Do you agree with Belichick's 4th down attempt?
View Poll Results: Do you AGREE with Belichick's 4th down attempt?
Yes
344
64.06%
No
193
35.94%
his exact day to day is a closely guarded state secret, but his title is "Director of Football Research". from the linked article you can tell a large part of his job is quant analysis and implications thereof...
Adams' official title is director of football research, and he does a lot of that, too, trolling the world for things that might offer the slightest advantage. A year or two ago, an Andover teammate ran across an obscure out-of-print book on nonlinear mathematics. He thought Adams might find a use for it, so he mailed it to him. Adams had already read it. Or there's Rutgers statistics professor Harold Sackrowitz, who got a call from Adams a few years back. Adams wanted to talk about some research Sackrowitz had just completed, dealing with how teams try two-point conversions far too often. Adams sent the professor the Patriots' when-to-go-for-two chart, and asked Sackrowitz to tear it apart. Of the 32 NFL teams, the statistician told the New York Times, only the Patriots called.
Here's another example: The academic paper of a Berkeley researcher, referenced in the same Times story, dealt with how teams punt on fourth down far too often. That paper ended up on Belichick's desk. Now, how do you imagine it got there?
On game day, Adams wears a headset in the press box, a direct line to Belichick. Adams advises Belichick on which plays to challenge, and charts trends.
Here's another example: The academic paper of a Berkeley researcher, referenced in the same Times story, dealt with how teams punt on fourth down far too often. That paper ended up on Belichick's desk. Now, how do you imagine it got there?
On game day, Adams wears a headset in the press box, a direct line to Belichick. Adams advises Belichick on which plays to challenge, and charts trends.
doubt this will require much thought
And as I stated in my post, he is the master at this so he may be able to make it a non factor. But just remember when he let Lawyer Milloy go right before the first game of the season with Buffalo. He eventually made that controversial move a non factor but the team was clearly distracted that first game after the firestorm. In that situation it was definitely worth sacrificing one game because of the long term implications, but in the present case, there are no long term benefits of his call. Only a short term gain of a 3% better chance of winning the game they were playing.
Here's the decision with all variables:
MAKE > 1 - LONG/SHORT
MAKE = Prob of making the 4th down conversion
LONG = Prob of other team driving for TD after a punt
SHORT = Prob of other team driving for TD after failed 4th down conversion
So if left side is greater than right side, go for the first down.
http://waynewinston.com/wordpress/
MAKE > 1 - LONG/SHORT
MAKE = Prob of making the 4th down conversion
LONG = Prob of other team driving for TD after a punt
SHORT = Prob of other team driving for TD after failed 4th down conversion
So if left side is greater than right side, go for the first down.
http://waynewinston.com/wordpress/
peter king says:
Why use the historical average net punt of 38 yards when Hanson's average net for the night, for four punts, was 44 yards?
Why use the historical average net punt of 38 yards when Hanson's average net for the night, for four punts, was 44 yards?
And the expected value needs to consider averages that are the result of the choices of offensive play and defensive response, which you can obviously apply game theory to. All in all it's a silly thing to nitpick about that seems like it was chosen to give some sort of purchase for arguing with Phone Booth.
The Peter Kings of the world are numerous and we're lucky to have them at our poker tables. "I don't care about your mathematical theories or long-term statistics. I've been hitting my draws all night. I call!".
that freddie clown needs to meet a grease fire. i will admit he is good at stringing together big words in a coherent fashion that is utterly ******ed.
I doubt that creating an algorithm to make play calls would lead to better decisions; unless it was extraordinarily specific. You wouldn't have time to plug in the variables. If you got Randy Moss matched up against a poor cornerback and noticed the defense was in a specific alignment you wouldn't want to run whatever the algorithm determines and obviously change the play.
my grandfather drives me nuts with the result oriented thinking
i was bitching about some of girdardis post season pitching changes and he defends them if they yankees won the game with "they won so it was the right move"
i then asked him if a manager pinch hits for pujols in the ninth inning with the pitcher and the pitcher hits a home run was it the right move- he says yes i smash my head into the wall
i was bitching about some of girdardis post season pitching changes and he defends them if they yankees won the game with "they won so it was the right move"
i then asked him if a manager pinch hits for pujols in the ninth inning with the pitcher and the pitcher hits a home run was it the right move- he says yes i smash my head into the wall
my grandfather drives me nuts with the result oriented thinking
i was bitching about some of girdardis post season pitching changes and he defends them if they yankees won the game with "they won so it was the right move"
i then asked him if a manager pinch hits for pujols in the ninth inning with the pitcher and the pitcher hits a home run was it the right move- he says yes i smash my head into the wall
i was bitching about some of girdardis post season pitching changes and he defends them if they yankees won the game with "they won so it was the right move"
i then asked him if a manager pinch hits for pujols in the ninth inning with the pitcher and the pitcher hits a home run was it the right move- he says yes i smash my head into the wall
when i have a grandson he better be smashing his head into a wall too
No no! If you stop smashing your head into the wall, you will stay coherent and then your grandson won't have to get frustrated and feel the need to smash his head. Break the cycle.
