2015 Packers notes

[Deleted member 6794]

If I remember correctly, McGinn had him leading the DL with most pressures per snaps. But IMO he has to increase his "splash" plays - TFL and sacks. The sportingcharts site had him playing about 29% of snaps, he should be earning more snaps considering how often Capers uses the sub packages.
http://www.sportingcharts.com/nfl/stats/player-defensive-snap-count-stats/2014/

PFF credited Jones with a total of 19 pressures on 213 pass rush snaps (one every 11.2) which only trailed Daniels (one every 8.6). I agree that he has to make more splash plays though.

 

[HardRightEdge]

Demovsky has compiled one-on-one pass rush scoring. Given that he's enlisted a data scientist for help, there's some kind of black box math involved, though I can't imagine what it might be. A level-of-competition weighting? Anyhow, more grist for the mill:

http://espn.go.com/blog/green-bay-p...-are-telling-for-packers-linemen-pass-rushers

 

[thisisnate]

Any black box that spits out a player who is 0-7 in a drill above someone that is 8-5 is broken.

 

[Deleted member 6794]

Demovsky has compiled one-on-one pass rush scoring. Given that he's enlisted a data scientist for help, there's some kind of black box math involved, though I can't imagine what it might be. A level-of-competition weighting? Anyhow, more grist for the mill:

http://espn.go.com/blog/green-bay-p...-are-telling-for-packers-linemen-pass-rushers

Here´s some information about the formula used by Demovsky to rank the players in this drill:

In order to better put those records into perspective, reader Luke Stanke -- a Green Bay native and graduate student at the University of Minnesota who has an interest in statistical analysis -- offered to devise a formula, similar to the Ratings Percentage Index (RPI) used in college basketball. The formula ranks the players involved in the drill based on my subjective results of each head-to-head matchup. Stanke has two metrics: a regular RPI and a weighted RPI.

He explained them this way: “The regular RPI is one-quarter winning percentage plus one-half opponent winning percentage and one-quarter opponents’ opponents winning percentage.” The weighted RPI, he said, is “a bit more complex, but the interpretation is the same as RPI.” He said the weighted RPI “does the best job of predicting the winner of a matchup.”

Any black box that spits out a player who is 0-7 in a drill above someone that is 8-5 is broken.

The one-on-one pass rushing drill clearly favors the offensive linemen. Therefore it would be better to rank offensive and defensive players in separate lists.

 

[Deleted member 6794]

Here´s McCarthy talking about the Packers´ approach on going for one or two points after a TD entering the season:

"It's not philosophically how we're going to approach it," McCarthy said, "but how we think other people are going to approach it. I think it's only natural you'll see more 2-point conversion opportunity just because of the personality of certain people. If you do it strictly by the numbers, you should really – from a risk assessment standpoint – should still kick the extra point. But going to the 1 adds a whole other element. That takes the numbers to a different level.

I´m kind of surprised about that statement by the head coach, who told reporters during his pre-camp press conference that he spent a ton of time analyzing various statistics in the offseason. Over the last five seasons teams made 94.1% of the field goals from the 15-yard line (expected point average of 0.941) while converting exactly 50% of the two-point conversions (EPA of 1.000).

It took me a total of five minutes to figure that one out, I expect the head coach to do better than that.

 

[PikeBadger]

Demovsky has compiled one-on-one pass rush scoring. Given that he's enlisted a data scientist for help, there's some kind of black box math involved, though I can't imagine what it might be. A level-of-competition weighting? Anyhow, more grist for the mill:

http://espn.go.com/blog/green-bay-p...-are-telling-for-packers-linemen-pass-rushers

Interesting that so far, Walker and Gaston are doing quite well. I hope we see further updates on this chart as the pre-season progresses.

 

[HardRightEdge]

The one-on-one pass rushing drill clearly favors the offensive linemen. Therefore it would be better to rank offensive and defensive players in separate lists.

It better favor the OL. Imagine what a game would look like if any group of pass rushers won anything close to half against any O-Line.

 

[Deleted member 6794]

Interesting that so far, Walker and Gaston are doing quite well.

Taking a closer look at the rankings it seems both Walker and Gaston have faced pretty easy competition in the drill so far.

 

[Deleted member 6794]

It better favor the OL. Imagine what a game would look like if any group of pass rushers won anything close to half against any O-Line.

That's why it would make sense to have seperate rankings for offensive and defensive players.

