pats lost by 7 in a game where they were favored by 4 despite the niners having a couple ridic punt returns and fumbling 4 snaps

"what if the niners had really, really, really ****ed themselves on plays that had absolutely nothing to do with the patriots instead of just kinda ****ing themselves?" is not much of an argument if you're trying to show that the patriots are the better team

nswf

Stay classy, America.

Should probably put spoiler tags and an nsfw warning around those tweets.

I still think the Pats are the better team, but they sure didn't show it tonight.

Earlier tonight, the Pats went for it on 4th and 2 with over 2:20 left and 2 TOs from their own 12 yd line.

You guys are crazy if you think they should have punted there. That's a clear go for it in a tie game in the 1st quarter. Down 7 facing the best running team in the NFL, punting would be absolutely horrific.

lol

Show your work tho

no u

I want to be proven wrong! I would be very happy if someone worked out a full argument/maths behind this. I don't care which side is right here. Im not smart enough obv

If I find the math for you why you should go for it there in the 1st quarter, is that good enough or do I also need to explain to you why it's better to go for 4th downs when you're down 7 with 2:00 left than when you're tied at the start of the game.

NO i think going for it is obv the right decision. I'm with you on this one. I wanted to be proven wrong that a coach has never gone for it when punting was superior.

As an aside, someone needs to get some artificial neural networks on football analytics. Brian Burke's linear regressions are obviously not optimal.

I would LOVE to see this

I thought Iggy was just making fun of the Brian Burke cult following with that comment, I hope he doesn't think it is a clear go for it in the first quarter. But it is probably really close!

Here's a link to the 4th down study on advancednflstats. I think the EP model is really good, and gives a very accurate picture of what's correct in which part of the field.

Part 1 (if you want to read the whole article):
http://www.advancednflstats.com/2009...dy-part-1.html

Part 4 (section with chart by yard line)
http://www.advancednflstats.com/2009...dy-part-4.html


What are you referring to here? Are you referring to the end-game calculator or the original 4th down study? If you're talking about the 4th-down calculator with the WP model, then yeah I think that's definitely quite sloppy. He's working with way too small of a dataset, and it doesn't end up with reliable data at all. But the original 4th-down study with the EP model was very good IMO, and involved such a large data-set that any problems would just be minor stuff around the edges, several orders of magnitude smaller than the stuff that coaches do wrong on a regular basis. If you have issues with that, I'd be interested to hear what they were.

Well, the study says it's close so "clear" is probably too strong of wording, but given how much offense has improved with the recent rules changes since the years used for that study, I think if anything their estimations are likely to be on the conservative side. I would say, at the very least, it's >90% that it's going to be the right decision to go for it.

I don't have issues with Burke's models, I think they're great. And we're pretty lucky as fans to even have them! In general, artificial neural network models do better than regression models for stuff like this is all I was getting at. And we could probably input a lot more variables with one. No shot I am gunna make it anytime soon though.

Also, in before Agame points out that in Burke's model it says to go for it on 4th and 2 at your own 20 but punt at your own 30.

Actually, you're supposed to go for it whenever you're on the line so 4th and 2's always a go for it unless you're inside your own 10. It's 4th and 3 where the inconsistency lies.

LOL @ 4th and 2 inside your own 10. Why does he even extend the line that far back. Go for it on 4th and 2 from your own 5? I want to see his data!

Right in the feelings.

the study says it's close for a league average offense facing a league average defense. the niners pass d is good but the patriots pass o is ****ing amazing so i doubt going for it is all that close in reality.

Dude, calm down. My post "all the good posters said to punt" was obviously tongue-in-cheek. Sheesh.

lollllllllll

Pats should have gone for it, especially because scoring a TD with less than 2 minutes left and 80+ yards to drive is very tough. On the other hand i'd bet Burke is overstating how much league average teams should go for it on 4th down, as others have pointed out there are lots of problems with his model.

Agame,

If the Pats punt and stop the 49ers and the 49ers punt back what do you think the Pats average field position is?
How often do the Pats convert on 4th and 2?
How often do the Pats force the 49ers to punt?

Is Burke's formula going to take into account that the typical possession for the 9ers there looks like this:

Run left for 1 yard
Run somewhere for -2 yards
Fumble snap, lose 6 yards
Punt - ???

Not sure how well historical data inputs tell the story there.