tallguyindc says:

Whenever an announcer refers to the "numbers" or the "math" stop and do the calculations yourself.  95% of the time (OK I made that number up but it feels right) they have no mathematical basis for saying what they are saying and haven't ever stopped to look at a single number....which is pretty pathetic for someone being paid big money to analyze football for a living.

mischkin04 says:

 FG there at least keeps hopes alive by my math which is not as deep as yours. You don't understand the difference between a 2 score game and a 3 score game? Your figuring doesn't include the intangibles, such as turnovers.

  Phillips decision to go for two, down by 12, was ridiculous beyond the 4th and 6 failure. I seriously doubt any other head coach would have done that. Seems to reenforce the thought he doesn't recognize the logic of staying in a game.

 

shudawg says:


I agree.  I do not see where this is hard.  They were down three scores.  They needed a FG somewhere along the line.

They should have went ahead and kicked it, and been down two scores.

I also agree with you on the two point conversion - you don't go for two until you have to.

They would have been down 11 kicking the extra point - that is a TD and two point conversion and a FG.  If they would have missed the extra point, they would have been down 12 and it takes the FG out of the equation.

I agree with you.  It seems like easy math to me.

gamblingallday says:
first of all it was tirico and jaworski that kept saying you got to kick the field.

tallguyindc says:

"Staying in the game" is a ridiculously overrated concept.  So is overreliance on the 2 score/ 3 score thing.  The bottom line is all scores aren't created equal.   With the field goal, the Cowboys would still need to score two touchdowns, hold the Giants scoreless, convert a two point conversion and win in overtime.

Scoring the touchdown there gives a much easier path to victory.  At some point, you need to try to actually win the game and not just "stay in it".  Staying in it seems like a great way to constantly lose close games. 

This game was a perfect example.  Bottom line, if they had caught the touchdown there, they would have won the game.  If they had kicked the field goal, they would have still had less than a one percent chance of winning the game in the last minute...even with all the other things that went right for them in the fourth quarter.

I mean just do a thought experiment.  After all the other stuff happened, the replay booth calls down and says they made a mistake on that long ago play.  Dallas can redo it. 

They have the ball on the 6 yard line down 6 with 45 seconds left.

They can either try to win the game with one play (about 40%) or they can kick the fg, kick the onside, drive the length of the field, kick another fg, play overtime and win in overtime.

Are you seriously saying that the latter course of action has a greater probability of success?  I mean it would definitely help them "stay in the game" longer so I guess that means they should do it.

The bottom line is that there is a serious problem in the NFL that coaches are far far too timid and they are constantly trying to avoid losing instead of trying to win. 

 

 

lakers34kb says:

Our field goal kicker is unreliable, he made the right call.

tallguyindc says:

You shouldn't go for two unless you think you have >50% chance of making it....since the NFL average is <50%....why would you think I would say that?

I think you are missing the big point here.  "Two scores" isn't really just two scores.  Its two touchdowns....a 2 point conversion....holding the Giants to zero scores...eventually scoring another one in overtime. 

My scenario isn't absurd at all.  In fact its the very scenario that ended up playing out.  In order for the field goal to work, you need a whole bunch of other things to work as well and you have effectively not done anything to dramatically improve your odds of winning the game.  A touchdown would have helped a whole lot more.

If you do the long form math, you'll see that this is really true and it was not even a remotely close decision.  

 

tallguyindc says:

That was also very clearly the correct decision and not by a close margin either. 

What would be the conceivable logic in not going for two in that situation?!?!?!

You are probably going to have to get the 2 eventually.  If you make the two, you know you can kick the fg if you get into a situation later.  If you miss the two, you know you need to go for it. 

If you kick the extra point, then next drive its 4th and 5, you kick the fg, then you make the td and miss the 2.  What did that get you?

If you look at as I'm going to go for 2 eventually.  Make or miss, I'm better off knowing now.

"Was refering to your  " thought experiment " of  2 FGs to tie as a completely unconsiderable option."

Why was it a totally unconsiderable option?  I agree it was a stupid option, but it appeared to be the option that Tirico and company wanted to take.  Phillips was smart not to take it.

The only difference was the timing.  It wouldn't have been at all obvious that they lost the game because of the fg, but thats what would have happened.  Doing that would have painted them into the corner where they needed all those things to happen.

The fact is calling 15 points two scores is a mirage.  Its only two scores under ridiculously good circumstances and you are going to need to get extremely lucky to win the game when you are down by 15.   When your down by 11, you are considerably better off than when your down by 15 and if you stop to think about what it really entails being down by 18 isn't all that bad compared to 15. 

Here is another way to think about it.

If you are down by 11, 2 TDs and a scoreless opponent, you have a 100% chance of winning the game.  Down 15, 2 TDs and a scoreless opponent, you have a 24% chance of winning (48% 2 pt conversion * 50% OT)

If you are down 11, 2 TDs and an oipponent FG, you have a 50% chance of winning the game.  If your down 15, 2 TDs and an opponent FG and you have a 0% chance of winning the game.

