Others have said we need ISU to have a good record....

Ree4 said:
Hahaha where the hell are you coming up with this math?
Click to expand...

Iowa went undefeated and won the conference in 1921 and 1922. They've never gone undefeated and won the conference again. They have played football for 130 years. 2 / 130 = .015 = P(A).

The CFP has been played five times, so whoops on the .25, it's actually .20. In those 5 times, 3 undefeated P5/ND teams has occurred once, so 1 / 5 = .20 = P(B).

1 - loss Big ten champs have made it in 2 / 5 times, so that's actually 2 / 5 = .4 = P(C).

The probability of any two independent events is equal to the probability of event A and the probability of event B occurring. You get this by multiplying P(A) by P(B). P(A) AND P(B) = P(A) x P(B).

Multiplication is associative. This means that X x Y x Z = (X x Y) x Z = X x (Y x Z). This means that the probability of 3 independent events occurring simultaneously is P(A) AND P(B) AND P(C) is P(A) x P(B) x P(C). Thus, to get the chance that op asked fo we multiply .15 x .2 x .4 = .012.

The percentage formula is X x 100. Thus a 1.2% chance of Iowa being undefeated and being a Big Ten Champ; three other P5/ND being undefeated; and a Big Ten Team making the playoffs.

We can get even more detailed, since we can independently calculate the odds of Iowa going undefeated this year with partial knowledge of being 4 - 0 and how the seasons ended before. Iowa went undefeated twice, and didn't a bunch of other times (6 I think, it would take some digging). Let's assume 8 times, so 2 / 8 = .25. That gets us to a 2% chance of going undefeated and winning the playoffs.

We could look at all the teams in our conference and figure out the dependent probability of going undefeated, since there can be only one, and get a really accurate estimate, but that's just too much work. I'll leave it at 1.5%.

@Fryowa.

I do data analytics for a living.
 
Last edited:
cincyhawk said:
Iowa went undefeated and won the conference in 1921 and 1922. They've never gone undefeated and won the conference again. They have played football for 130 years. 2 / 130 = .015 = P(A).

The CFP has been played five times, so whoops on the .25, it's actually .20. In those 5 times, 3 undefeated P5/ND teams has occurred once, so 1 / 5 = .20 = P(B).

1 - loss Big ten champs have made it in 2 / 5 times, so that's actually 2 / 5 = .4 = P(C).

The probability of any two independent events is equal to the probability of event A and the probability of event B occurring. You get this by multiplying P(A) by P(B). P(A) AND P(B) = P(A) x P(B).

Multiplication is associative. This means that X x Y x Z = (X x Y) x Z = X x (Y x Z). This means that the probability of 3 independent events occurring simultaneously is P(A) AND P(B) AND P(C) is P(A) x P(B) x P(C). Thus, to get the chance that op asked fo we multiply .15 x .2 x .4 = .012.

The percentage formula is X x 100. Thus a 1.2% chance of Iowa being undefeated and being a Big Ten Champ; three other P5/ND being undefeated; and a Big Ten Team making the playoffs.

We can get even more detailed, since we can independently calculate the odds of Iowa going undefeated this year with partial knowledge of being 4 - 0 and how the seasons ended before. Iowa went undefeated twice, and didn't a bunch of other times (6 I think, it would take some digging). Let's assume 8 times, so 2 / 8 = .25. That gets us to a 2% chance of going undefeated and winning the playoffs.

We could look at all the teams in our conference and figure out the dependent probability of going undefeated, since there can be only one, and get a really accurate estimate, but that's just too much work. I'll leave it at 1.5%.

@Fryowa.

I do data analytics for a living.
Click to expand...
Sorry, I should have said numbers, the math was there but the numbers were off. I thought it was a weird answer to the question that was posed and didn't quite answer it. 1 out of 5 should've answered it. I've spent a lot of time in analytics myself, stopped just short of a math minor in college and can appreciate the thought you've put into this.

Unfortunately I feel that stats/analytics/probabilities like this are terribly flawed. This assumes that the probabilities don't change - in the way flipping a coin is ~ 50/50. However, drawing the probability of Iowa going undefeated this year from two occasions that happened almost 100 years ago is incredibly flawed. Iowa was 7-0 in 21 & 22. To go undefeated this year and to win the championship would require us to be 15-0. That alone is a monumental difference. Scholarship limits changing in the 90s, the increase of parity, etc.

