Suspicious Decision Making By John Harbaugh Last Night SNF

[KVB]

It was 4th and inches so it's like 90% chance they convert maybe even higher

Dude, wtf. More like ~ 60%, and it's actually less near the goal line. Easier in the middle of the field because the defense has more ground to cover. You go for it if you are trying to maximize points, but when it's late the score comes into play.

I want to go into this a little bit.

I've done some extensive work in this specific field and at the outset I will say that relatively recent available data will show that it matters whether it's 4th and inches 4th and 1 or even 4th and 2. This distance matters more than where you are on the field and even the discrepency in teams, although all can play a role.

For just overall data, without getting into team specifics or field position I will offer that 4th and 2 (+/- .5 yards) is more like the 60%er and 4th and 1 (+/- .5 yards) is more like 70% to 75%.

For 4th and inches (that would be .5 yards or less) the probability is more like 75% to 80%.

Now, years ago when we were looking at this stuff originally (when it was just 4th and 1 or 2) I argued that large numbers were a starting point and that we should use team comparisons to help narrow it down even further for specific games. I argued we should use our methods for using the closing spreads as a decent team comparison. I further argued that we could incorporate the Totals as well, similar to how us higher level analysts work with push rates, and also to imitate what the books may be using as a comparison. I then said we could make it better by and even using our own rating systems, which can be more predictive or as predictive as the market.

It turns out that the NextGen stats guys (powered by AWS) who help prepare the analytics in conjunction with modern NFL stats created machine learning models and in order to hone them began incorporating the betting lines in their team comparisons.

lol.

Also, when it comes to analytics remember that the formula is not just the chance the play succeeds, there has to be a probability of winning the game, a current at that moment probablity, and how the success or failure of each decision (whether to go for it or kick, etc.) affects those probabilties that determines the ultimate anlaytic decision or final answer of best probablity to WIN the game.

In the end, it becomes about what the best probability to win the game will be and the idea is that a number of small decisions at each moment, with the info avaliable at each moment, that may improve your probability to win in that moment will payoff overall, with the big picture, in the end. This philosophy expands to not just within the game, but within a season or set of games or even seasons.

I have many opinions on all of this, the application of it, and and how we can best use it all to serve us.

When I see these guys in the booth paid to just analyze those probabilties and are likely just dialing into next gen stats and even AWS, sometimes I wonder. Sometime I wonder how their betting accounts are doing...lol.

I could do their job, could they do mine?

[TheMoneyShot]

You're reading way too much into it. Harbaugh probably just had last week's (and last seasons) game(s) on his mind. 4th and inches should be automatic for him but I suspect he didn't want ANOTHER one of those to go against him after last week and the 10 or so times he went for it and failed with the game on the line last season

When Baltimore went for it on 4th down VS Buffalo... I said NO NO NO in my mind. I was very surprised Harbaugh went for it. To me that was a perfect situation to kick the FG. Again, why does Harbaugh believe Jackson is god at times??? I don't know. Are we going to play psychologist and look into emotions of what we believe Harbaugh is thinking from one week to another? Or are we just going to look at simple play calling? Harbaugh doesn't give a sh## what anyone thinks of him. I'm only suggesting that his decision making is suspicious.

I don't believe I'm looking way into it. This new "Analytics" word... is simply a smoke screen for what's really being done with the spread involved. A simple cover up. Just like Andy Reid going for a 2pt conversion tonight... going up by 9??? lol Wait... what??? You don't kick the Extra Point.... and go up by 8??? 8 isn't a good number anymore??? Oh someone's defense is... they don't have a good kicker?? Then why was he hired for??? Like I said... no coach goes for a 2 pt conversion to go up 9.... when the tragedy of it all... you're only up by 7.

Like I said.. Analytics is a new word just to say... "OH!! That's why he did it. Analytics said to do it." No... it's because it helps Vegas in certain situations.

When John Harbaugh had a Prime Flacco.... wow... Harbaugh was the most conservative HC in the NFL. He would always KICK the FG 4th and 1. He probably lead all coaches NOT GOING FOR IT. But mysteriously.... "Analytics" tells you HOW TO LOSE A FOOTBALL GAME. Yep, it certainly does.

[TheMoneyShot]

It's all about money. Yes, what Vegas wants it.

If you looked at the Chargers-Browns game in which Chargers went for 4th down in their own territory during the last minute of 4q while leading by 2 pts, how could it be possible? They wanted Browns to win the game.
No wonder, why Vegas accepts 1million bets for NFL.

I watched the highlights of that game.... and there's truly no excuse going for it on 4th down. I don't even know how that HC has a job this week. The only reason to go for it in that situation would be purely gambling related. Every coach would punt in that situation.