The problem is the rounding to "two scores".  Its a simplistic shorthand and the announcers like to count scores as if they are all roughly equivalent.  Dig deeper and you see that 2 scores at 11 means you score twice, you probably win.  2 scores at 15 means you score twice, you almost certainly still lose.  

 

tallguyindc says:

What good is it to be close to 8 if you are going to miss the 2 point conversion?

Think about it a different way.  I flip a coin.  Heads your going to get your next two point conversion.  Tails and you are going to miss it. 

Its now 4th and 5 on the 20 yard line.  Do you want to know what the flip was or not?

If you are a smart coach, you probably want to know.  If you are like most NFL coaches, probably not.

I guess my point is that the 2 point conversion will be a huge play regardless of when you try for it.  Try for it early and fail and you can still alter your strategy to have at least some shot.  Try late and fail and you are completely screwed.

If you think about it, this isn't that different than college OT strategy.  We both agree the second team has a big advantage because they will never kick when they need a TD. 

Same basic principle.  You are always better off knowing how many points you need.

I'm literally shocked that so many coaches seem to take "never go until you absolutely have to" at face value.  They are losing games because of it.   

casheasy says:

ok Tallguy hit on something that all announcers, and some of you, are completely lost on. I don't care whether you kick the fg there, you can look at it both ways. But to say you're down 2 scores at 15 points is wrong. You are down 3 scores, UNLESS you get the 2 point conversion after a TD. Once you assume that, then you are also making the argument that you should absolutely go for 2 at any time in the game, since you are already admitting you're going to make it anyway.

Kind of related to this, seeing as how the conversation went this way. LOL These same announcers told Minnesota to kick the extra point when the were down by 2 against NYJ 2 weeks ago (15-13), yet they're yelling to kick the FG here in this game. WHY? If you already admit that it's 100% certain you're going to make the 2 point conversion (which you are saying by thinking you're only down 2 scores when the lead is 15!!!!)  .....then go for 2 now, when you're down 2 in the 2nd half. Or anytime for that matter.

mischkin04 says:
 

Same basic principle.  You are always better off knowing how many points you need.

 

    The score changes throughout the game so you don't know what that will be until the end.

 

I guess my point is that the 2 point conversion will be a huge play regardless of when you try for it.  Try for it early and fail and you can still alter your strategy to have at least some shot.  Try late and fail and you are completely screwed.

 

      Why pass up a guarenteed point early in any contest for a less than 50/50 chance to get 2? Wouldn't that be like throwing points away? You might not have had to alter your strategy if you had taken the sure thing earlier. Of course there's failures, it's not a perfect world.

 

 

Barnstorm says:
And you were wrong about Gruden. He's the one that agreed with Wade in going for the TD.

tallguyindc says:

 

Can you come up with a reasonably realistic scenario where you go for 1 and eventually end up with a tie or a one point game without getting a two point conversion later?

No you can't.   The closest would be that somehow NY scores two more field goals and you score 3 more TDs.  That is extremely unlikely with the amount of time left.

The advantage of the early 2 greatly outweighs it. 

tallguyindc says:

 

The "optimistic score" and the "pessimistic score" are part of the problem here.  Its like the announcers are saying you have to be pessimistic with AA.  They could get broken.  You should be aggressive with 99.  AK might hit but you might still win with a river 9.

The way to make these decisions logically is to rationally calculate the probabilities of all the possible events.  Go with the one that gives you the greatest likelihood to win.  Obviously you can't pull out Excel at that moment, but it really helps to do a post mortem because 95% of coaches are making the same wrong decisions repeatedly.

The mistake that they are making is always assuming the worst about the current decision's probability of success while counting "optimistic scores" in the future.   

casheasy says:

 

Mischkin, if you choose to be optimistic, then you go ahead and assume you're going to make the 2 pointer, however foolish that is. BUT, if the announcers are going to do that, then they MUST admit they were wrong 2 weeks ago in the Vikings at Jets contest. That's when they said you should NOT go for 2 until the absolute last play of the game (which is the dumbest idea I've come across from football announcers and coaches on a widely accepted basis.) If you already know you're going to make it, then who cares when you go for 2????? By the way, the Vikings did NOT make it, proving the Vikings correct for trying it at 15-13, instead of waiting for 22-20, when it would have been too late. Remember, you can't say you're only down one score if you're down by 8 (or 2 down by 15) unless, you can see the future and KNOW you're going to make it. Since nobody can see the future, the correct decision is to go for 2 earlier, so you can find out if you make it or not. Then you can adjust your strategy. If you wait until the last second, it's too late when you miss it.

 

I realize you like to be "optimisitc,"  and "inclusive." But that's not real life. That's not how football works.

 

THEMUGG says:
 Maybe Wade should try to figure out the odds of being in that position after having a 20-7 lead......

casheasy says:

Also you cant factor in what happened afterwards to buttress your argument. The transpired afterwards was in part due to the decision made on that play. Not only are the coaching decisions after dictated by time and score, the player mentalities are different as well. Do you think the defense plays as hard up by 6 verse up by 13 with a minute to play? do you think the defensive alignment would be the same going down the field at up 3, 6, or 13?

 

You are using flawed numbers. The difference would be are we down by 2 or by 1....or by 8 or by 7..... I don't think you understood what I was saying in my original post.

