I mean…we’re about to find out, no?

I see mediocrity in New England’s future

- Get ripped (emotionally)
- Read 12 books (but maybe not because maybe I don't feel like reading today or any day)
- Love the hell out of my wife and kids

Bravo!

Nope - the Christmas party is the central plot point. Also the redemption of the marriage through adversity is certainly in keeping with the Christmas spirit.

When Lemonnier talks, I assume the opposite of whatever he says is true.

Correct - and SMU even chose to go to a tougher conference! They literally worked to get the toughest schedule they had access to!

be fed into the models, there is no reason ESPN, as a journalistic enterprise, couldn't find more nuance in it's reporting.

... he can put more resources and practice time into special teams. He can spend more cap space on a punter. He can ensure more top tier players on ST and so on. General models are nice and have obviously shifted the game towards more aggression on these calls. But with so much more data that can

Maybe - but the way you frame it continues to indicate a misunderstanding of my main point. Saying MT 'gets away with' poor 4th down decision making implies that he doesn't intentionally prep his team for the specific situations that make punting lead to more successful outcomes. For instance...

I'm so glad we have a chance to see Mike Tomlin coach. Absolutely a master class - I can't wait for next week!

...Historical data should absolutely be the foundation, but we could weight current game/season data at say 10% to help with those razor-thin 51-49 decisions.

Mike Tomlin rarely goes for it on 4th down, yet he continuously has winning seasons - what does he know that the ESPN model doesn't?

Good points. I understand the caution about small sample sizes, but I think there's room for a more nuanced approach. Even if coaching decisions average out over time, we can provide more accurate single-instance recommendations by incorporating additional parameters without overfitting...

Coaches (rightly) are taking into account much more information intuitively(?) to make these decisions. So when they go against the ESPN model recommendations, if it succeeds (and they win the game) it undermines analytics. If it fails, it unfairly places blame on the coach.

Found this article: www.espn.com/nfl/story/_/...

The issue I have is that the ESPN model does not take into account individual team stats on the year or even the game. There is so much data on this that it's impossible to have more parameters than data points.

Okay genuinely curious: what parameters go into the model for calculating these recommendations? Score, field position, distance to first, and time left certainly. What about offensive and defensive efficiency? How well has the team been on 3rd down? Etc?

Also, if you are looking at seeding in CFP, a bigger win might give them a home game if they lose to Oregon.

Interesting stats. I’m also curious about analytics - namely what models teams is for decision making and do they suffer from over-fitting, too few parameters, etc

Really sucking the joy out of watching your team score.