Different teams are different? No shit. There are probably some teams that wouldn’t benefit from going for it more often. However, the numbers say that the majority of teams would.
If you read the article, it notes that the 2002 Ravens and most other teams with very strong defenses and very poor offenses would have done worse by going for it.
Do these sims take into account that the teams will paly downs 1-3 differently if they already have decided to go for it on fourth downn?
Omni raises a valid point about a lack of abundance of data, but that would be more relevant if we were currently pretty close to optimal but were trying to strike the perfect balance. That’s not the case - right now we’re so far away from optimal that any step in the direction of aggression would be an improvement.
Coaches are ridiculously risk-averse, but the fans play a big role in that because they’re results-oriented idiots. If a coach does a QB sneak on 4th and 1, that say… has a 70% chance of success and is much more valuable to his chances of winning the game than punting, then either it’s a good decision or not. It doesn’t become a good or bad decision when it succeeds or fails - but that’s how fans treat it. Coach makes a risky decision and it works - “That man’s a genius! That takes balls!” or if it fails “WHAT THE HELL!? THAT IDIOT! NO ONE DOES THAT THERE! FIRE HIS ASS!” - yet in each case it’s the same decision, only the results happened to work or not work in their favor this time. The same decision cannot both make the coach OMG A GENIUS and OMG DUMBEST COACH EVER, yet that’s how knuckle dragging fans react.
That. I’ll take Bill Belichick going for it and failing against the Colts over Jack Del Rio punting on 4th and 1 at the opponent’s 35 any day of the week.
Doubtful, but that would increase the chance of converting, if anything.
I often think that if I was an NFL coach (which I’m not), and I had a 3rd and 1, I would call two plays. If the 3rd down play doesn’t work, why not immediately line up again without a huddle and try a dive or a sneak or something on 4th down. It seems to me like you could catch the defense off guard and get the play off before they’re set.
I would only do this if I had already some kind of reasonable field position, of course. I wouldn’t try it from my own 15 yard line, but from the 50, I think it’s orth a shot. Guys start running off the field after a 3rd down stop. I really think you could get them out of position. They aren’t going to be expecting you to line right up again. If nothing else, you might force them to have to waste a timeout.
That Bellichick going for it against the Colts thing is the perfect example of this sort of thinking. At worst, they were roughly equal decisions. Bellichick’s decision to go for it was probably the best route. But since it didn’t work, the outcry was hysterical. OMG BELLICHICK WORST COACH EVER, etc. Yet when I asked SDMBers to actually plug in their own numbers into the situation, no one could come up with any scenario in which the decision was more than marginally bad. Yet they were still insisting WORST DECISION EVER. Any rational analysis couldn’t overcome their gut feeling about conventional wisdom.
You can see discussion here and here
Someone even made a website with a simple formula to demonstrate the merits of your decision even plugging in your own numbers.
And yet… OMG WORST DECISION EVER. Ironically Bellichick had done the SAME THING, SAME DECISION, a few weeks earlier, but it worked that time, so not outcries of WORST DECISION EVER. It was the same decision both times, so how can one be fine and one the worst ever? Oh right, the average football fan is more fucking results oriented than gamblers who have a gambler’s fallacy dildo stuffed deep into their anus.
Sometimes I wish I had a tivo set up at the sports bar or something and right when a team lined up to go for it on 4th and one, I pause it and then poll the bar whether or not it’s a good decision here. Their heads would explode. I CAN’T DECIDE IF IT’S A GOOD DECISION UNTIL I SEE THE RESULT!
Edit: That isn’t to say they couldn’t criticize the specific playcall. That’s open game. But for simplicity’s sake I’m just talking about whether or not they think going for it on fourth there (assuming the optimal playcall) is a good idea.
Statistically? Good decision. But in reality, on the road in a dome, coming out of a timeout, with no running game and the crowd going crazy? Not nearly as solid.
This is exactly what I was talking about. It’s like somebody said in a thread about fumble probability, the statistical likelihood of someone fumbling doesn’t really matter when the game is on the line and James Harrison is karate chopping your arm.
Then plug in your numbers, justify them, and see what the result gets you.