 

I am saying that IF you say "hey, let's just go for 1 here because if we're down by 8 later, we're only down 1 score..." then you absolutely go for 2 now, since, who cares? We're going to make it! Because the only way 8 points is 1 score is IF you make the 2 pointer.

 

Now go to what tallguy was saying earlier. Because he mentioned what the announcers were arguing. They did NOT say, "Well, they should kick the FG because later on we'll be down by 16 instead of 17."   They said they'd be down 15 instead of 18, and Tirico called that "two scores."  And I didn't even get to ask them, "ok guys, what happens if NYG kicks a FG later? I'd rather have 7 than 3 here, if that's the case. Having said that....

Again, if you read my original post, I said that in last night's game, you could look at it both ways. In fact, I would have kicked the FG for a momentum lift to let my defense know that we're in this game, etc.   But there are mathematics behind what tallguy is saying. I actually do NOT completely agree with that method of determining an outcome or decision, but there's some merit to looking at it that way. And he hit on a few good points. That's all.

 

 

gabuchu says:
How did you factor in the time left in the game into your probability outcome? I am sure there are other important factors did you are not accounting for. I for one will fully admit I am not smart enough to calculate the percentages because there is too many variables.

Please don't compare it to poker %. That is apples to oranges.

gamblingallday says:

jgo22 says:
Going in the booth has cost gruden a future coaching job he is the biggest brown noser and will come out of the closet before seasons end. Why do you think he is not coaching every team he has gone to he has destroyed!!!! And he I mean tampa won the super bowl hmmmm. he is super f'n gay everyone is the best and last week jaws finally had enough and basically told him to shove it!!!!

winner2 says:
100% shoudl of kicked the fieldgoal. no doubt about it should of kicked it

Phor20 says:
i don't know how you can argue that it was wise to not kick FG. take points anytime you can. kick FG made it 2 possession game not 3, duh!!! people are SOFA KING we Todd did

tallguyindc says:

 

OK.  You need to read the post a little bit closer.  Gruden (actually I think it was Jaworski but I thought it was Gruden when I heard it) said after they got to six points "If only they had kicked the field goal" it would now be 3 points.  Implicit in that statement was the assumption that everything else would have gone the same way as it did.  I recognize that assumption is faulty.  But lets just run with it a second. Plus he said it not me. 

Obviously with everything else playing out the way it did, the Cowboys would have been up by a point if they made it.  The way I look at it, if you are going to assume everything else plays out the same, it proves that Phillips did make the right decision.  If they had made it, they would have won the game.  If they kicked, they would still be in desperation mode as the kick wouldn't have helped much.  Even with his own assumption in place and the ability to revisit it with the knowledge of the future scores, he still thought kicking was the right move.  I found it to be a ridiculous statement and I still do.  

Lets step back and look at this from a more macro level.

Have you ever seen a situation when a team could kick a field goal but goes for a touchdown and misses and eventually loses by 3 points or less?  What did the announcers say about it?  

Whenever this happens, they talk about it endlessly.  Often, they completely ignore the game at hand in order to keep talking about it.  In the post game press conference, there are no other reasons presented for the loss.  Thats the only one.  The coach could have kicked and he didn't.  Talk radio lines light up all day Monday with people's anger. 

Have you ever seen a situation where a team could kick a field goal but goes for a touchdown and makes it and eventually wins by a 4 points or less?  What did the announcers say about it?

Absolutely nothing.  They never mention the play the rest of the day.  When the team ultimately wins, other factors are always presented.  Nobody ever acknowledges that the coach's "gamble" (actually a well calculated +EV risk) was a key factor in the victory. 

How often does a team kick and then eventually lose by 4 or less?  How often do you hear about it then?

Perfect example:

Opening weekend.... The Redskins took points off the board and then didn't score.  They won anyways, but Collinsworth couldn't shup up about it the rest of the game.  If they lost, you knew that would be the reason and every newspaper in the country would put "Shanahan gambles game away" in the headline.

The very next night, the Ravens took 3 points off the board and replaced them with 7.  They ultimately won by a single point.  I watched all the postgame shows and read a lot of papers the next day. Not one single reporter focused on it.  Nobody acknowledged that this was a key to the victory. 

I bet you watched that game.  Do you remember them doing it?  Do you remember anybody talking about it?  I bet somehow you blocked it out from your memory because nobody else made any sort of big deal.  I can guarantee you they did it and the Ravens would have likely lost if they didn't.

For reasons that I don't at all understand, its like the announcers have some sort of pact to never mention these calculated risks unless they fail.  I think the coaches pick up on that and say..."I know the odds are in my favor but I really don't want the blame if it doesn't work" so they wimp out.  Since nobody does it, it becomes even more of a story when somebody does. 

As somebody that does study the math and knows the truth, I can tell you that the "gamble" works both in theory and in practice. 

Far far more teams have ultimately lost games they could have won because they didn't gamble than because they did.  Smart fans need to start noticing this and doing a few of their own calculations.

You say you know math well and you would run circles around me.  Great.  Go for it.  Calculate the probability that the Cowboys win the game when they are down 15 at that point.  At 11. At 18.

I'll even make it easy on you.  I'll let you round the 18 probability down to 0. It wasn't zero and that helps my case but I'll still let you do it for simplicity.

If it was the wrong decision, it should be easy to prove with a statement like "They had a 4% probability at 11 and a 3% probability at 15.  40% of 4% < 3%"  But you can't make that statement.  Can you?  You know why...because its not true. 

Nearly any way you calculate it, whatever number you calculate for their prob at 15, their prob at 11 was four times higher. 

I've been called "stupid", "idiotic" "a moron", etc. etc.  but nobody has presented anything near a coherent argument as to exactly where I'm wrong.  You said you were going to run circles around me.  Start running!!! 

 

 

tallguyindc says:

 

I'm familiar with the volatility argument.  At first, it makes a lot of sense.  I think its a holdover from the financial industry.   A lot of math people worked in finance at some point.  Basically, you'd much rather a mutual fund that provided a 10% annual return than one that had a 50% chance at quadrupling and a 50% chance of going to zero.  Makes sense in that context.

I don't think its as relevant in football as people think.  The problem with worrying about volatility is that one way or another you are going to win the game or you're going to lose the game.   Volatility would make sense if a coach could go to the other coach and say "Hey you take 0.6 wins.  I'll take 0.4 wins".  The other coach would say "OK deal...I was scared I'd lose too".  You can't do that.  No matter what you do, there is going to be risk.  No matter what decision you make, its going to be a gamble.  If you have to gamble, you might as well make the bet that gives you the best chance at winning. 

tallguyindc says:

 

Mischkin,

You're quickly becoming my favorite poster on this site.  You don't quite buy into what I'm selling but at least you're trying to figure it out and not just instantly dismiss it.

Its a very good question.  The leap in importance is really from 15 to 13.   Its a much smaller leap from 13 to 12 and then from 12 to 11.

Remember the implied assumptions behind 15.  You don't only need to score two TDs and hold the other team scoreless.  You also need to make the 2 and win in OT.  None of  that is true at 13.

I don't see much difference at all between 13 and 12.

12 to 11 is important because it gets you the ability to use a fg and a 2 point conversion, but you still need the two point conversion.  That makes 11 to 10 a much more important jump.

I've been focusing on the importance of 11 instead of 15 but I would have thought Phillips should have gone for 2 if he got it.  I guess arguably this whole thing should have been about the difference between 15 and 12/10, but that would have just confused the math and I think most people reading this are confused enough as it is.

Bottom line: 11 isn't a key number.  13 is. 

 

 

 

tallguyindc says:

 

OK.  You're obviously not very perceptive.  Read my username closely.  It will tell you exactly where I'm from.  Hint: It makes it difficult to say the Cowboys did anything right.

You're clearly slow so I'll spell it out for you.

Obviously it would have been better to have the FG than the missed attempt. 

But that wasn't the question.  The question is which is worth more in that situation.  3 points or 1 play from the 6 yard line for a shot at 7.

Now think back to the end of the game.  You're lining up to do an onside kick down 6 points.  I offer you either 6 points or one play from the 6 yard line.  Which do you pick? 

tallguyindc says:

I see not a single person in the "FG first" crowd has taken me up on my challenge to actually post the probabilities.

Thats because almost no matter how you calculate it being down 11 gives you a 4 times greater chance than being down 15 at winning the game. 

Poker analogy:

You're ridiculously short stacked. You can go fold or go all-in on a draw.  You only have a 40% chance of hitting that draw, but if you hit it, you immediately quadruple your meager stack.

You're saying that the right move is to fold?!?!?!?!??!

If so, I really want to play against you.

The problem is that nearly all coaches and announcers are really weak-tight poker players.  They don't take gambles even when the pot odds are ridiculously aligned in their favor.

Then when somebody does take a calculated risk they talk endlessly about what the "numbers" say without mentioning a single number.

I challenge anybody to post a mathematical reasoning as to why going for it there wasn't a smart move. 

tallguyindc says:

One more person that tries to answer the question without any mathematical basis.

 

This is getting repetitive.

Whats the prob at 15?  11? 18?

tallguyindc says:

 

Actually, this is sort of a legitimate point.  Within the confines of the absurd BCS system (don't get me started), this makes sense.

Without a doubt though this move took his win probability from slim to none. 

I often see NFL coaches make similar moves and there is no logic behind it at all ....other than I don't want to get fired for being blown out.  I know this is the logic but I'd love to hear a coach explain on the podium that he places his own job security higher than his team's chance of winning the game.

tallguyindc says:

"Because realistically down by 18 halfway through the 4th quarter the possiblity of a tie or win is very limited anyway. "

True but realistically being down by 15 halfway through the 4th quarter and the possibility of a tie or a win is very limited as well.

Thats the real point on all of this.  Substituting being "down 15" with being "down 2 scores" is intellectually lazy and extremely misleading. 

The bottom line is that there is a huge difference between being down 11 and down 15 and there isn't much difference at all between being down 15 and down 18. 

Being down 11 means I need two TDs and I probably win the game.

Being down 15 means I need two two TDs and I very likely lose anyway.

