2025 College Football Playoffs

[GATTACA!]

STOP THE COUNT!!

 

[SportsingHard]

All I care about is, what are the chances of ND winning the last five games. Here's the best ChatGPT could come up with:

Alright, SportsingHard — here’s a model-based take (SP+/FPI-style logic, ~2.5–3 pts HFA baked in). Use these as placeholders until books post moneylines.

Game	Site	Est. win %
at Boston College (Nov 1)	Away	80%
vs Navy (Nov 8)	Home	85%
at Pitt (Nov 15)	Away	60%
vs Syracuse (Nov 22)	Home	70%
at Stanford (Nov 29)	Away	65%

Math you asked for​

• Chance to win all five = 0.80 × 0.85 × 0.60 × 0.70 × 0.65 = 18.6%
• Expected wins over these five = 0.80 + 0.85 + 0.60 + 0.70 + 0.65 = 3.6 wins

I'm gonna give us better than a 65% chance of beating Stanford, but on the whole, I find this analysis depressingly realistic.

 

[InKellyWeTrust]

All I care about is, what are the chances of ND winning the last five games. Here's the best ChatGPT could come up with:

Alright, SportsingHard — here’s a model-based take (SP+/FPI-style logic, ~2.5–3 pts HFA baked in). Use these as placeholders until books post moneylines.

Game	Site	Est. win %
at Boston College (Nov 1)	Away	80%
vs Navy (Nov 8)	Home	85%
at Pitt (Nov 15)	Away	60%
vs Syracuse (Nov 22)	Home	70%
at Stanford (Nov 29)	Away	65%

Math you asked for​

• Chance to win all five = 0.80 × 0.85 × 0.60 × 0.70 × 0.65 = 18.6%
• Expected wins over these five = 0.80 + 0.85 + 0.60 + 0.70 + 0.65 = 3.6 wins

I'm gonna give us better than a 65% chance of beating Stanford, but on the whole, I find this analysis depressingly realistic.

I'd take those odds all day. In fact, I'd wager we beat every team left on the schedule by 2+ scores.

 

[Dale]

That is not how likelihood odds work lol. Thank you ChatGPT for the “model”

 

[SportsingHard]

That is not how likelihood odds work lol. Thank you ChatGPT for the “model”

Likelihood odds are all bullshit. I just wanted to see if the numbers it arrived at were similar to mine. It's that 18.6% number that's important to me, and I don't believe it's too far off.

 

[SportsingHard]

That is not how likelihood odds work lol. Thank you ChatGPT for the “model”

Out of curiosity, since essentially zero thought process was communicated here in terms of computing single-game likelihoods of victory, what exactly are you criticizing? What is "not how likelihood odds work"?

 

[stlnd01]

I am not a mathematician but the likelihood that we win all five games is vastly higher than 18.6 percent.

 

[Dale]

Out of curiosity, since essentially zero thought process was communicated here in terms of computing single-game likelihoods of victory, what exactly are you criticizing? What is "not how likelihood odds work"?

I think the odds you used for the games as well as the process are both silly. ChatGPT is using a simplistic “independent events just multiply them together” which I don’t find useful here. Can easily point to Fanduel odds to make the CFP being like 75% implied odds. The odds ND goes 10 - 2 are likely even higher than that then according to oddsmakers. As far as the games, again odds likely significantly higher of a win in every game posted.

 

[NumbersGuy0520]

If the individual win probabilities are correct, the 18.6% probability of going 5-0 is correct.

The underlying issue is GPT’s individual game probabilities are way short. A win probability of 80% in CFB typically corresponds to a spread of -11 or so. A win probability of 60% is like -3 or -4.

There is no way we will only be -11 @ BC or -3 at Pitt lol. If we are, bet the mortgage on us to cover.

 

[SportsingHard]

I think the odds you used for the games as well as the process are both silly. ChatGPT is using a simplistic “independent events just multiply them together” which I don’t find useful here. Can easily point to Fanduel odds to make the CFP being like 75% implied odds. The odds ND goes 10 - 2 are likely even higher than that then according to oddsmakers. As far as the games, again odds likely significantly higher of a win in every game posted.

Those aren't odds I used. Those came from ChatGPT. In regards to the A * B * C * D * E = F part, that is objectively correct, mathematically, so I'm not understanding the problem there. You don't see the games as independent events?

 

[Bane]

Oh, we talkin math in here. I'll see myself out.

 

[Dale]

Those aren't odds I used. Those came from ChatGPT. In regards to the A * B * C * D * E = F part, that is objectively correct, mathematically, so I'm not understanding the problem there. You don't see the games as independent events?

There is 0.0% chance the odds of ND going undefeated down the stretch is 18% is what I know.

 

[BleedingGold]

Hosting Miami or Texas Tech in ND during the 1st round would be just to much fun. Supposed to be a cold, wet winter.

 

[Armyirish47]

There is 0.0% chance the odds of ND going undefeated down the stretch is 18% is what I know.

Absolutely, it’s 50%. It happens or it don’t.

 

[GATTACA!]

I think it's funny how much FPI must hate Indiana. I ran the sim several times and I don't think I saw them be predicted to win their first game a single time.

 

[IrishColts17]

Here's what you're looking for. 74% chance to go 10-2 according to Kelley Ford.

 

[SportsingHard]

If the individual win probabilities are correct, the 18.6% probability of going 5-0 is correct.

The underlying issue is GPT’s individual game probabilities are way short. A win probability of 80% in CFB typically corresponds to a spread of -11 or so. A win probability of 60% is like -3 or -4.

There is no way we will only be -11 @ BC or -3 at Pitt lol. If we are, bet the mortgage on us to cover.

I see what you're saying. I know math more than I know Vegas, but it seems to me fans (not gambling pros) consistently underrate the likelihood of any given upset to their own favorite team. What kind of likelihoods would you assign to each of those five games?

 

[SportsingHard]

Here's what you're looking for. 74% chance to go 10-2 according to Kelley Ford.

View attachment 3060725

Wow. Judging by overall posting patterns, I'm more of an optimist than 98% of this board, but 97-98% chance of victory in 4 of our last 5 games? In a sport where injuries, officiating, and random bounces account for so much? That strikes me as crazy talk.

 

[NumbersGuy0520]

I see what you're saying. I know math more than I know Vegas, but it seems to me fans (not gambling pros) consistently underrate the likelihood of any given upset to their own favorite team. What kind of likelihoods would you assign to each of those five games?

As of right now, I have us as close to 20+ point favorites in every game except Pitt.

20+ point favorites win outright like 95% of the time (+- a few percentage points, as it’s hard to estimate these tail-end events). It’s just reality. There is always the potential to lose these games (Penn State did it twice in a row this year), but it’s incredibly rare.

 

[stlnd01]

As of right now, I have us as close to 20+ point favorites in every game except Pitt.

20+ point favorites win outright like 95% of the time (+- a few percentage points, as it’s hard to estimate these tail-end events). It’s just reality. There is always the potential to lose these games (Penn State did it twice in a row this year), but it’s incredibly rare.

Yep. We could lose one, because it's college football and we lost to Northern Illinois last year and Marshall a couple years before. But aside from maybe Pitt we should be huge favorites in every game, and even against Pitt we are vastly more talented - how many Pitt players would start for us? - and if we're locked in, as we largely have been over two years now, we win that one by multi scores.

The odds we drop one are certainly higher than 2 percent, but they're closer to 20 percent than they are to 80 percent.

 

[SportsingHard]

As of right now, I have us as close to 20+ point favorites in every game except Pitt.

20+ point favorites win outright like 95% of the time (+- a few percentage points, as it’s hard to estimate these tail-end events). It’s just reality. There is always the potential to lose these games (Penn State did it twice in a row this year), but it’s incredibly rare.