hell smash his head from hearing tim mcarvers head floating in a jar analyzing baseball
The whole "it didn't work so it was the wrong decision" has got to be the most frustrating angle anyone is taking in regards to this.
Oh noes, a kicker missed a 50 yard FG as time expired while down 2, should have thrown a hail mary instead. Just a ridiculous argument.
Oh noes, a kicker missed a 50 yard FG as time expired while down 2, should have thrown a hail mary instead. Just a ridiculous argument.
Quote:
peter king says:
Why use the historical average net punt of 38 yards when Hanson's average net for the night, for four punts, was 44 yards?
Well, I think he may have accidently hit on something here, I would imagine his net punting average is higher in domes when wind/weather is not a factor (return man gets a faster track to counterbalance but so do the gunners/punting team).
peter king says:
Why use the historical average net punt of 38 yards when Hanson's average net for the night, for four punts, was 44 yards?
Well, I think he may have accidently hit on something here, I would imagine his net punting average is higher in domes when wind/weather is not a factor (return man gets a faster track to counterbalance but so do the gunners/punting team).
I'd like to see a list compiled for which media types have come out for and against BB's decision.
PB- You're assuming that what is true of the general is also true of the specific.
You're right that the average person in a position of authority is there for a reason and the "elitist" common man is far too quick to be "LOL baseball GM what an idiot," but we aren't talking about the position as a whole. We are talking about a specific situation.
You're right that the average person in a position of authority is there for a reason and the "elitist" common man is far too quick to be "LOL baseball GM what an idiot," but we aren't talking about the position as a whole. We are talking about a specific situation.
You're also kneejerking too far the other way against people who reject authority.
It's the height of arrogance to think, for example, that I could become a greater authority in economics than, say, a Nobel Prize winner because I read 3 articles on mises.org, but CEOs/GMs/politicians are just people like anybody else. It's not like we're Betas incapable of understanding their Alpha decisionmaking processes. That there are variables that we are unaware of from our collective position as "idiots on the internet" doesn't mean those variables don't exist.
Earlier you implied that Robert Kraft knows more about math and game theory than us.
Setting aside that this isn't a game theory situation and the math involved is multiplication(iow, the skills necessary to make the correct call here/realize the correct call was made are psychological/cognitive skills), how do you know?
Furthermore, it's likely that Robert Kraft is far more emotionally stable - there are likely far fewer gaps and holes that cause him to refuse to think about certain possibilities because they are uncomfortable. This matters a ton, when it comes to applying math correctly to high pressure situations. This is the sort of thing that TomCollins was talking about - there's no reason why his lawyer friend shouldn't be able to understand the reasoning here. My experience is that twoplustwo posters as a group aren't very good in this department - they have such a strong need to validate their ego that much of their belief system works strongly to reject certain thoughts. They may not be any worse than average people - I have no idea - but top business people are a high target - the business world very strongly rewards people with few ego-driven emotional glitches that affect thought processes.
He's got a MBA from Harvard so he's reasonably intelligent, but according to Wikipedia he's never worked as a mathematician or anything.
He's been successful in business, but obviously we don't know the specific history of his business ventures. For his business we know the most about(the Patriots), it seems that Kraft's success stems from making excellent hiring decisions.
Given my dealings with MBA students of today, I'd be stunned if his math skills were better than that of a typical 12th grade math teacher.
and by what authority do you now make this claim?
yes, they did, and that's precisely my point. people well outside the traditional 'decision making' arena nudged their way into that arena.
but of course, you go on to argue that sabermetrics has little value, while alternately arguing that baseball teams are smart by snapping up sabermetricians.
and this is just a brutally terrible argument. you are saying that because one guy knew that 'hey, walks are good', that really, i mean, those sabermetrics guys didn't really do anything - it was the cigar-chomping GMs who understood the game through and through.
http://www.baseballthinkfactory.org/...o_old_idea.htm
Onbase percentage as a statistic, I believe, was invented by Branch Rickey.
