Reiss: Brady's target & completion %

[KoolAid]

Interesting stats. In all fairness to Moss he is consistently doubled, runs high risk/high reward routes, and would get quite a few OH Sh!t! chuck the ball down field throws coming his way. Imagine if Moss had TO's hands, he would be like 20%

I think these stats bode well for the well rounded, spread the ball everywhere offense we should see from here on out

 

[convertedpatsfan]

I am surprised that Moss has about the same percentage last year as he did in 2007. It looked like last year that many of the passes he would have connected with Brady in 2007 were not made last year (some being Brady's fault and other being Moss).

Peter King mentioned how much Larry Fitzgerald was missing Kurt Warner in today's column, but it showed that Fitzgerald was catching 62% of his targets from Warner, similar to Moss's figures. It also showed Fitzgerald down to 48% without Warner this season. So Moss's completion ratios are reasonable minus this season, where there's such a small sample size.

Team overall miss rate 33/120 .....27.5%
Moss miss rate ......13/22 59%
Moss was targetted 18.3% of the time (22/120) but had 39.4% of the misses.
Am I reading the number right?
Disclaimer- obviously bombs and the deep ball will have a lower success rate than a checkdown but does this account for the entire difference?????

There are lies, damn lies, and then statistics. You can twist those numbers to say whatever you want sometimes, but it's important to note we're talking about an extremely small sample size. Moss has had some drops this season, but there have also been some bad throws. If only 4 of those passes are a bit better and he catches them, suddenly he's above his average completion %. A bad pass on a 5 yard slant can still be caught. A bad pass on a 40 yard bomb is hard to catch up to. Regardless though, we're talking about a small sample size.

As for the numbers you mentioned, part of it is on Moss, but part of it is also on Brady, who has never been a great deep ball thrower. And I say that as a huge Tom Brady fan who cheered for him at Michigan, and only became a Pats fan when he got drafted.

Another thing to note here is the number of times he was targeted compared to previous years. His stats project to 88 throws his way, compared to 138 last year. Even if you scrap the Miami game and go by the first 3 (21 targets), he still projected to be targeted only 112 times. That's a 20% reduction.

Does that make Wes Welker suck as well? His targets are down 27% from last season (8.5 in 2010 vs. 11.5 in 2009). And this drop-off isn't related to Edelman, who has also seen a drop in targets.

We are shifting to a more balanced offense, and adding new targets in Hernandez and Gronk, along with a healthy Tate. Of course the target numbers to Moss and Welker should go down. If they didn't, something would be seriously wrong.

I know Moss wasn't playing great, but it was early, it was a small sample size, and people are drawing all sorts of unfair conclusions from it.

 

[emoney_33]

I believe Moss played the Jets twice last year, as well as the Dolphins (0 for 1 this year). And he still came out over 60%. So you can be as selective as you want, but even with Cassell as his QB, the success rate was still 55%. There's a pretty big difference between that and 41%.

He had 10 targets in the Jets game this year which accounts for 45.5% of his total targets. I'm not being "selective", I'm simply pointing out facts including the paltry sample size of 22 targets.

 

[TheGodInAGreyHoodie]

I think Hernandez % will go down as he gets more balls thrown his way, the same will probably be true of Edelman.

Probably. But the Hernadez sample size is not small, he caught the ball 18 out of 20 times. That not a fluke, like the Woodhead or Crumpler sample size.

 

[Kontradiction]

You might as well have just started a thread and called it "Brady's target and completion % to Randy Moss". After all, isn't that all you wanted to get across?

 

[TheGodInAGreyHoodie]

You might as well have just started a thread and called it "Brady's target and completion % to Randy Moss". After all, isn't that all you wanted to get across?

That's half the point, but not the only point. The more important point is that the guys who will be playing in the Raven's game do a pretty darn good job at catching the ball.

 

[emoney_33]

That's half the point, but not the only point. The more important point is that the guys who will be playing in the Raven's game do a pretty darn good job at catching the ball.

I've certainly been impressed with the hands of Gronk and especially Hernandez. A reliable TE or 3, ahh life is good

 

[eagle eye]

We'll see how they do now that Randy is gone. I'd expect more coverage in the middle now which makes it likely that their completion % will go down along with their yac.

Its all about Tate now or Edelman or whoever else can run flat out. If that person starts hauling in the deep balls then the numbers will remain consistent as opposing defenses will have to view them as a similar threat to what Moss was but until that happens there will be more players up around the box.

 

[patsfaninpittsburgh]

My guess is the staff views the importance of the Moss double teams much less important than the average fan or mediot. If the double coverage was the single most important factor in the offense; Moss would have been signed to an extension.

Besides, Tate with single coverage is more likely to produce big gains vs Moss double covered. Besides, "stretching the field" may now mean 15-20 yard completions.

The NFL constantly changes. Anybody notice how the "big play" defense is now the rage?

 

[PatsDeb]

We'll see how they do now that Randy is gone. I'd expect more coverage in the middle now which makes it likely that their completion % will go down along with their yac.

Its all about Tate now or Edelman or whoever else can run flat out. If that person starts hauling in the deep balls then the numbers will remain consistent as opposing defenses will have to view them as a similar threat to what Moss was but until that happens there will be more players up around the box.

This is why I am wondering whether the Pats will come out 5 wide against the Ravens and throw deep to Tate on the first play, to show everybody they haven't lost anything and everyone still has to be worried about the deep threat.

 

[lamafist]

It's important to remember that, whatever the clever quip says, statistics don't really lie themselves -- it's people that add the lies into their motivated interpretation of the statistics. Much of the time, this comes in the form of people skipping ahead several steps and reaching conclusions of their choosing among a wider array of possible interpretations of the stats.