Go and put the probabilities into Excel if you don't believe me.

Going from 18 to 15 is an improvement but not a substantial one. 

You are much better off gambling for 11.  You'll probably lose but you probably lose with all these scenarios.  At least this way, you have a slightly better lottery ticket.

tallguyindc says:

Did any of the math seep through?!?!  There is a lot to think about. And "two scores" does not always equal "two scores"!!!

 

I didn't watch the second half of the Oklahoma game.  What did Stoops do?

tallguyindc says:

 

All of this is true.  What you really need to do is estimate the probability of all events and run simulations many times from the key decision point.  Obviously, you can't do it at the time but you can do it after the fact.  Post mortems help for future games. 

When you run this decision through a simulator, you'd pick up many more wins from going for it and making it than you'd lose by going for it and missing.   

Whats interesting about this is that announcers etc.  always (and I mean always) bring up the fact that a team would have been tied or winning if they went for it instead of kicking the fg, missed and then lost by 3 or less.    The "anything can happen" disclaimor never seems to apply there.

I have never once heard an announcer bring up the fact that making that "go for it" decision was the reason they won when a team goes for it, makes it and then eventually wins by 4 or less.

Implicitally, the assumptions seem to be that if you kick everything else will go your way later.  If you go, everything else will go against you.  This is just stupid.  Do you make these sort of assumptions when you are capping games?  If you do, thank you.  You're helping to make me big profits. 

tallguyindc says:

Did you love the fact that they went for it....or just the fact they missed?

Remember, the Cowboys ultimately lost by 6. 

Kick a chip shot and lose anyway....go for it and win.

hit_a_parlay says:
I had Giants ML so I LOVED that they didn't kick the chip shot

misfit_aka says:
regardless, u needed the FG, so kick the damn thing and maybe just maybe you can convince ur offense to score a TD, making it easier to then score another one (because of momentum, all caoshces ant every level will tell u is a very real thing)...call reeked of desperation...

mischkin04 says:

 

           tallguy,

    You constantly state the importance and much better odds of 15 vs.11 in your posts. The key number being 11. This leads me to believe a 1 point conversion try is the right choice when down by 12 after scoring.

    Why have you argued that going for 2 early, in that spot, is the proper play then? There are times to be aggresive, but this is not one of them. You always manage to assume there will be time enough to drive the entire field twice if neccesary. I'd see much more probability against that and use my 40% chance of making the 2 pointer to tie later.

   

   

         

RJSizzle says:


Implicitly, the assumption is that if you kick you only need one possession to have a chance to tie the game.

Explicitly, if you dont kick, you need two possessions to tie the game.

I dont make any assumptions while I am capping games.

I mean, I dont see your threads, but I highly doubt I am the one giving you huge profits. If anything, you are the people that get me the lines I do. I love people who play games on paper.

Try this with someone who isnt one of the more successful people at Covers.

RJSizzle says:


You are correct about Phillips. He shouldnt have gone for two until he needed to. Games arent played in classrooms or on paper, math doesnt apply in this situation.

I dont think you can fault him for punting that ball away. Mizzou scores again if they dont get it. The victory looks that much better and it hurts OU in the BCS. That was a move with the big picture in mind. Couple that with OU not being able to complete a pass in the 4th quarter. Stoops was playing for the future.

ApocalypseLater says:
amazes me that people who actually bet money on the games don't know that Wade was a complete imbecile to pass up the FG...

it's a 3 score game, you make it a 2 score game... all your post does is show that someone who is mediocre at math can use it incorrectly in order to appear to make an argument...

your math leaves much to be desired...

ApocalypseLater says:



it's sad to me that it is immediately obvious to everyone (including yourself) that what you have done here has nothing to do with math... it's a fucking joke!

the Cowboys were down by 18 points... you are claiming that in scenario #1, in order to win, they need to register positive on 5 consecutive propositions (if they take the FG)... but somehow, if they complete a pass (not for a TD, but for a first down), that is it!  that's ballgame... 40% and you win the ball game...

you are ridiculous to even be talking about math!  you clearly didn't graduate high school algebra... so get out of here with that crap!

ApocalypseLater says:



the only thing this paragraph is telling me is that it is a complete waste of time to read further...

i don't care what Gruden said... i didn't hear it, it is completely irrelevant, even in light of this thread that YOU created...

what is relevant is whether or not Wade Philips made the right coaching decision when he passed up on the FG... and he did NOT!

period, end o' fucking story!  math, no math, Gruden, no Gruden... none of that shit matters... it was a retarded decision that no NFL coach should ever be guilty of making...

i could literally write a book on how bad of a decision that was, and yet you are here trying to endorse it!

unbelievable!  let this thread go away, dude... say... "my bad!" and fucking drop it!

MYSTICRICH says:

it comes down to time and the number of possession you will have , down 3 scores ,  you have to kick the field goal , stupid not kicking it at that point in the game .

the only reason the giants ended up with only a 6 point win , is that they didnt play the game correctly ...........no team with 3:20 left in the game up by 18 throws the ball , there was no purpose to do that .

gabuchu says:


I think I see what is throwing you off now. You are comparing -11 to -15 straight up. You can't do that because when you go for it on fourth down and not make it then you are at -18. So you have to factor that in.