It's also quite likely that many front office people before him and after him had similar ideas and were open to even others. The problem has always been that of culture - you need guys who understand baseball well at every level and a lot of these guys weren't numbers people. It's fairly clear that if you had to have either traditional baseball people or stats people, but not both, you'd go with the former. Culture clash causes emotional friction, political issues, etc, which often leads to subpar decision making, especially in knowledge-oriented ventures. Furthermore, owners had no real basis with which to evaluate baseball executives other than on-field results, personnel management, organizational structure that seems to reward people who understand the game, etc.
What changed isn't the introduction of new knowledge, but rather a way for outsiders to understand things that were exclusively within the domain of insiders, allowing for a greater level of accountability throughout the organization and to the owner. Generally successful business people are the type of people who would be open to ideas like this and predictably, they moved quickly. Note that traditional baseball people were still an important part of the organization and simple sabermetrics appeared to work better than it would in a neutral setting because it wasn't widely used. In a market setting, if two people had equal knowledge, the one with more variant knowledge (further away from the market norm) does better. For the same reason, statistical projection sometimes works well because teams aren't using the same statistical projection to promote players. In this sense, I'd argue that front offices as a group moved about as fast as they could given the constraints and at no point were they far behind in terms of utilizing proven knowledge. While cultural inertia does hold innovation back, market incentives are aligned the other way - knowledge that other teams don't have is more crucial to success than having knowledge that all teams have.
considering the importance of sabermetric thought in baseball today, and the number of people employed by major league teams - the most stodgy and traditional of old boys' clubs - who never played baseball, i think you are way, way off base on this. sabermetrics is a totally new paradigm through which to evaluate players.
http://sports.espn.go.com/mlb/column...lan&id=1835745
There's no new paradigm, but gradual increase in the ability of people without intuitive, specific knowledge, but with strong comprehension mathematics and data analysis to judge performance of others with such knowledge. This can be a GM evaluating splayer, scouts or managers; a manager evaluating players matchups, roles, an owner evaluating GMs. Most of this simply has to do with advances in data gathering, gradual increase in the availability of and capabilities of personal computers, wider acceptance of statistical and computational techniques throughout the business world. Very little has to do with outsiders being right and insiders being wrong.
And the expected value needs to consider averages that are the result of the choices of offensive play and defensive response, which you can obviously apply game theory to. All in all it's a silly thing to nitpick about that seems like it was chosen to give some sort of purchase for arguing with Phone Booth.
this is the guy I was referencing
his exact day to day is a closely guarded state secret, but his title is "Director of Football Research". from the linked article you can tell a large part of his job is quant analysis and implications thereof...
his exact day to day is a closely guarded state secret, but his title is "Director of Football Research". from the linked article you can tell a large part of his job is quant analysis and implications thereof...
Originally Posted by ESPN Article
By now, after years of evolution, Adams sees film differently. Not just as random actions, but a genealogy of the game of football. When a defender moves, he recalls watching or having read about the first time a defender moved like that, even if it was 50 years ago, and he knows why, which tells him how to counteract the move. He has a photographic memory. Perkins tells a story of Adams' memorizing the Giants' thick playbook. In one night.
Originally Posted by ESPN Article
Pats owner Robert Kraft, a successful businessman in his own right, discusses economics with Adams.
I think an analogy that sums up the position of most of the people who acknowledge the math and are backing off calling it a terrible decision, but still aren't willing to say it was the right call is playing with a rigged deck.
Yes, historical statistics show that pockets Jacks have an edge over AK, but what if their are six aces in the deck we are playing with (obviously could be the other way and there are only two aces in the deck)? Given all the variables of this spot what proof do we have that we are playing with a standard deck? I mean this Casino seems pretty sketchy and the dealer has tattoos on his neck, are we sure our odds are the same here? And what is the sample size of spots that are sufficiently similar to use in this situation. The response is that factoring in Brady and Manning only makes it more of a no-brainer, but how do we know that the other factors in this game do not outweigh the Brady and Manning factor?
The key question that swings this from being a slightly bad call to probably being a good call is whether Tom Brady converts in this spot more than 50% of the time. My intuition would tell me that he does, but I haven't seen any stats that support this and the 45% league average conversion rate for two point conversions seems much more analagous than the 4th and 2 60% number cited by Advanced NFL stats since their are many varieties of 4th and 2 which are not analagous at all, for example many spots that still have the defense giving a small cushion or even playing prevent.
I've been looking for Tom Brady's two point conversion stats, but other than sifting through game logs, I haven't found a source for this. While still being a woefully inadequate sample size and different in many respects, I would think this would really be the most relevant historical stat you could use in this situation. Limiting it to above average defenses would be even better, but the sample size would be tiny.