[TheMoneyShot]

I want to go into this a little bit.

I've done some extensive work in this specific field and at the outset I will say that relatively recent available data will show that it matters whether it's 4th and inches 4th and 1 or even 4th and 2. This distance matters more than where you are on the field and even the discrepency in teams, although all can play a role.

For just overall data, without getting into team specifics or field position I will offer that 4th and 2 (+/- .5 yards) is more like the 60%er and 4th and 1 (+/- .5 yards) is more like 70% to 75%.

For 4th and inches (that would be .5 yards or less) the probability is more like 75% to 80%.

Now, years ago when we were looking at this stuff originally (when it was just 4th and 1 or 2) I argued that large numbers were a starting point and that we should use team comparisons to help narrow it down even further for specific games. I argued we should use our methods for using the closing spreads as a decent team comparison. I further argued that we could incorporate the Totals as well, similar to how us higher level analysts work with push rates, and also to imitate what the books may be using as a comparison. I then said we could make it better by and even using our own rating systems, which can be more predictive or as predictive as the market.

It turns out that the NextGen stats guys (powered by AWS) who help prepare the analytics in conjunction with modern NFL stats created machine learning models and in order to hone them began incorporating the betting lines in their team comparisons.

lol.

Also, when it comes to analytics remember that the formula is not just the chance the play succeeds, there has to be a probability of winning the game, a current at that moment probablity, and how the success or failure of each decision (whether to go for it or kick, etc.) affects those probabilties that determines the ultimate anlaytic decision or final answer of best probablity to WIN the game.

In the end, it becomes about what the best probability to win the game will be and the idea is that a number of small decisions at each moment, with the info avaliable at each moment, that may improve your probability to win in that moment will payoff overall, with the big picture, in the end. This philosophy expands to not just within the game, but within a season or set of games or even seasons.

I have many opinions on all of this, the application of it, and and how we can best use it all to serve us.

When I see these guys in the booth paid to just analyze those probabilties and are likely just dialing into next gen stats and even AWS, sometimes I wonder. Sometime I wonder how their betting accounts are doing...lol.

I could do their job, could they do mine?

Great post KVB. All I'm saying is... no matter how logical... Star Trek... Dr. Spock you are with your interpretation of what you should wager... you all should of been a millionaire by now. The reason why you aren't.... because your PURE LOGIC is being destroyed by a FIX in the equation. The FIX is phantom holds... and illogical decisions by head coaches. But, we'll continue to say it's Murphy's Law or some sh## to make it believable in our minds. Yeah.

[jackpot269]

Why is everything here taken out of context? Kicking for the 3 is great but I think they had just got 1st and 10 if it was 4th down sure kick it! Greatest kicker in the history of football! Ok how about the snapper also the best all time? No way a kick ever gets blocked?

Manning and the colts (wish I could remember the year ) did same on Sunday night once with even more time on the clock with a very good kicker (not the best all time tho) they could have run 5 or 6 plays with a 1st but instead just ran it down to 3 sec and kicked it, wide right and lost, I think it cost them the number 1 Seed that year. This work out this time, but you are giving yourself 1 option in stead of 2. When you have time on the clock and downs to work with

[KVB]

Great post KVB. All I'm saying is... no matter how logical... Star Trek... Dr. Spock you are with your interpretation of what you should wager... you all should of been a millionaire by now. The reason why you aren't.... because your PURE LOGIC is being destroyed by a FIX in the equation. The FIX is phantom holds... and illogical decisions by head coaches. But, we'll continue to say it's Murphy's Law or some sh## to make it believable in our minds. Yeah.

Ahh, but I never mentioned anything about it being fixed or not. How do you know "my logic" is being destroyed by a FIX in the equation? Long time FlyMe readers might disagree with you there. I purposely skipped addressing that topic, the topic, of this thread.

Notice I said "I have many opinions on all of this, the application of it, and and how we can best use it all to serve us."

I've given plenty of posts on that topic, even if some leave the reader to come to the obvious answer. I get the feeling my answers and conclusions, much of it math, results, and chart based, isn't what you think it is. In fact, it's probably far more extreme than...

...In Plato's Parable of the Cave you may be so blinded you turn around and go back inside, lol.

...

...

I wrote more, twice, but deleted it both times.

Like I said, readers might be disappointed in my answers..

...

...

[TheMoneyShot]

And someone please tell me why Josh McDaniels goes for the 2 point conversion after the TD late in the 4th quarter?

What does Analytics say to a players emotions if YOU DON'T CONVERT??? And what does it actually change systematically IF YOU CONVERT? Anyone???