 

[HardRightEdge]

Here´s McCarthy talking about the Packers´ approach on going for one or two points after a TD entering the season:

"It's not philosophically how we're going to approach it," McCarthy said, "but how we think other people are going to approach it. If you do it strictly by the numbers, you should really – from a risk assessment standpoint – should still kick the extra point.

Of course. The miss rate on FGs from that distance in recent history history is quite low. I doubt a 5-10% higher chance of missing the EP kick will change many coaches' thinking.

 

[vince]

I´m kind of surprised about that statement by the head coach, who told reporters during his pre-camp press conference that he spent a ton of time analyzing various statistics in the offseason. Over the last five seasons teams made 94.1% of the field goals from the 15-yard line (expected point average of 0.941) while converting exactly 50% of the two-point conversions (EPA of 1.000).

It took me a total of five minutes to figure that one out, I expect the head coach to do better than that.

Perhaps you should have taken another five minutes to look up the actual success rate for two-point conversions.

2014 28-59 47.5%
2013 33-69 47.8%
2012 29-58 50.0%
2011 23-50 46.0%
2010 26-53 49.0%

Add in an expected negative point impact of fumble and interception returns on competitive two-point attempts, as well as blocked kick returns (likely close to nil on such short kicks), and I don't know what the real formula shows but I suspect the coach knows what he's talking about regarding their analysis on the new rule.

 

[HardRightEdge]

Any black box that spits out a player who is 0-7 in a drill above someone that is 8-5 is broken.

If an OT loses 5 to third stringers with poor scores while a DE wins none against the high scoring starters, I'd conclude that both have performed poorly in an indistinguishable way. In this case, the rating has them separated by a negligible 0.01. If that is in fact the scenario, the comparative ratings make sense to me.

Demovsky does not explain "SOS". I suspect it's a strength of opponent factor, but that's just a guess.

 

[HardRightEdge]

That's why it would make sense to have seperate rankings for offensive and defensive players.

It appears the RSI adjusts for the offensive advantage since the distribution of offensive players and defensive players is fairly equal.

 

[Deleted member 6794]

Of course. The miss rate on FGs from that distance in recent history history is quite low. I doubt a 5-10% higher chance of missing the EP kick will change many coaches' thinking.

It should though as the expected points average going for two is higher compared to kicking the extra point.

Perhaps you should have taken another five minutes to look up the actual success rate for two-point conversions.

2014 28-59 47.5%
2013 33-69 47.8%
2012 29-58 50.0%
2011 23-50 46.0%
2010 26-53 49.0%

Add in an expected negative point impact of fumble and interception returns on competitive two-point attempts, as well as blocked kick returns (likely close to nil on such short kicks), and I don't know what the real formula shows but I suspect the coach knows what he's talking about regarding their analysis on the new rule.

Even if your numbers should be correct (I got mine from Pro Football Reference's play index) the expected point average is still higher going for two (48.1%, 0.962 EPA) than kicking the extra point (0.941 EPA).

There's no reason to worry about extra point kicks or two-point conersations being returned for two points as it happened only five times on a total of 3,632 such regular plays since 1998.

 

[HardRightEdge]

Perhaps you should have taken another five minutes to look up the actual success rate for two-point conversions.

2014 28-59 47.5%
2013 33-69 47.8%
2012 29-58 50.0%
2011 23-50 46.0%
2010 26-53 49.0%

Add in an expected negative point impact of fumble and interception returns on competitive two-point attempts, as well as blocked kick returns (likely close to nil on such short kicks), and I don't know what the real formula shows but I suspect the coach knows what he's talking about regarding their analysis on the new rule.

Yeah, the net point outcome differences are statistically insignificant.

Considering the PAT kick had only a slight edge over the 2-pointer under the old rules, and considering how rarely the 2-pointer was attempted outside of late game strategy, there is a strong "bird in the hand" tendency among NFL coaches. I doubt that will change. Maybe if the wind is blowing hard, the kicker has missed already that day or the field is icy, we might see some coaches favor the 2 pointer more often.

Studies in behavioral finance gives us some clues. If you give people a choice being taking $1,000 and walking away or taking a chance on a coin flip for $2,000, the majority will take the $1,000 even though the propositions are statistically equal, with equal payouts expected if the proposition is performed an infinite number of times. In an isolated instance, however, most people are wired to take the bird in the hand.

 

[Deleted member 6794]

It appears the RSI adjusts for the offensive advantage since the distribution of offensive players and defensive players is fairly equal.