Yes, historical statistics show that pockets Jacks have an edge over AK, but what if their are six aces in the deck we are playing with (obviously could be the other way and there are only two aces in the deck)? Given all the variables of this spot what proof do we have that we are playing with a standard deck? I mean this Casino seems pretty sketchy and the dealer has tattoos on his neck, are we sure our odds are the same here? And what is the sample size of spots that are sufficiently similar to use in this situation. The response is that factoring in Brady and Manning only makes it more of a no-brainer, but how do we know that the other factors in this game do not outweigh the Brady and Manning factor?
The key question that swings this from being a slightly bad call to probably being a good call is whether Tom Brady converts in this spot more than 50% of the time. My intuition would tell me that he does, but I haven't seen any stats that support this and the 45% league average conversion rate for two point conversions seems much more analagous than the 4th and 2 60% number cited by Advanced NFL stats since their are many varieties of 4th and 2 which are not analagous at all, for example many spots that still have the defense giving a small cushion or even playing prevent.
I've been looking for Tom Brady's two point conversion stats, but other than sifting through game logs, I haven't found a source for this. While still being a woefully inadequate sample size and different in many respects, I would think this would really be the most relevant historical stat you could use in this situation. Limiting it to above average defenses would be even better, but the sample size would be tiny.
Christ. I really wanna copy paste that into my Word doc and word count it.
However, I'm a bit afraid that it'll stall my machine counting that crap.
Phone Booth,
Do you have an academic background?
However, I'm a bit afraid that it'll stall my machine counting that crap.
Phone Booth,
Do you have an academic background?
Calling him a stat guy is just wrong
they both seem smart and knowledgeable enough about their subject matter to marry all types of data, quant and qual
kidcolin,
any dispassionate hyper-rational analyst whose one and only objective was to answer the question, ''what should bill belichek have done in this particular instance if his one and only objective was to win this specific football game?" would not be able to provide a definitive answer. they would instead offer up a probability distribution of answers given what we know about the usefulness and reliability of the available evidence and methods. unsurprisingly, that sort of talk is not seen because it is boring, unpersuasive, and signals an unwillingness to accept the ''spirit of the conversation'', i.e., a refusal to participate in the social game that has sprung up around this super exciting event (i think he should probably just punt the god damn ball, if you're wondering; expressed well, it makes the most friends).
any dispassionate hyper-rational analyst whose one and only objective was to answer the question, ''what should bill belichek have done in this particular instance if his one and only objective was to win this specific football game?" would not be able to provide a definitive answer. they would instead offer up a probability distribution of answers given what we know about the usefulness and reliability of the available evidence and methods. unsurprisingly, that sort of talk is not seen because it is boring, unpersuasive, and signals an unwillingness to accept the ''spirit of the conversation'', i.e., a refusal to participate in the social game that has sprung up around this super exciting event (i think he should probably just punt the god damn ball, if you're wondering; expressed well, it makes the most friends).
what i find so interesting about this thread is how people choose to play this game (this relates directly to pb's talk of business; perhaps more later) around the game. "I Believe In EV!" "Conventional Wisdom Aint So Crazy" "Meta-Analysis of EV vs. CW/Im Smarter Than All", all of these are techniques people use to..
The key question that swings this from being a slightly bad call to probably being a good call is whether Tom Brady converts in this spot more than 50% of the time. My intuition would tell me that he does, but I haven't seen any stats that support this and the 45% league average conversion rate for two point conversions seems much more analagous than the 4th and 2 60% number cited by Advanced NFL stats since their are many varieties of 4th and 2 which are not analagous at all, for example many spots that still have the defense giving a small cushion or even playing prevent.
Right. It sounds like he's one of the most knowledgeable people about qualitative aspects of football as well. He's a nerd, but he's an all-around football nerd not just a stat/numbers/quant nerd.
Quote:
peter king says:
Why use the historical average net punt of 38 yards when Hanson's average net for the night, for four punts, was 44 yards?
Well, I think he may have accidently hit on something here, I would imagine his net punting average is higher in domes when wind/weather is not a factor (return man gets a faster track to counterbalance but so do the gunners/punting team).
peter king says:
Why use the historical average net punt of 38 yards when Hanson's average net for the night, for four punts, was 44 yards?
Well, I think he may have accidently hit on something here, I would imagine his net punting average is higher in domes when wind/weather is not a factor (return man gets a faster track to counterbalance but so do the gunners/punting team).
I am not in the habit of inventing odds and pretending that they have the authority of mathematics, unlike most of the people in this thread. I think the Patriots had a better chance to win had they punted
Feedback is used for internal purposes. LEARN MORE