I may have missed it but I am curious to know what you the the % of winning the game when down 11, 15, and 18.

gabuchu says:


That is not even the same. You're telling me a field goal means folding and going for it means betting. If home team wins 3-0, how did that happen when they only fold and never bet?

gabuchu says:


Really? Everybody is saying your math is flawed not that it is confusing.

Where are you getting the numbers for those events? If you don't have hard numbers with stats to back those up then you're using coming up with a fictitious probability with made up numbers.

It is that difficult. That is why nobody besides you have attempted to do it because it is impossible. It doesn't matter what I think the probability is because what is needed is the actual #s and nobody can provide that.

gamblingallday says:
a couple of things...

andarmac99 says:


The flaw is your argument is that you are making a massive amount of assumptions based on stats that do not take into context how this game has played out. How tired are the players, are the starters in there? Would these %'s change if Romo was in over Kitna? What about the Giants defense? You seem to display the same symptoms that so many hardcore math guys do when betting sports, you don't understand the situation in the game and that they are played by human beings.

1.  At -11, DAL wins 100% of the time.  At -15, DAL wins 24% (48% 2 pt conv * 50% OT win)

What? How do they win 100% of the time at -11? How do you know Dallas will convert the 2pt conversion 48% of the time? That makes ZERO sense just because that is the league average does not mean that is the % it will happen in this situation with these two teams. 50% in OT? Says who? Again what do past games have to do with this situation?

Hey, I absolutely agree with you with other things like Belichick going for it on 4th down, etc. Look at last night's Florida State/NC State game. 4th down NC State has the ball on the 6 inch line, they initially line up for the game tying FG but then go for the touchdown and convert. That was the right move 100% of the time because they were playing to win and still had outs if they stop Florida State because there was enough time left. In the Dallas game time was running out and you need to out yourself in a position to get back in the game. You kick the FG there to make it a two possession game, it's that simple. 2 possessions late in a game is a hell of a lot better off than 3.

Occam's Razor my friend: one should not make more assumptions than the minimum needed.

2 possessions or 3?

theclaw says:

 

Those mathmatical formula's give a very rough idea of the odds of each senerio, but are not accurate.

It seems you read that on a web site, it sounded good to you so you bought into it without alot of further thought.

First, he goes off recent historical averages, which "DO NOT" account for the skill and ability of each opponent in this game. It's flawed already

Second, and the "biggest flaw", is teams "do not" perform at a average level each game, some games they perform well above their avearages, these are games they normally win, then they have games of performing well below their averages, these are games they might lose, that's how stats go in the real world.

Let's look at Dallas's average 3rd down percentage, then let's break it down to a game to game and see how many games  Dallas performed at their average.

I use 3rd down's because it's a larger sample size than 4th down plus, if teams chose to go for it on 4th downs every time as the web site states would be the right play, then they would have a much larger sample size on 4th downs very similar to 3rd downs now.

Dallas season average 39.7 %.

Game 1 --- 38%

2. --- 46%

3.--- 50%

4. --- 50%

5. --- 52%

6. --- 0% 

As you can see by the info above Dallas played at it's average just 1 game out of 6, and they played extremely off their average in 4 of 6 games. This is how stats work in the real world.

Now, how then can putting in a historical average give you an accurate odds of Dallas on "this day" ????

Simply not possible !!!

For that mathmatical formula to be accurate, each team would have to perform at an average level in each situation that makes up the formula, and as the info above clearly without any question shows, it does not work that way in the real world.

In other words, in games that a team is playing above their averages, these are games the team is highly likely to win, when a team is playing above it's average that would increase it's odds  in  the formula if you inserted these above averages into the formula which would be representive of how the team was playing on this day.

And it would be the opposite when teams are playing below the averages on a given day.

Human beings playing a football game are not like a coin, where regardless of how many times a coin comes up heads, it's still 50-50 it comes up heads on the next flip. Coins  not posses skill, statergy, heart or desire and motivation.

A football game involves skill and ability, coaching, energy, heart, motivation and desire on this given day, all those things come together at different levels each game, to "assume" each team plays at a historical average each and every game is silly.

tallguyindc says:

 

You're right.  You need to account for the skill level of each team, but remember this isn't just a decision whether you can make the fourth down.  Its a decision between which is more likely making the fourth down now or making the kick + 2 point conversion + winning in OT.

I explain the logic a little better in this post.

http://www.covers.com/postingforum/post01/showmessage.aspx?spt=21&sub=100864546

If you think the Cowboys are far worse offensively than historical averages than you have to apply that same logic to their chances of converting the 2 and their chances of winning in overtime.  When you do that,  the skill adjustments work out to basically a wash and you should go anyway.

As a matter of fact, I'd argue that the worse a team is the more they should go.  On any one play, the skill differences between teams isn't that great.  The reason better teams typically win is because those skill differences at up over time. 

Think about it this way, if you could play Kobe Bryant in one-on-one, would you rather play first shot wins or would you rather play to 100.  First shot wins you have a legitimate chance.  Playing to 100, he's got you for surel.