The second event you mention is that only 1 out of 5 years there was 3 or more undefeated teams. That is an incredibly small sample size (but it's all we have). While it's fun to think about and analyze, any conclusion drawn would be flawed because the entire thing is based on a flawed assumption. When you multiply flawed numbers by other flawed numbers it only exacerbates the margin of error.

Like I said, I can appreciate the thought you put into this, you used the data that existed, but I'd argue that any confidence in the results would be minimal.
 
Last edited:
cincyhawk said:
Iowa went undefeated and won the conference in 1921 and 1922. They've never gone undefeated and won the conference again. They have played football for 130 years. 2 / 130 = .015 = P(A).

The CFP has been played five times, so whoops on the .25, it's actually .20. In those 5 times, 3 undefeated P5/ND teams has occurred once, so 1 / 5 = .20 = P(B).

1 - loss Big ten champs have made it in 2 / 5 times, so that's actually 2 / 5 = .4 = P(C).

The probability of any two independent events is equal to the probability of event A and the probability of event B occurring. You get this by multiplying P(A) by P(B). P(A) AND P(B) = P(A) x P(B).

Multiplication is associative. This means that X x Y x Z = (X x Y) x Z = X x (Y x Z). This means that the probability of 3 independent events occurring simultaneously is P(A) AND P(B) AND P(C) is P(A) x P(B) x P(C). Thus, to get the chance that op asked fo we multiply .15 x .2 x .4 = .012.

The percentage formula is X x 100. Thus a 1.2% chance of Iowa being undefeated and being a Big Ten Champ; three other P5/ND being undefeated; and a Big Ten Team making the playoffs.

We can get even more detailed, since we can independently calculate the odds of Iowa going undefeated this year with partial knowledge of being 4 - 0 and how the seasons ended before. Iowa went undefeated twice, and didn't a bunch of other times (6 I think, it would take some digging). Let's assume 8 times, so 2 / 8 = .25. That gets us to a 2% chance of going undefeated and winning the playoffs.

We could look at all the teams in our conference and figure out the dependent probability of going undefeated, since there can be only one, and get a really accurate estimate, but that's just too much work. I'll leave it at 1.5%.

@Fryowa.

I do data analytics for a living.
Click to expand...


I was going to tell you to go out and get some fresh air before reading you do analytics for a living. Now fresh air wouldn't matter. That said none of us know analytics so we have no idea WTF you're talking about. Carry on.
 
kicker22 said:
I was going to tell you to go out and get some fresh air before reading you do analytics for a living. Now fresh air wouldn't matter. That said none of us know analytics so we have no idea WTF you're talking about. Carry on.
Click to expand...
Eh, it's incredibly simple what he's doing honestly. He's looking at three separate stats. He's taking the number of times something occurred and dividing by the number of times each could have happened in total.

Number of times Iowa was "undefeated champs" : 2 times in 130 years
Number of times there were 3 undefeated teams in playoff: 1 time in 5 years
Number of times there were 1 loss B10 teams in playoff: 2 times in 5 years.
2/130 x 1/5 x 2/5
 
Ree4 said:
Sorry, I should have said numbers, the math was there but the numbers were off. I thought it was a weird answer to the question that was posed and didn't quite answer it. 1 out of 5 should've answered it. I've spent a lot of time in analytics myself, stopped just short of a math minor in college and can appreciate the thought you've put into this.

Unfortunately I feel that stats/analytics/probabilities like this are terribly flawed. This assumes that the probabilities don't change - in the way flipping a coin is ~ 50/50. However, drawing the probability of Iowa going undefeated this year from two occasions that happened almost 100 years ago is incredibly flawed. Iowa was 7-0 in 21 & 22. To go undefeated this year and to win the championship would require us to be 15-0. That alone is a monumental difference. Scholarship limits changing in the 90s, the increase of parity, etc.

The second event you mention is that only 1 out of 5 years there was 3 or more undefeated teams. That is an incredibly small sample size (but it's all we have). While it's fun to think about and analyze, any conclusion drawn would be flawed because the entire thing is based on a flawed assumption. When you multiply flawed numbers by other flawed numbers it only exacerbates the margin of error.

