When the Texans play their "A" game they will beat just about anyone else in the NFL on their "A" game. So far this year they've had too many "F" game efforts. The Broncos and Saints might be the only two teams who would have a decent chance to beat the Texans "A" game. Maybe Green Bay, too. GOAT, Brees, and Rodgers are just incredible at their best.

You are right but don't act like real life does not provide that platform too. Any time money is involved it does imo. I mean the most powerful force in evolution is based around that exact concept (natural selection).

It just may happen faster on the Internet in this case.

lol @ this noise. Both the Pats and the Packers destroyed the Texans last year. And Texans played solidly. Rodgers just threw 6 TD passes that were basically not defensible - one of the sickest QB/WR performances I've ever seen. And then Brady and Co. made the defense look foolish and unprepared with his quick offense.

This is not the same team as in 2011.

O-line is worse, Schaub is worse, Foster is worse, Andre is worse (older, really), Daniels is worse, linebackers outside Cushing are absolutely awful, and coaching is worse.

A few things are better. Secondary is probably a little better, and DeAndre Hopkins is better than Kevin Walter. Down linemen a little better.

So this caused me to start looking at random pointless stuff on profootballreference for fun..Since 1970, there have been 70 games in the regular season where a team had 10 rushing attempts or less in a game, and those teams combined record is 2-68.

The two wins:

Jim Sorgi led Colts (with OC Peyton Manning?)

http://www.pro-football-reference.co...0601010clt.htm

18-1 Patriots

http://www.pro-football-reference.co...0712090nwe.htm

My favorite loss from that list upon first glance

http://www.pro-football-reference.co...9909120cle.htm

Cushing was out for that Pats game and maybe the Green Bay game? Their defense seems to fall apart when he's not on the field.

40 total yards, jesus christ

25 plays in a full game is maybe even more amazing.

The NFL is a billion dollar a year industry. How many teams even have an analysis department or even just 1 guy crunching meaningful data?

Their realm is rife with words and phrases like "establish", "trust", "pin our ears back", "to a man", "heart", "guts", "courage", "instinct", etc.

Hell, the only number they routinely use is mis-applied: "We gave it 110%"

I think it's just the nature of the beast. To get 300 pound dumb guys to play hard you have to use emotional words to fire them up. You can't show Wilfork what his WPA is on rushing plays to get him fired up. You have to say things like "move the pile" and "get after 'im"

I think it's just super hard to introduce this type of language (statistical data) into an industry that doesn't operate with it and almost never has.

You wouldn't go into a room of traders and try to motivate them by yelling words like "let's pin our ears back boys and play with heart today!" No. You'd tell them about the dollar amount of the fat bonus check they can make if they hit quantifiable goals.

Little primer for what figures to be a loltastic seahawks fan OP for the game tonight

Seahawks are 2-1 on the road this year

12-7 @ CAR
23-20 (OT) @ HOU
30-34 @ IND

Cardinals are 2-0 at home

25-21 vs DET
22-6 vs CAR

Doesn't almost every team have at least something now in some form? Steelers being notable exception

I don't see how this negates what I said. The incentive and platform is still there. It will just take a while due to the "barriers of entry" as you said. The ideas that lead to winning will eventually win out.

I just disagree with your point that money is involved therefore they must be doing everything they can to win*. I mean we literally know this is untrue - especially in sports.

Sports leagues aren't really the same as public or private companies trying to operate in America. They've got it pretty damn easy. A private entrepreneur operating in the real world has way more incentive to exploit edges and try to acquire all the best data than, say, the Dallas Cowboys.

*You didn't say this but this is what I'm interpreting.

lol patriots saying they didn't stat pad on that season. peyton and welker breaking TD record would be awesome

jmill, good stuff

looking up random games on PFR

please for the love of god. it's gone from the theoretical to the fantastical when people are seriously advocating that young stud QBs should throw the coach under the bus in pressers or wave off the punt team.

afaict the discussion isn't very relative at all; it's why I asked whether 4th & 4 at the 50 was debatable. In his original post Boot mentioned that the advantages for good offenses in that scenario would be greater. I'm assuming that also means that the advantage for bad offenses is smaller. I can't tell whether EP factors in the quality of the opposing defense, your own defense, or the opposing offense, but that will also influence the decision. Imo a more complete analysis would utiilize a matrix that includes all of these factors.

I'm also not sure why Baids' excellent work showing that the calculator might be way off (btw, it says the +EP for 4th and 5 is .38) isn't getting more consideration. We seem to be accepting discrete EP/WP calculations as fact, when in reality they have ranges and are a long-term calculation that requires a coach to go for it on 4th and 5 from midfield every other game for 6+ years just to get to ~50 trials. (Obviously I have a problem with EP -- imo its issues wrt evaluating line play are a fundamental flaw, and if they aren't, the solution sure as hell isn't to just ignore line play -- but that's another conversation).

I think we need to find a better balance between the theoretical and practical aspects of this debate. We seem to be assuming that all decisions are made in a vacuum, ignoring outside influences and consequences. But we know that emotion is often a factor in football and the emotional effects of 4th down decisions on a team are a real consideration, even if they can't be quantified. Instead of restating what Horrorshow noted, I'll just quote him:

The incomplete information in football far surpasses that of poker and no HU NLHE pro will say the highest EV decision they can think of is best because it doesn't factor in that they may be assessing the situation incorrectly or worse that the downside of their decision failing could strengthen their opponent more than that of an alternative non-self-weakening decision.