Even if you convert the 2 pt conversion... and you're winning the game by 1 point. Mahomes has all the time in the world to get into FG range... and win the game. Getting the 2 pt conversion DOESN'T WIN YOU THE GAME. It doesn't change a player's mindset that they will STOP Mahomes. You were winning the game by 17 points earlier... what does 1 point mean to you? Nothing. What happened to your damn lead?

Just kick the Extra Point... and try to play solid defense. Did we lose the meaning to "grinding out" a game??? Like... I don't believe the sh## we're watching. This isn't a football game... this is strictly entertainment.... all the moves coaches are making are "gambling related".

You only go for a 2 point conversion extremely late in a game... like under 20 seconds or so.... and you make a "coach's decision" do I want to win it right here... or lose... do or die. Who the hell goes for a 2 pt conversion to take the lead with 4 minutes and 27 seconds to go in a game??? Worse case scenario you are still LOSING THE GAME... to a HIGHLY EXPLOSIVE OFFENSE. Get the fukk outta here with this. Why not give yourself a fighting chance... kick the extra point... and TIE IT UP! Your team has fought awfully hard to just deflate them with an unsuccessful try. Again, even if you convert... WHAT THE FUKK HAVE YOU WON??? There's almost a third left of a quarter. You haven't won sh##! This isn't analytics. This is all gambling related.

[KVB]

Yeah, many of us understand that do or die situation, not wanting to go to OT with KC even with a kicker advantage, but do or die with 4 1/2 minutes left seems a little strange.

Going for the win and getting out of a hostile road environment makes sense, but that’s usually for the end of the game, not five minutes left when there could still be a couple of scores.

[d2bets]

Yeah, many of us understand that do or die situation, not wanting to go to OT with KC even with a kicker advantage, but do or die with 4 1/2 minutes left seems a little strange.

Going for the win and getting out of a hostile road environment makes sense, but that’s usually for the end of the game, not five minutes left when there could still be a couple of scores.

Agreed. I can see doing with 10 seconds left, but 4.5 minutes was not a good scenario. They were fortunate to get another shot. If you tie it, you put KC in a bit of a trick bag, wanting to score but not wanting to give it back with time for Raiders to score.

[asiagambler]

Like I said... no coach goes for a 2 pt conversion to go up 9.... when the tragedy of it all... you're only up by 7.

Pete Carroll did the exact same thing against New England

[asiagambler]

Yeah, many of us understand that do or die situation, not wanting to go to OT with KC even with a kicker advantage, but do or die with 4 1/2 minutes left seems a little strange.

Going for the win and getting out of a hostile road environment makes sense, but that’s usually for the end of the game, not five minutes left when there could still be a couple of scores.

He did it because he felt his defense had a better chance of making a stop if KC is looking to run the clock (ie run, run, pass on 3rd and long). KC was going up and down the field all 2nd half

[d2bets]

He did it because he felt his defense had a better chance of making a stop if KC is looking to run the clock (ie run, run, pass on 3rd and long). KC was going up and down the field all 2nd half

So, he was banking on missing it and then getting it back -- because KC had been gaining yards easy? That's your story?
And what if they make it. And then KC goes up and down the field to win anyway?
If they tie it, then KC also has to worry about the clock.
Only explanation is that he was quite confident (65%+) they'd convert.
I think it's gotten to the point where some coaches are going for it to "look smart".

[bonzaii]

It's all about money. Yes, what Vegas wants it.

If you looked at the Chargers-Browns game in which Chargers went for 4th down in their own territory during the last minute of 4q while leading by 2 pts, how could it be possible? They wanted Browns to win the game.
No wonder, why Vegas accepts 1million bets for NFL.

lol. They went for it because they wanted to end the game and they don't trust their defense.

[asiagambler]

So, he was banking on missing it and then getting it back -- because KC had been gaining yards easy? That's your story?
And what if they make it. And then KC goes up and down the field to win anyway?
If they tie it, then KC also has to worry about the clock.
Only explanation is that he was quite confident (65%+) they'd convert.
I think it's gotten to the point where some coaches are going for it to "look smart".

Where did I say he was banking on missing it ?

You are incredibly dense

If they kick the extra point then it's a tie game. Then KC marches straight down the field, gets in the red zone runs the clock down to near zero and likely kicks the game winning FG. That's called time management. There is not really pressure on the kicker because if he misses, then game goes to OT.

If they go for two and make it, then it's a little different. Yes KC still probably marches down the field and gets in the red zone but do they really feel confident in kicking a field goal as time expires with their backup kicker? If they get it into chipshot territory then it's still probably all the same but it's different being up 1 here as opposed to a tie game. Once in the red zone, KC might elect to go for a TD instead in which case there's likely going to be time left for Oakland to respond.