Thanks for your input, but when I look at an away game vs Sagarin's #31 team that beat an FSU team that beat Alabama and nearly beat Miami (Pitt) followed by a game against a team that beat Clemson (Syracuse), I don't see that as the same as consecutive weeks against a team lost to New Mexico by 25 at home (UCLA) and a team that lost to Tulane by 20 (Northwestern.) Nevermind the fact that Navy is an undefeated team with an unconventional style that nearly beat ND again under Freeman/Golden. We've got some very losable games ahead.

 

[InKellyWeTrust]

Thanks for your input, but when I look at an away game vs Sagarin's #31 team that beat an FSU team that beat Alabama and nearly beat Miami (Pitt) followed by a game against a team that beat Clemson (Syracuse), I don't see that as the same as consecutive weeks against a team lost to New Mexico by 25 at home (UCLA) and a team that lost to Tulane by 20 (Northwestern.) Nevermind the fact that Navy is an undefeated team with an unconventional style that nearly beat ND again under Freeman/Golden. We've got some very losable games ahead.

Lots of words. ND is going 5-0.

 

[SportsingHard]

Lots of words. ND is going 5-0.

 

[MPClinton22]

Thanks for your input, but when I look at an away game vs Sagarin's #31 team that beat an FSU team that beat Alabama and nearly beat Miami (Pitt) followed by a game against a team that beat Clemson (Syracuse), I don't see that as the same as consecutive weeks against a team lost to New Mexico by 25 at home (UCLA) and a team that lost to Tulane by 20 (Northwestern.) Nevermind the fact that Navy is an undefeated team with an unconventional style that nearly beat ND again under Freeman/Golden. We've got some very losable games ahead.

FSU and Clemson are terrible, so wins against them are irrelevant. Syracuse lost Angeli and is also now terrible. Don't disagree Pitt is definitely the toughest game we've got left, but they're not actually a good team.

Semi-off topic, but that Alabama loss to FSU is truly unbelievable. Still have no clue how that happened.

 

[BleedingGold]

Can’t deny I’d be disappointed to lose to navy, cuse, Stanford, or BC. Wouldn’t be the first time we laid an egg against an inferior team.

Pitt is hard to place. They have that win over FSU, which again is hard to figure out as a team.

 

[SportsingHard]

FSU and Clemson are terrible, so wins against them are irrelevant. Syracuse lost Angeli and is also now terrible. Don't disagree Pitt is definitely the toughest game we've got left, but they're not actually a good team.

Semi-off topic, but that Alabama loss to FSU is truly unbelievable. Still have no clue how that happened.

Well, like I said, FSU nearly beat Miami, so it's not like the Alabama game was their only sign of life. Dokter Entropy has them as the #25 team in the country.

 

[MPClinton22]

Well, like I said, FSU nearly beat Miami, so it's not like the Alabama game was their only sign of life. Dokter Entropy has them as the #25 team in the country.

FSU is also a team that notoriously quits when the going gets tough, and this year is not looking any different. By the time Pitt played them, they'd already lost 2 straight and the spiral had fully begun. Now allegedly Pitt has really turned around since switching QBs - maybe that's true. But they did also lose to an absolutely pitiful WVU team, so I also really don't know how to assess them. But I do know if ND is as good as I think/hope they are, they should beat Pitt by 2+ scores.

 

[NumbersGuy0520]

Can’t overlook the others, but @Pitt is by far the spookiest game left on the schedule.

As things stand right now, I think the line would be close to where USC opened. Definitely not a gimme.

 

[SWirishfan]

There is 0.0% chance the odds of ND going undefeated down the stretch is 18% is what I know.

That is emotion.

The calculation is correct here as each game is an independent event.
The probabilities, like 80% win prob for BC, are way low. That skews the final total probability.

 

[Jeffwantpoppers]

Well, like I said, FSU nearly beat Miami, so it's not like the Alabama game was their only sign of life. Dokter Entropy has them as the #25 team in the country.

Wait FSU didn’t nearly beat Miami. That game was domination until Miami stopped playing. FSU didn’t belong on the same field.