I get that.

 

[HardRightEdge]

It should though as the expected points average going for two is higher compared to kicking the extra point.



Even if your numbers should be correct (I got mine from Pro Football Reference's play index) the expected point average is still higher going for two (48.1%, 0.962 EPA) than kicking the extra point (0.941 EPA).

There's no reason to worry about extra point kicks or two-point conersations being returned for two points as it happened only five times on a total of 3,632 such regular plays since 1998.

Even if your numbers are right, the difference is 0.02. That would amount to exactly a 1 point difference in a 50 TD season. I don't see where anybody would be particularly swayed.

 

[Deleted member 6794]

Even if your numbers are right, the difference is 0.02. That would amount to exactly a 1 point difference in a 50 TD season. I don't see where anybody would be particularly swayed.

If my numbers are correct the difference is 0.059 in favor of going for two. The Packers scored a total of 60 TDs, so that would result in an additional three or four points. While that doesn't sound like a whole lot it could result in an additional win.

The difference is bigger when taking a look at teams running the ball on a two-point conversion with a success rate of 57.1%. That would result in additional 12 points for a team like the 2014 Packers. But for whatever reason teams have only run the ball on 28.3% of the two-point conversions in the last five years.

 

[HardRightEdge]

If my numbers are correct the difference is 0.059 in favor of going for two. The Packers scored a total of 60 TDs, so that would result in an additional three or four points. While that doesn't sound like a whole lot it could result in an additional win.

The difference is bigger when taking a look at teams running the ball on a two-point conversion with a success rate of 57.1%. That would result in additional 12 points for a team like the 2014 Packers. But for whatever reason teams have only run the ball on 28.3% of the two-point conversions in the last five years.

Right...I was quoting you quoting vince. If vince's numbers are correct, the difference is 0.02, or a one point differential over 50 TDs.

Regardless, whether it's 1 point or 3 points over 50 TDs, that differential can't be captured unless you go for 2 every time, regardless of the game situation. As for your 6 points when running the ball, if a team went for 2 every time and ran the ball on each occasion, we could expect the small advantage to disappear in a regression to the mean, so I discount it. Whether running or passing, if a team decides to go for 2 every time, opponents will be accumulating tape and will focus more closely on it in game planning.

When you start applying go-for-two to every game situation, there are potential unintended consequences. Down 7...score a TD...kick to tie: momentum is still on your side. But if you fail in going for 2 you give the opponent an emotional lift.

Which brings to mind the established habits of NFL coaches in late game opportunities. When having the choice of kicking an EP and going to overtime vs. going all-or-nothing on a 2 pointer, they nearly always kick. I doubt a tiny improvement in odds differential will change that thinking. Now project that thinking to other game situations.

I doubt this very minor change in the calculus is going to result in noticeable behavioral changes. It took coaches decades, going back to the days before the current golden age of place kicking, to figure out that attempting a FG on 4th. and 1 on the opponents 35 (or worse...punting) is an unfavorable proposition.

I see this very minor change in the calculus resulting in very minor changes in coaching behavior.

There is one prediction I'll make with a high level of confidence. If Crosby has an EP blocked on an opening drive TD and the Packers lose the game by one point, there will be a contingent arguing the Packers (or any other team for that matter) should have gone for 2 while ignoring all of the the other woulda, shoulda, coulda in the intervening 55 minutes of football.

 

[vince]

http://espn.go.com/blog/nfcnorth/post/_/id/71293/dont-expect-mike-mccarthy-to-go-for-two

• Since McCarthy took over in 2006, the Packers have converted just 38.9 percent of their two-point tries in the regular season. That is well below the NFL average of 46.9 percent during that time, and it ranks 23rd in the league, according to ESPN Stats & Information.
• Last season, the Packers were just 1-of-4 on two-point tries (including playoffs).
• McCarthy has attempted two-point conversions after just 4 percent of the Packers' touchdowns, a rate that ranks tied for 17th in the league since the start of the 2006 season.

However, McCarthy's feelings apparently would change in the case of a defensive penalty, which would move the line of scrimmage to the 1-yard line.

"Going to the 1 adds a whole other element," McCarthy said. "That takes the numbers to a different level. I think with that, that opportunity going to the 1, you'll definitely see more of those opportunities go up."