 

tallguyindc says:

The way I look at it, you have the field goal.  Those are the Y chips.  Going for it means risking those Y chips.  If you lose, you have nothing, if you win you have 6Y.   Do you go all-in to get a chance at 6Y chips or do you fold to keep the Y chips?  That is the question.

From what I understand, you are saying that by folding you maintain 100% chance of keeping all of your existing chips you should always fold.  If thats true, well 30 out of 32 NFL coaches agree with you.  You won't lose any chips this hand, but you'll be blinded out very soon. 

 

tallguyindc says:

 

I haven't pulled a single number out of anywhere.  I said the Cowboys and Giants would have the same probability of winning in OT because they would have had the same probability of winning in overtime.  The game started with the Cowboys as a 3.5 point favorite.  The starting QB's injury is typically worth about 3-4 points.  That would make the game basically a pickem.  Thats over a whole game.  In a much shorter sudden death overtime, the randomness factor is even higher.  Plus, by that point the two teams would have scored the exact same number of points in regulation.  I have a tough time seeing how you can make any sort of skill level adjustment.  I mean maybe you could argue one team was 51 or 52% but anything more than that is ridiculous.  If it had got to overtime and anybody had offered anything over +115 or so on either team before the coin flip, I would have gladly accepted.

I've also pointed out that skill level adjustments don't have as large of an effect as you would think offhand.  Whatever skill level adjustment, you need to put it in uniformly.  In other words you can't argue the Cowboys are less likely than average to get the touchdown but more likely than average to get the 2 point.  When you do that, the adjustments become mostly a wash and its obvious you shouldn't kick. 

On a more macro level, apparently you enjoy betting on games without performing any form of probability analysis whatsoever.  I think this is insane and I can't imagine its worked out too well for you, but I can only say thank you, because its people like you that keep me in business.

tallguyindc says:

This is a little bit of a Catch-22.  When I try to describe the math, everybody says "slow down you're confusing me".  When I just jump to the conclusions, nobody trusts the numbers. 

Why don't you try to do them yourself? 

If you kick you need to

a) make the kick

b) Hold NY scoreless

c) Score 2 touchdowns

d) Make 2 point conversion

e) Win in OT

I could assign probabilities to each of those events but you wouldn't believe me anyway.  Why don't you assign your own probabilities to them and then multiply them together and tell me what you come up with?

Its not that difficult.  You need all 5 things.  What do you think the probability is on each?

 

tallguyindc says:

 

You got it exactly right.  Except replace the generic "improving" to "more than quadrupling"

in other words, going for it on 4th down gave them a 40% chance of more than quadrupling their chances...

I couldn't have said it any better myself.

 

tallguyindc says:

Andarmac,

I want to keep talking to you about this because I feel like you are a rational person and if you just sat down and thought about it for a little while, you'd see my point of view.  I think there are a couple of specious arguments on around here.

1.  Did you see what I was talking about when I said that  -15 isn't really 2 scores...its actually 5 scores?

2.  I acknowledge the fabric of the game argument, but I think its misapplied.  People are always saying "Anything can happen" and building that into their argument.  But what they usually mean is "anything good can happen when you follow conventional wisdom.  Anything bad can happen when you don't".  They let those implicit assumptions start messing with their logic.  You really kind of have to apply neutral assumptions to everything but the key variables.

3. Did you read the debate about volatility in post #98?  Basically "staying in the game" is highly overrated.  At the end of the season, nobody cares how many games you stayed in.  They care about how many games you won.  When you are down 18, you are going to need to increase the volatility.   Basically, you're going to need to start doing some risky things and you're going to need to get lucky as hell. Kicking is basically saying I'm, going to lower the volatility".  You're going to need to catch a couple breaks along the way.  Might as well start now when they are relatively high probability. 

 

tallguyindc says:

It should be closer to 35%.  The math works at anything above about 14%.  

http://www.advancednflstats.com/2009/09/4th-down-study-part-3.html

 

I really don't want to get into a general fourth down discussion but I agree with him.  You're right the math says that you should generally go for it on 4th and 1 and that you need to apply game-specific factors.  But if on average you should go for it and coaches apply game specifics and never go for it you are quickly approaching Lake Woebegone territory where they proudly boasted that "every child is above the town average"

 

tallguyindc says:

During that whole Belichick thing, I met math people who couldn't believe he was right at first and then ran the numbers to try to refute me and said "wow that surprised me...why don't more coaches go" and I said "Would increasing your probability of winnning the game by 5% be worth listening to a bunch of morons call you a moron?" . 

Last post you said you could "run circles around me in any math field".  You actually got my hopes up.  I've been wanting somebody reasonably smart to come on here and try to argue the other side.

Instead all I got was a bunch of curse words.  Oh well....disappointing.

 

tallguyindc says:

 

Thank you for starting to think about it from a purely math point of view.  The volatility argument is actually pretty relevant.  The reason why NBA teams don't shoot a 3 every shot is game theory.  There isn't a guy open every shot.  To a certain extent, you need to take what the defense gives you.  The "3 only strategy" only works when you actually can convert 40% of the shots.  The NFL analogy is the prevent defense.  I know its hated but there is a logic behind the prevent.  The offense should (and if there was no prevent) would throw a bomb on every play when they are behind.

