ESPN to do 30,000 simulations and come up with an estimate of the number of wins, you could have just used the percentages directly without doing simulations. The percentages could be used to compute a probability distribution. Your 30,000 simulations estimate that probability distribution. Of course, the usefulness of either depends on how good ESPNs percentages are in the first place.

ETA: In fact, for curiosity sake I did it and here are the two results side by side. Hopefully these line up correctly and are readable.

Your 30 K sims Actual Prob
Prob of 12 wins 0.14% 0.14%
Prob of 11+ wins 2.20% 2.29%
Prob of 10+ wins 13.2% 12.9%
Prob of 9+ wins 36.8% 36.7%
Prob of 8+ wins 65.6% 65.5%
Prob of 7+ wins 86.6% 86.7%
Prob of 6+ wins 96.5% 96.4%
Click to expand...

One could have done that too -- you are absolutely correct right. Your method is exact and precise, mine is admittedly a close estimation. There are only 4,096 (2^12) different possible win-loss orders (e.g., WWWLLWLWWWLL) that a 12-game season can follow, so that wouldn't have been too complex to figure either.

I'm admittedly biased toward Monte Carlo because it's fairly easy to set up in Excel: honestly, it only took me about 3 minutes to do it this way. 30,000 rows of random numbers, a few IF and SUMIF statements, and there we are.
 
You must log in or register to reply here.
ADVERTISEMENT

Latest posts

ADVERTISEMENT
Top Bottom
Back