If they go for two and miss, then as I said McDaniel probably felt his defense had a better chance making a stop in this game situation as opposed to in a tie game. It seemed to work as KC was going up and down the field all 2nd half and yet they were stopped here after 1 1st down

So I'm not even giving my opinion on which scenario is preferable (I'll leave that up to you). I'm just giving my opinion on why I think McDaniels decided to go for two

That's as simple as I can make it for you. I don't even know why I bothered with a long response given that you think 4th and inches has a 60% conversion rate

[bonzaii]

These coaches get paid so much money. The chance of a coach throwing a game is astronomical.

[d2bets]

Where did I say he was banking on missing it ?

You are incredibly dense

If they kick the extra point then it's a tie game. Then KC marches straight down the field, gets in the red zone runs the clock down to near zero and likely kicks the game winning FG. That's called time management. There is not really pressure on the kicker because if he misses, then game goes to OT.

If they go for two and make it, then it's a little different. Yes KC still probably marches down the field and gets in the red zone but do they really feel confident in kicking a field goal as time expires with their backup kicker? If they get it into chipshot territory then it's still probably all the same but it's different being up 1 here as opposed to a tie game. Once in the red zone, KC might elect to go for a TD instead in which case there's likely going to be time left for Oakland to respond.

If they go for two and miss, then as I said McDaniel probably felt his defense had a better chance making a stop in this game situation as opposed to in a tie game. It seemed to work as KC was going up and down the field all 2nd half and yet they were stopped here after 1 1st down

So I'm not even giving my opinion on which scenario is preferable (I'll leave that up to you). I'm just giving my opinion on why I think McDaniels decided to go for two

That's as simple as I can make it for you. I don't even know why I bothered with a long response given that you think 4th and inches has a 60% conversion rate

They would have to manage that perfectly to do that. 4.5 minutes and Balt had 3 TO's plus the 2 minute warning. And if that's your thinking, they can do the same from down 1. That's what Balt did Sunday night with less time.

4th and 1 is almost 70% is NON-goal line situations. In goal line situations, it's actually a fair bit harder because the D has less ground to cover and can jam. It's definitely less than 70% and nowhere close to 90%.

[asiagambler]

They would have to manage that perfectly to do that. 4.5 minutes and Balt had 3 TO's plus the 2 minute warning. And if that's your thinking, they can do the same from down 1. That's what Balt did Sunday night with less time.

4th and 1 is almost 70% is NON-goal line situations. In goal line situations, it's actually a fair bit harder because the D has less ground to cover and can jam. It's definitely less than 70% and nowhere close to 90%.

Not really that perfectly. For a team like KC that was getting anything they wanted in the 2nd half, they can easily kill 4.5 minutes with 5-6 yard gains at a time and staying in bounds. Should not be hard to imagine. One 1st down in the red zone and it's game over. Of course they can do the same from down 1 but it's different than a tie game. I'm not really going to repeat myself because I explained all this already.

I'm not even going to look at your made up numbers for 4th and 1 (which are also probably wrong). 4th and inches is close to 90% conversion rate. It's not definitely less than 70%. You would literally have to be a lunatic to think that

[TheMoneyShot]

Pete Carroll did the exact same thing against New England

If you are referring to Tom Brady with New England.... the only thing I could think of... Pete Carroll was trying something totally out of the ordinary to shut down Tom Brady. Like... ok... 1 possession isn't enough on this guy... I need to be up 2 possessions. But for Reid to do it last night on MNF... just totally insane. His kicker made a 59 yarder... but he can't make a 33 yard extra point?? It's Derek Carr the tooth fairy. Being up 8 is pretty solid.

[asiagambler]

If you are referring to Tom Brady with New England.... the only thing I could think of... Pete Carroll was trying something totally out of the ordinary to shut down Tom Brady. Like... ok... 1 possession isn't enough on this guy... I need to be up 2 possessions. But for Reid to do it last night on MNF... just totally insane. His kicker made a 59 yarder... but he can't make a 33 yard extra point?? It's Derek Carr the tooth fairy. Being up 8 is pretty solid.

Well I was just saying there's precedent for it. I can understand their thinking and it's not like Derek Carr is a complete bum and he in fact went straight down the field and got a touchdown. Both games actually were quite similar. Both failed the two point conversion to make it a 9 point lead but both ended up winning the game. In Seattle's case, because they were able to make a goal line stop and I distinctly remember the final New England play Brady tried to get it to Gronk in the end zone and they called for a flag but refs let it go