 

[Deleted member 6794]

Regardless, whether it's 1 point or 3 points over 50 TDs, that differential can't be captured unless you go for 2 every time, regardless of the game situation. As for your 6 points when running the ball, if a team went for 2 every time and ran the ball on each occasion, we could expect the small advantage to disappear in a regression to the mean, so I discount it. Whether running or passing, if a team decides to go for 2 every time, opponents will be accumulating tape and will focus more closely on it in game planning.

When you start applying go-for-two to every game situation, there are potential unintended consequences. Down 7...score a TD...kick to tie: momentum is still on your side. But if you fail in going for 2 you give the opponent an emotional lift.

Which brings to mind the established habits of NFL coaches in late game opportunities. When having the choice of kicking an EP and going to overtime vs. going all-or-nothing on a 2 pointer, they nearly always kick. I doubt a tiny improvement in odds differential will change that thinking. Now project that thinking to other game situations.

I doubt this very minor change in the calculus is going to result in noticeable behavioral changes. It took coaches decades, going back to the days before the current golden age of place kicking, to figure out that attempting a FG on 4th. and 1 on the opponents 35 (or worse...punting) is an unfavorable proposition.

I see this very minor change in the calculus resulting in very minor changes in coaching behavior.

I agree that game situation should factor into the decision to either kick the extra pick or go for two. That's why teams won't be able to settle for one point every time either.

McCarthy has already mentioned that he won't change his approach but I expect more aggressive coaches to go for two more often. Once thing McCarthy has to do in my opinion though is to get more diverse once trying a two-point conversion as passing it all the time is way too predictable.

 

[Deleted member 6794]

Since McCarthy took over in 2006, the Packers have converted just 38.9 percent of their two-point tries in the regular season. That is well below the NFL average of 46.9 percent during that time, and it ranks 23rd in the league, according to ESPN Stats & Information.

Last season, the Packers were just 1-of-4 on two-point tries (including playoffs).

McCarthy has attempted two-point conversions after just 4 percent of the Packers' touchdowns, a rate that ranks tied for 17th in the league since the start of the 2006 season.

I don't know where Demovsky's numbers are from but the Packers only tried a total of three two-point conversions last season including playoffs. As long as not proven otherwise I assume Pro Football Reference's numbers to be correct compared to the ones you have brought up so far.

 

[HardRightEdge]

But for whatever reason teams have only run the ball on 28.3% of the two-point conversions in the last five years.

There are a couple of reasons this might be the case.

The favored model in the pass-first game leans toward pass blocking OL players over road graders and (2) TEs valued for pass catching ability over run blocking. Many teams favor zone blocking schemes to accommodate the more athletic/finesse OL players valued for pass blocking over road grading. Many teams look for a primary RB in the Chris Johnson/Jamal Charles mold...they complement the passing game and provide some long gainers (which pad the average by the way), but are a crap shoot in short yardage. Many teams want what was once considered "3rd. down" skills on every down.

The compromise in the pass-first game is in personnel not suited to ground-and-pound. It's gotten to the point where 2nd. and 5 and 3rd. and 2 are predominantly passing downs.

Another consideration is in the hedge-betting in throwing the ball on short yardage. If you run with 2 yards to go, it's an all-or-nothing proposition on opening the hole or cutting a D-Lineman for a dive play. If you play action or roll out, the thinking goes, there's time for the QB to select from a few options among the receivers or run it in.

Certainly there are teams that balance pass/run in their personnel and scheme, either for lack of a decent QB or settling on a game managing option. They are in the minority, however.

If running is more successful than passing in short yardage, yet only 25% of the teams favor it, it's worth considering that 25% or fewer teams run a relatively balanced attack.

 

[HardRightEdge]

I agree that game situation should factor into the decision to either kick the extra pick or go for two. That's why teams won't be able to settle for one point every time either.

McCarthy has already mentioned that he won't change his approach but I expect more aggressive coaches to go for two more often. Once thing McCarthy has to do in my opinion though is to get more diverse once trying a two-point conversion as passing it all the time is way too predictable.

"Every time"...they don't do that now. "More often"...probably. But not with any frequency that will draw any particular notice.

 

[vince]

I don't know where Demovsky's numbers are from but the Packers only tried a total of three two-point conversions last season including playoffs. As long as not proven otherwise I assume Pro Football Reference's numbers to be correct compared to the ones you have brought up so far.

A fumbled snap in the second quarter of the Eagles game resulted in a failed two-point attempt rather than a missed extra point as PFR inaccurately classified it.