You mention whether a team should prefer a high volatility or low volatility game.  It brings up a very good point and helps me make my case.  The greater your chance of winning the game, the lower volatility you want it.  If the ref called the coaches and said, "we're bored with this....every play I'm going to award a TD to whoever performs the best" the Cowboys would obviously cheer and the Giants would obviously boo.  There is only one win to go around.  If the Giants want low volatility, the Cowboys want high volatility. 

Bottom line, the lower your chance of winning the more risks you need to take.  You really don't see this in the NFL.  Coaches should be getting desperate a whole lot earlier.  But they just don't want to get blown out so they do incredibly stupid things like kick FGs down 18.  Instead of getting mad at them, most people seem to get mad at the guys that actually try to win. 

Stop and think about it.  You'll see I'm right.  I know I making a big deal of this but its really a bigger issue.   In my opinion, this is the biggest problem in football.  Basically, teams give up too early.  They disguise it by saying they are "just staying in the game" but in truth the coach is just narrowly lowering the losing margin.

When guys like Belichick and Phillips actually think about how to win and challenge the order, we really need to focus on what happened and why.

Volatility is a very good point.  From now on, watch a game and say "If I was Team A would I want high volatility or low" Do the same for the other one.  The answer needs to be the opposite for the other team.

tallguyindc says:

Again, I'd like to see your probabilities.  Being down 15 wasn't really 2 scores.  It was 5 scores. In addition to the 2 touchdowns, you also need the 2 point conversion, your defense to provide a 0 socre to the Giants and a score in overtime

I'm curious what you have as the probabilities on each of these occuring.

You're right. To a certain extent, the Cowboys were let back into the game because the Giants were stupid and they were lucky.

But lets be honest, -11, -15, -18.  You don't come back from that far behind without a little luck.  The question is just how much you need...

tallguyindc says:

 

God after your first post.  I had such high hopes for you.  I really wanted to argue with someone that some understanding of math.

Reread post #91.  Look at it a little closer because I've already answered all your issues.  I did not say at all that it was a certainty that they would score 2 more touchdowns and the Giants would score none.  I just said none of this would matter unless Dallas scored two more TDs and the Giants scored either a FG or nothing.

That should be obvious.  I would think even you would agree with that.  If the Cowboys score once more, does it matter?  If they score 5 times more, does it matter? 

No. It only matters in narrowly defined scenarios so it makes the math easier to just look at those scenarios.

mischkin04 says:

 

               

ApocalypseLater says:


reposting to make sure you don't miss it this time...

ApocalypseLater says:


i already did run circles around you in my previous post... you probably missed it, because shooting down your argument mathematically was extremely easy...

read Andarmac's post, but notice that i already pointed out exactly the same thing...

this is your argument: two scenarios...

1) Dallas kicks the FG (a 98% proposition, according to you), putting them 11 pts behind, if successful... at this point they still need 4 more things to happen in order to win (according to you)...

2) Dallas goes for the 4th down conversion (a 40% proposition, according to you), putting them 15 pts behind (still), if successful... at his point (according to YOU), if they have succeeded on converting the first down... wait for it... THEY WIN!  that's it... even though they are still 15 pts behind, if they get that 40% proposition, they win the game (magically!)...


are you still not seeing where i ripped your argument apart?  i'm not claiming i did it with Differential Calculus... it wasn't needed... basic algebra was all that it required...



you have officially been encircled!

ApocalypseLater says:
this is exactly what your argument boils down to:

IF the Cowboys had gone for it on first down, and succeeded (a proposition that you say is roughly 40%), they would have improved their situation more than if they had just kicked a FG...

in other words, going for it on 4th down gave them a 40% chance of improving their chances...

there is nothing you have said that goes any deeper than that, no matter how much you insist that you are using math here...

ApocalypseLater says:
so i suppose Wade Phillips was also a genius when he... (a) used a challenge flag on a play that, if he had won it, would have left GB on the 1 inch line... (b) called a timeout (final TO) on the next play, so they could save 13 seconds instead of just letting it run until the 2 minute warning...

thus leaving them with no timeouts and no challenge flags...

these are the things that terrible coaches do... and Wade Phillips is the most terrible coach in the NFL... period!

pillyw0nka says:

LOL

listen to yourself tallguy. More than quadruple , 11 is about 6x better than 15.

LOL you just make up numbers in a game where mathematic percentages get beat all the time. A game where you cannot account for failed snaps, picks with 3 min to go. You are a joke, please do us all a favor and never use numbers again.

The fact of the matter is that you still fail to see the subsequent events. THEY DID NOT CONVERT. And if they don't get the touchdown they lose the game there. Do you not understand that?

Wade Phillips gave up, he knew he was cooked. In his dumb ass mind on the field with all the lights shining on him, he made a dumb decision and went for it when the obvious play was to kick the fg and make it 2td game, some people call that TILT.

Another thing that someone mentioned before is that both teams might have played differently being in a 18pt 3td deficet compared to 15pt 2td deficet, how can you account for those variables with your flawed high school probability math?