We all agree that coaches should go for it more frequently, in a wider range of field positions, than they currently do. However, when and where is more debatable than SE thinks.

Meh, is it really that ridiculous? He's the top asset in all of football, and he plays for arguably the worst coaching staff in the entire league. Why should he give them a pass on throwing the game away because they never bothered to spend half an hour learning how to handle simple in-game decisions? Not by the standards of people who can use the internet, but by even by the standard of the other risk-averse neanderthals running football teams? If anything, Luck should be putting as much effort as possible into forcing them out.


I think this was mentioned earlier in the thread, but no one here is treating the 4th down calculator as fact. We all admit that the WP calculations are based on small samples, and are often way, way off. The EP calculations are still pretty good, and when they gibe with other 4th-down studies (specifically the chart that's been linked a couple times in the thread lately), we tend to accept them as fact. Even still, the decisions that are remotely close by the chart OR expected points don't get a lot of complaint in this thread. It's only when teams make decisions that are wildly, wildly off that we make a big deal about it. For instance, the Colts decision to punt that 4th and 2 probably halved their chances of winning the game. That's an incredibly terrible decision.


The emotional effects of 4th down decisions aren't some completely unquantifiable effect like you're assuming. There have been studies done. Specifically, Barnwell posted a momentum study on Grantland not too far back, and it showed that teams who give up the ball after missing a 4th down on average do slightly better the next drive relative to their field position than teams that punt the ball away. Now, I'm not saying the reverse momentum effect is real; it was probably just a fluke that would net out to nothing with a larger sample. But the fact that the evidence has it going in that direction though would be enough to show definitively that there's no reason to give up direct EV for a phantom momentum gain, let alone the giant amounts of it that are routinely given up during NFL games.

Also, I don't see how a NLHE player occasionally making slightly -EV decisions in order to tilt his opponent or avoid getting him off tilt would correlate to an NFL football game. For one thing, the mistakes that get attention on these forums are only the massive mistakes. For example, kicking a field goal early on 4th and goal from the 2 is analagous to folding a hand that's 51% against an opponent's range when we only need 37% to breakeven on the call. Obviously, no remotely, remotely decent poker player would ever willfully turn down that kind of edge.

Furthermore, it's not done to any purpose the way a good poker player's decision might be. It's not like the coaches are thinking they can get the opposition to make -EV decisions if they take less risk. They're just being scared nits that are afraid to do anything that could be perceived as risky. Their decision making is in the stone age, roughly where nitty poker players were in the 70s and 80s. The fact that they refuse to move into the new millennium is going to be a continual source of frustration to those of with analytical bents since they're flushing millions of dollars of equity down the toilet due to willful ignorance in a field where lots of us are very emotionally invested.

Issue with Romo are his interceptions aren't epically low to account for his insanely high fumble rate. His combined TOV rate is pretty high I believe. 53 fumbls + 94 picks in 99 career games is pretty bad. Romo is a really efficient QB though. His issues are mainly mental and TOV related.

1.) I was afraid to make that comp because I didn't wanna derail this thread with a discussion on GTO as it pertains to HU NLHE, but I specifically said choosing a +EV decision (x) over another slightly higher +EV decision (y) when the risk of x is lower than that of y.

2.) I specifically noted leaving thin value on the field. An extremely thin spot is a Gronk/Amendola/Vereen healthy Pats team up by 6 over a neutral opponent with 4:30 remaining facing 4th and GL on the 2. Kicking has to yield somewhere in the 0.85 EP range less than going for it (if not a full point with Brady/BB/Gronk success rate) and slightly less +WP than going for it. You really saying kicking is LOLneanderthal here?

Horrorshow: I alluded to this in my earlier response, but I don't think "risk" is particularly meaningful here. If something is +WP, the "risk" doesn't matter at all. You should make that move. That's the definition of +WP in football. By taking a lesser WP option, you will always increase the only relevant risk: that of losing.

I wonder what will happen first:

• Week 6 rankings posted
• Week 7 games finish
• SRM's werewolf game reaches an actual conclusion

It took me well over a year of reading 2p2 to finally come around to the statistical side of sports. You guys are way underrating how long it takes(and how hard it is) to unlearn stuff you've been taught your whole life.

I play professional football, and I despise everything about it

http://www.reddit.com/r/confession/c...and_i_despise/

Denied.

http://www.pro-football-reference.co...0910110buf.htm

17 passes
2 completions
1 interception
0 touchdowns
Your winning QB, Derek Anderson

You don't increase the risk of losing by kicking in that spot over the status quo only over that of going for it. I guess I'm asking what the WP difference is there, too, and if it's thin, when does success rate override that thin value?

Not sure if it was itt or in last week's but in a response to Assani, I said this is where a football games play out optimally more similarly to STTs or even GTO usage in HU NLNE than MTTs or cash games.

Are you saying a team would rather be down 9 100% of the time with 4:30 remaining and getting the ball back at their own 20 than down 13 about 55-60% of the time and the ball at that same 20 along with the 40-45% chance of getting the ball back inside their own 5 down 6?

If so, I'll plead ignorance but I would think the team that's behind would rather you go for it.

Horrorshow, I honestly still don't understand how what you're saying relates to the post of mine that you quoted....I kinda agree with you that there are some spots(although rare) where you'd want to pass up a slight % edge in order to minimize variance, but its just such a small aspect that I'm not sure why its even relevant.

You said in jest that Luck would make a flipping coins reference followed by an LOL math. I'm saying that "LOLmath" and variance reduction are two separate mindsets.