They are very different because it’s much easier to gain 3-4 yards on the 45 yard line than it is to gain 3-4 yards on the 3 or 4 yard line. The statistical analysis does 1 of 2 things, I don’t know which without seeing Accuscore’s algorithm. They either treat the theoretical 4th and 4 play from the 45 yard line like every other play the team ran and predict the results versus various teams results. Or they treat the 4th and 4 play like every 4th down play the team ran and run it against various teams 4th down defenses. Both systems are inherently and irreparably flawed.
If they use the first system it’s flawed because the defense coaches much differently in a 4th down situation and the offense uses a different set of plays when the “must” gain 4 yards as opposed to on 1st thru 3rd down where gaining less than 4 yards is an acceptable outcome. Defenses will commit more defenders to stopping the offense short and allow more risk of a deep play on 4th downs.
If they use the second system its flawed because of a dramatic lack of data. The randomness and possibility calculations that were fine tuned using hundreds of thousands of plays for typical situations aren’t going to hold across the 1 or 2 4th down plays run in a game, if that. Not to mention those few 4th down plays tend to represent very strange circumstances like 4th and 17s with time expiring and a team down 2 touchdowns.
The efficiency of offenses decreases dramatically within the 10 yard line. Those statistics would more accurately reflect what plays run on 4th down anywhere on the field look like than an average of all plays or an average of late game, desperation 4th down plays.
By my calculation, there will be a seismic shift in play calling beginning in about 15 years, when the Madden generation starts taking over NFL teams.
ETA: Omni, you’re missing the point. If you’re going for it on every fourth and short you can afford to throw deep on some of them, and thus take advantage of defenses stacking the line.
Heh. You got me. Well, it does work on Madden.
I’m not calling you out, or anything. I think it would work even better in real life. Certainly does for the Colts, and they’re the worst power rushing team in the league.
While statistically, I agree with many of the above points regarding going for it on 4th and share the frustration with coaches who don’t maximize their chances of winning, something I’ve reluctantly come to accept is that winning is not really the coach’s goal. The coach’s goal is to keep his job. And if you don’t have either the job security of Bill Belichick or a statistically savvy owner, it’s probably better to lose conventionally than risk losing unconventionally when your performance review comes up.
Also, regarding Suh and Welker’s extra points kicks - don’t those teams have punters? I mean, I know that punting and placekicking are different skills, but, I mean…they’ve gotta be closer than a wide receiver or defensive tackle who played soccer 15 years ago, right?
You’d think so. Considering how worthless NFL punters are when it comes to punting, you’d think they’d have time to learn how to kick an XP.
No, I’m not. That is irrelevant to the point. I’m criticizing the Accuscore methodology. It’s based on what historically actually happened. Not what might happen or what coaches could do. While I agree that a more liberal 4th down policy would be good and that teams would adjust accordingly on both sides of the ball, that isn’t the point. The point is that you and TMQ can’t quote the Accuscore data as any way authoritative, it may very well be misleading depending on what their algorithms actually are. The fact is, there simply isn’t any data available to determine what the results would be of a more aggressive 4th down policy.
To the extent that data is available, it shows that a more aggressive fourth down policy would work. Is there enough data? Probably not. Doesn’t mean you can dismiss the results.
For all intents and purposes there is no data available. I’m not aware of a single instance of a team going for it on 4th and 3 on the 45 yard line in the middle of a game. Maybe somewhere a team had an injured punter or something but that’s not data in the scientific sense.
That some teams have run 4th down plays on late 4th quarter drives against prevent defenses isn’t relevant. That some teams have run 4th down plays on 4th and inches isn’t relevant. That teams have gone for it on 4th and goal isn’t relevant. All those situations have nothing in common with what that article is espousing.
The only way there’s any data available is if you stipulate that the difference between a 4th and 4 play and a 1st and 10 play or a 3rd and 4 play is negligible. This I’m afraid is patently absurd.
Media reporting that Wade Phillips is fired, Jason Garret promoted to Head Coach.
That’s the most logical solution. Wade was obviously done and Jerry thought 3 years ago that Garrett was the coach-in-waiting all along. Garrett has been a disappointment but based on what he was paid Jerry might as well give him a shot at coaching and see if he’s got a chance of being any good at all.
I never really understood why so few media folks mentioned the much discussed coach succession plan from a few years back when speculating about Wade’s future these last few weeks.
Oh, if you’re not aware of it then clearly it never happened. :rolleyes: