A few folks really struggling with some basics of win probability.
The point is to win, right? So if you have an opportunity to win the game with one play, and a play that you have a high likelihood of success at (gain 5 yards), then you should gain an equal or better likelihood of success by passing up that opportunity.
This is where Lake fucked it up. Because going up by 6 instead of 3 is pretty marginal in terms of winning. An OSU touchdown still means you lose.
So you give up a chance to win outright, and instead just make it so OSU has to gain appx 30 more yards to beat you. You also give OSU a kickoff return opportunity, which basically guarantees them better field position than the 5.
I think they returned it to the 20, so of the 30 extra yards you make them go to score a touchdown vs a field goal, 15 of it was wiped out instantly anyways (starting on the 20 vs the 5).
Not sure why some fancy model is needed to realize that kicking that field goal from the 5 was a loser move.
This doesn't take into account that OSU would no longer need a TD to beat you. All they would need is a FG to push it to overtime, which negates pretty much everything you typed in the following paragraph. From the 5, they'd need only gain 65 yards to get into very makeable field goal range.
The field goal was as close as a guarantee as you can get. Call it 95%+. What's the success rate on plays from the 5 yard line? I don't know, but I'd guess less than 50%. And forget all of that. In this case, you have the benefit of actually watching the game and using prior performance to predict future performance. OSU was stacking the box and shutting down the run. The three previous plays would suggest that running the ball would be high risk from that distance. The quarterback averaged under six yards per attempt on the game, 3.8 yards per attempt in the redzone (helped greatly by a dumpoff pass to the RB from the 19 and another 13 yard pass from the 19 that was dropped and given back by the stripes), 25% and less than a yard per attempt inside the 10 (a single 3-yard diving completion), and zero passing TDs on the game. Receivers dropped four passes in the game and had another two broken up (both in the end zone), so you had to have confidence in them at that point, right?
You take the guaranteed points and make the opponent go the length of the field.
A few folks really struggling with some basics of win probability.
The point is to win, right? So if you have an opportunity to win the game with one play, and a play that you have a high likelihood of success at (gain 5 yards), then you should gain an equal or better likelihood of success by passing up that opportunity.
This is where Lake fucked it up. Because going up by 6 instead of 3 is pretty marginal in terms of winning. An OSU touchdown still means you lose.
So you give up a chance to win outright, and instead just make it so OSU has to gain appx 30 more yards to beat you. You also give OSU a kickoff return opportunity, which basically guarantees them better field position than the 5.
I think they returned it to the 20, so of the 30 extra yards you make them go to score a touchdown vs a field goal, 15 of it was wiped out instantly anyways (starting on the 20 vs the 5).
Not sure why some fancy model is needed to realize that kicking that field goal from the 5 was a loser move.
This doesn't take into account that OSU would no longer need a TD to beat you. All they would need is a FG to push it to overtime, which negates pretty much everything you typed in the following paragraph. From the 5, they'd need only gain 65 yards to get into very makeable field goal range.
The field goal was as close as a guarantee as you can get. Call it 95%+. What's the success rate on plays from the 5 yard line? I don't know, but I'd guess less than 50%. And forget all of that. In this case, you have the benefit of actually watching the game and using prior performance to predict future performance. OSU was stacking the box and shutting down the run. The three previous plays would suggest that running the ball would be high risk from that distance. The quarterback averaged under six yards per attempt on the game, 3.8 yards per attempt in the redzone (helped greatly by a dumpoff pass to the RB from the 19 and another 13 yard pass from the 19 that was dropped and given back by the stripes), 25% and less than a yard per attempt inside the 10 (a single 3-yard diving completion), and zero passing TDs on the game. Receivers dropped four passes in the game and had another two broken up (both in the end zone), so you had to have confidence in them at that point, right?
You take the guaranteed points and make the opponent go the length of the field.
No. And this is a weird hill you are choosing to die on.
Again, it’s about the trade-off. You are advocating giving up the chance to win the game by gaining 5 yards.
A chance that, even if you fail, as you point out, the beav has to go 65 yards just to *tie.* So if you don’t convert, you can still win by stoping their drive, they miss the field goal, or - worst case - you go to overtime. Because with the time on the clock & no timeouts, the beav would be playing for overtime, anyways. Not for a touchdown,
So if you give up the chance to win the game with one play, you better end up with a scenario for winning that is significantly better than the one above.
But a field goal to go up 6 doesn’t drastically change what the beav needs to do. Like I said, kickoff return negates the yardage they need to get into a scoring situation. And going up 6 also forces them to play for a touchdown, not just a field goal to go to overtime. So they are now using 4 downs, etc. Not settling for a potentially difficult field goal.
Again, having an opportunity to win by getting 5 yards is a phenomenal opportunity. To pass that up, you have to demonstrably improve your opportunity to win. Going from a cushion of 3 to 6 doesn’t to that.
A few folks really struggling with some basics of win probability.
The point is to win, right? So if you have an opportunity to win the game with one play, and a play that you have a high likelihood of success at (gain 5 yards), then you should gain an equal or better likelihood of success by passing up that opportunity.
This is where Lake fucked it up. Because going up by 6 instead of 3 is pretty marginal in terms of winning. An OSU touchdown still means you lose.
So you give up a chance to win outright, and instead just make it so OSU has to gain appx 30 more yards to beat you. You also give OSU a kickoff return opportunity, which basically guarantees them better field position than the 5.
I think they returned it to the 20, so of the 30 extra yards you make them go to score a touchdown vs a field goal, 15 of it was wiped out instantly anyways (starting on the 20 vs the 5).
Not sure why some fancy model is needed to realize that kicking that field goal from the 5 was a loser move.
This doesn't take into account that OSU would no longer need a TD to beat you. All they would need is a FG to push it to overtime, which negates pretty much everything you typed in the following paragraph. From the 5, they'd need only gain 65 yards to get into very makeable field goal range.
The field goal was as close as a guarantee as you can get. Call it 95%+. What's the success rate on plays from the 5 yard line? I don't know, but I'd guess less than 50%. And forget all of that. In this case, you have the benefit of actually watching the game and using prior performance to predict future performance. OSU was stacking the box and shutting down the run. The three previous plays would suggest that running the ball would be high risk from that distance. The quarterback averaged under six yards per attempt on the game, 3.8 yards per attempt in the redzone (helped greatly by a dumpoff pass to the RB from the 19 and another 13 yard pass from the 19 that was dropped and given back by the stripes), 25% and less than a yard per attempt inside the 10 (a single 3-yard diving completion), and zero passing TDs on the game. Receivers dropped four passes in the game and had another two broken up (both in the end zone), so you had to have confidence in them at that point, right?
You take the guaranteed points and make the opponent go the length of the field.
No. And this is a weird hill you are choosing to die on.
Again, it’s about the trade-off. You are advocating giving up the chance to win the game by gaining 5 yards.
A chance that, even if you fail, as you point out, the beav has to go 65 yards just to *tie.* So if you don’t convert, you can still win by stoping their drive, they miss the field goal, or - worst case - you go to overtime. Because with the time on the clock & no timeouts, the beav would be playing for overtime, anyways. Not for a touchdown,
So if you give up the chance to win the game with one play, you better end up with a scenario for winning that is significantly better than the one above.
But a field goal to go up 6 doesn’t drastically change what the beav needs to do. Like I said, kickoff return negates the yardage they need to get into a scoring situation. And going up 6 also forces them to play for a touchdown, not just a field goal to go to overtime. So they are now using 4 downs, etc. Not settling for a potentially difficult field goal.
Again, having an opportunity to win by getting 5 yards is a phenomenal opportunity. To pass that up, you have to demonstrably improve your opportunity to win. Going from a cushion of 3 to 6 doesn’t to that.
Not enough fuck offs to give here. The Beavs getting screwed on both 3rd and 4th down saved the game for #mydawgs
And it didn’t get any better from there with another non call after that.
No. And this is a weird hill you are choosing to die on.
It's no more of a hill to die on for me than it is to you. And my quarrel was never with you in the first place. It was with a young poster who came across as a cocky dick without having anything to back it up. It's the combination of these two things that bothered me. I don't mind being cocky and willing to back it up. I don't mind being wrong and open minded. I don't mind if I'm wrong when the other side of the argument isn't a dick about it. But if you're going to call out peoples' understanding of statistics in an appeal to authority to back up your shitty argument, you'd better be ready to produce said statistics.
Unlike him, I actually checked my work, finding a couple of these mythical win probability calculators online, entering the situational information, and having my argument confirmed (admittedly, by a razor's edge).
So whatever your (his) opinion, I'm fine with being proven wrong, especially by somebody like you who's framed it in a more constructive way and not, "If you mouthbreaters understood statistics, you'd agree with me" kind of way.
Two small things I wanted to add to this, and wasn't sure where to put them.
Dylan Morris isn't very good at selling the play action. I'm not expecting perfection, but it's pretty obvious within a split second that it's PA and not a run based on his effort. He barely sticks the ball out and immediately pulls it back in. I'm not sure if it is meaningful in the grand scheme of things, but it seems like something that should be looked at.
Our win expectancy at the end of the game was 98% at a site I was looking at. The metrics suggest we should have won by a lot more, but our WRs can't catch and our special teams are an abomination. Improve both of those two areas even a little and we win this game going away.
Best defense is the conference rather easily. #MyAnalytics showed that UW was better than ducks rather easily. I love maff
You needed analytics to tell you that a defense that gave up 29 points to Rolo and 35 to the ghost of Chip Kelly is bad?
No. Everyone knows their defense sucks. They just aren’t a good overall team. They have severe deficiencies in the pass game, Shough has 14 turnover worthy plays in 4 games.
Best defense is the conference rather easily. #MyAnalytics showed that UW was better than ducks rather easily. I love maff
You needed analytics to tell you that a defense that gave up 29 points to Rolo and 35 to the ghost of Chip Kelly is bad?
No. Everyone knows their defense sucks. They just aren’t a good overall team. They have severe deficiencies in the pass game, Shough has 14 turnover worthy plays in 4 games.
Best defense is the conference rather easily. #MyAnalytics showed that UW was better than ducks rather easily. I love maff
You needed analytics to tell you that a defense that gave up 29 points to Rolo and 35 to the ghost of Chip Kelly is bad?
No. Everyone knows their defense sucks. They just aren’t a good overall team. They have severe deficiencies in the pass game, Shough has 14 turnover worthy plays in 4 games.
Want a cool DWAG stat?
McDuffie, Keith, Kyler, Elijah, Asa, and Cook have allowed a combined 65 passing yards when targeted in 2 games.
Best defense is the conference rather easily. #MyAnalytics showed that UW was better than ducks rather easily. I love maff
You needed analytics to tell you that a defense that gave up 29 points to Rolo and 35 to the ghost of Chip Kelly is bad?
No. Everyone knows their defense sucks. They just aren’t a good overall team. They have severe deficiencies in the pass game, Shough has 14 turnover worthy plays in 4 games.
Want a cool DWAG stat?
McDuffie, Keith, Kyler, Elijah, Asa, and Cook have allowed a combined 65 passing yards when targeted in 2 games.
I'll take that and raise you one: ZTF has allowed ZERO when targeted!
Best defense is the conference rather easily. #MyAnalytics showed that UW was better than ducks rather easily. I love maff
You needed analytics to tell you that a defense that gave up 29 points to Rolo and 35 to the ghost of Chip Kelly is bad?
No. Everyone knows their defense sucks. They just aren’t a good overall team. They have severe deficiencies in the pass game, Shough has 14 turnover worthy plays in 4 games.
Want a cool DWAG stat?
McDuffie, Keith, Kyler, Elijah, Asa, and Cook have allowed a combined 65 passing yards when targeted in 2 games.
I'll take that and raise you one: ZTF has allowed ZERO when targeted!
Eddie has allowed 2 catches for 16 yards on 5 targets with 2 pbus, and a 48.8 rating against.
Comments
The field goal was as close as a guarantee as you can get. Call it 95%+. What's the success rate on plays from the 5 yard line? I don't know, but I'd guess less than 50%. And forget all of that. In this case, you have the benefit of actually watching the game and using prior performance to predict future performance. OSU was stacking the box and shutting down the run. The three previous plays would suggest that running the ball would be high risk from that distance. The quarterback averaged under six yards per attempt on the game, 3.8 yards per attempt in the redzone (helped greatly by a dumpoff pass to the RB from the 19 and another 13 yard pass from the 19 that was dropped and given back by the stripes), 25% and less than a yard per attempt inside the 10 (a single 3-yard diving completion), and zero passing TDs on the game. Receivers dropped four passes in the game and had another two broken up (both in the end zone), so you had to have confidence in them at that point, right?
You take the guaranteed points and make the opponent go the length of the field.
Again, it’s about the trade-off. You are advocating giving up the chance to win the game by gaining 5 yards.
A chance that, even if you fail, as you point out, the beav has to go 65 yards just to *tie.* So if you don’t convert, you can still win by stoping their drive, they miss the field goal, or - worst case - you go to overtime. Because with the time on the clock & no timeouts, the beav would be playing for overtime, anyways. Not for a touchdown,
So if you give up the chance to win the game with one play, you better end up with a scenario for winning that is significantly better than the one above.
But a field goal to go up 6 doesn’t drastically change what the beav needs to do. Like I said, kickoff return negates the yardage they need to get into a scoring situation. And going up 6 also forces them to play for a touchdown, not just a field goal to go to overtime. So they are now using 4 downs, etc. Not settling for a potentially difficult field goal.
Again, having an opportunity to win by getting 5 yards is a phenomenal opportunity. To pass that up, you have to demonstrably improve your opportunity to win. Going from a cushion of 3 to 6 doesn’t to that.
And it didn’t get any better from there with another non call after that.
Unlike him, I actually checked my work, finding a couple of these mythical win probability calculators online, entering the situational information, and having my argument confirmed (admittedly, by a razor's edge).
So whatever your (his) opinion, I'm fine with being proven wrong, especially by somebody like you who's framed it in a more constructive way and not, "If you mouthbreaters understood statistics, you'd agree with me" kind of way.
Dylan Morris isn't very good at selling the play action. I'm not expecting perfection, but it's pretty obvious within a split second that it's PA and not a run based on his effort. He barely sticks the ball out and immediately pulls it back in. I'm not sure if it is meaningful in the grand scheme of things, but it seems like something that should be looked at.
Our win expectancy at the end of the game was 98% at a site I was looking at. The metrics suggest we should have won by a lot more, but our WRs can't catch and our special teams are an abomination. Improve both of those two areas even a little and we win this game going away.
Against UW
Total yards - 252
Points -14
Gebbia - 85 yards, 1 int
Jefferson- 133, 1 td
Against Coog
Yards - 451
Points - 28
Gebbia - 329, 1 td
Jefferson - 120, 3 td
Against Cal
Yards - 360
Points- 31
Gebbia- 147, 1 td (3 total), 2 int
Jefferson - 196, 1 td
Against Oregon
Yards - 532 (Holy shit)
Points - 41
Gebbia - 263, 1 td (2 total)
Jefferson - 226, 2 td
McDuffie, Keith, Kyler, Elijah, Asa, and Cook have allowed a combined 65 passing yards when targeted in 2 games.