Like I said, I can appreciate the thought you put into this, you used the data that existed, but I'd argue that any confidence in the results would be minimal.
Click to expand...

Oh, I have no confidence in the results. There's not enough data. FiveThirtyEight has a simulation. If Clemson, Alabama,Oklahoma, and n ND win out, Iowa has a 7% chance to make it in: https://projects.fivethirtyeight.com/2019-college-football-predictions/. Wisconsin and Ohio State are more likely to make it in that scenario (presumably one will have beaten Iowa).
 
cincyhawk said:
Well, Iowa going undefeated and winning the championship has happened twice in 130 years, so that's .015 probability x .25 for >3 undefeated P5 champions + ND x .5 for a 1-loss BT team getting in = 0.1% chance of that happening. An undefeated BT team has never made it to the CFP, nor been left out, but several 1-loss Big Ten Teams have been left out.

If we take Iowa out of it, a 1-loss BTT team has been left out .25 x .5 x .25 = a 3.01% chance of it happening.

So 0.1% (for Iowa going undefeated and 3 other unbeaten powers and a 1 - loss Power) and 3.01% chance that any 1-loss Big champion misses the playoffs. Not big odds, certainly.
Click to expand...

I understand where you are coming from your work looks more like the Drake equation.
 
kicker22 said:
I was going to tell you to go out and get some fresh air before reading you do analytics for a living. Now fresh air wouldn't matter. That said none of us know analytics so we have no idea WTF you're talking about. Carry on.
Click to expand...

It's just simple probability, as @Ree4 pointed out. Someone asked for numbers, so I gave it to them. Took me like four minutes.

To get a better estimate, you would have to gather stats about all the current contenders and their historical records and calculate the dependent probabilities that they'd make it through their schedule undefeated. The math is simple, but the legwork is harder.

The model is fundamentally flawed, though - you can't really compare teams from year to year as there are too many variables. Every team every year is different.
 
I've seen a lot of thread derailments over the years. Maybe even caused a few...

But this is the first time I've seen a thread derailed by math!
 
Pure math is one of my favorite subjects. It's pretty cut and dry, unlike most things in this universe, there's really no room for disagreement unless you start getting into theoretical stuff, which I never did.

How to use math to try to explain the real world is a different beast altogether, just wayyy too many variables 99% of which we may never understand or be able to quantify.
 
Ree4 said:
Eh, it's incredibly simple what he's doing honestly. He's looking at three separate stats. He's taking the number of times something occurred and dividing by the number of times each could have happened in total.

Number of times Iowa was "undefeated champs" : 2 times in 130 years
Number of times there were 3 undefeated teams in playoff: 1 time in 5 years
Number of times there were 1 loss B10 teams in playoff: 2 times in 5 years.
2/130 x 1/5 x 2/5
Click to expand...

Analytics for Dummies...Thanks for the help. Makes much more sense now.
 
cincyhawk said:
IMHO, Iowa's not going to the playoffs this or any year. The real unwritten rule is, unless you are USC, Texas, Alabama, OSU, LSU, Florida, Gerogia, Auburn, ND, or Oklahoma you aren't ever going to be in. It's just not happening.
Click to expand...

Other than the fact that MSU, Oregon and Washington have already been in the playoffs.

2015 Iowa was one play from the playoffs.

The CFP is realistic. Unlikely, but realistic.

Your premise is flat out wrong.
 
The 2017 Outback Bowl had the Iowa sideline in the shade most of the game, and FLA baking in the sun.

We lost 30-3.
 
cincyhawk said:
I do data analytics for a living.
Click to expand...

That is scary.

Football teams are not actuarial tables or a deck of cards. You can't use 130 years of Hawkeye data to project future playoff chances.

Holy shit it is scary that you get paid for this dribble.
 
ICHawk24 said:
Two SEC teams have made it precisely once out of five years.

Wrong AGAIN for the hat trick!!!! Impressive.
Click to expand...

Dude, it's highly unlikely Iowa makes the playoffs. Be prepared to hear the arguments, because they're coming if we go undefeated.
 
You must log in or register to reply here.

Latest posts

Share this page

Top