But that doesn't mean you should ignore the statistics altogether.

Randy Moss' catch rate through the first four games is 42%. While this is a sample size of only 1/4 of the season, a decline that steep from his average with Brady (61%) and his career average outside of Oakland (59%) should not be written off to standard variation without justifying that conclusion.

In other words, writing off Moss' decline in catch rate to sample size without first verifying that he has had similar 4-game stretches of under-average performance in the past is just as much of an unfounded conclusion as people who are assuming that his catch rate will stay around 40% this season. Admittedly, I've only given his game logs a cursory glance, but no similar four game slump in his career w/ Brady has jumped out at me.

A further reason to suspect a factor beyond statistical variance is the number of dropped passes that have been credited to Moss so far -- an NFL-leading 5. Now, admittedly, dropped passes is a partially subjective stat, but, in the interest of fairness, the STATS Inc. game charters make a point of giving receivers the benefit of the doubt, and only awarding drops on the most obviously catchable passes. Moss 5 obvious drops through four games is, I believe, more than he had through 16 games in 2007 and 2009... and we haven't even played a game in bad weather yet.

Added to the drop in general catch rate, even though the sample size is only four games, Moss would already need to have an unprecedented -- even for him -- final 12 games of the season in order to finish the year tied with his averages with Brady. This, again, argues against Moss' statistical downturn being attributable to variance.

Also, I've seen arguments that a few bad passes from Brady could enough to account for Moss' sub-par showing. First, let's remember that since we're mostly comparing these four games in 2010 to Moss' 32 prior games with Tom Brady, the quarterbacks relative talent for throwing the deep ball is a constant. Let's also note that the quarterback in question is completing passes above his already-impressive completion percentage in his years with Moss, and also, most importantly, is not targeting Moss for a higher percentage of "bomb" passes than in years past, but simply completing fewer of them. Let's also remember that just as a receiver's ability to catch the ball is affected by the quarterbacks passes, the quarterback's passes are equally influenced by the receiver's ability to get separation. If a receiver has good separation at the time of the pass, a QB can hit him on the shoulder. If the receiver has only half a step, the QB needs to overthrow him, and hope he can catch up to the pass.

So basically, what all of this adds up to is that Moss' sub-par play through four games, while not enough to dismiss him altogether, is certainly cause to be cautiously worried if you're relying on him for the rest of the season, and something that would clearly factor into BB's decision to trade him.

 

[PatsDeb]

These stats seem kind of bogus. I was upset with the Moss trade because it makes our team weaker, at least theoretically, in the short run. I thought Moss played pretty hard for us on a consistent basis on the field, and that his problems were more off the field. Looking at the first 4 games of the season, which included one where the game plan was to use him almost exclusively as a decoy, doesn't seem fair. What if he gets 2 TDs and 130 yards tonight? Should those be added into this formula? Seems like it's time to just move on, rather than keep investigating why this trade happened. We'll never know for sure unless Brady or BB writes a tell-all after they retire from the Pats (not bloody likely).

 

[emoney_33]

In other words, writing off Moss' decline in catch rate to sample size without first verifying that he has had similar 4-game stretches of under-average performance in the past is just as much of an unfounded conclusion as people who are assuming that his catch rate will stay around 40% this season. Admittedly, I've only given his game logs a cursory glance, but no similar four game slump in his career w/ Brady has jumped out at me.

One only needs to find one bad game, maybe the Panthers game last year? Jets games last year? The Jets game in 2010 he was 2/10. 10 is 45.5% of his total targets and that gives a huge skew. For example if he were something like 10/10 he'd be at 77% and the argument that he was the best of his Patriots career in 2010 would be just as poor a conclusion.

He's 7/12 in the 3 other games (58%). If it were consistently 41% each game then it'd be a bit different.

Added to the drop in general catch rate, even though the sample size is only four games, Moss would already need to have an unprecedented -- even for him -- final 12 games of the season in order to finish the year tied with his averages with Brady. This, again, argues against Moss' statistical downturn being attributable to variance.

I agree that the uncharacteristic drops were a bit worrisome, but still the sample size is just too small to say that it's clearly a sign of deterioration. How many drops were in the Jets game?

I don't think anyone expected Moss' stats not to go down this year, as will Welker's.

 

[strngplyr]

Crumpler had a 22 yard sideline catch against the Bengals, so he has atlest one.


..hmm was that call reversed? Or did they give the credit to someone else? I can't find the stats of it.


edit: jk i think i was remember the 24yd catch on the sideline in the rams preseason game

 

[lamafist]

One only needs to find one bad game, maybe the Panthers game last year? Jets games last year? The Jets game in 2010 he was 2/10. 10 is 45.5% of his total targets and that gives a huge skew. For example if he were something like 10/10 he'd be at 77% and the argument that he was the best of his Patriots career in 2010 would be just as poor a conclusion.

He's 7/12 in the 3 other games (58%). If it were consistently 41% each game then it'd be a bit different.

But he didn't catch 10/10 passes against the Jets, he caught 2/10. And his 58% in the other 3 games is still only around his average. In fact, out of the 4 games he's played, in only one did he equal his usual average catch rate. The problem is that, unlike in previous years, he's gone an unusual number of games without having any above-average performances, which should be expected about as often as a sub-par performance.

And looking around the Carolina game, it looks like if you jigger it so that it misses a 5/6 performance against Buffalo, it does look like you might get somewhere around 43% over four games -- so how much better is it to say that Randy's first four games are only <em>just as bad as</em> his worst four-game stretch with us, and no worse? It's better, but not significantly better.

The problem is that the four-game stretch to start the season is a random selection, whereas picking the three worst games around a bad Carolina game isn't random, but selected for poor performance. So to have the first four games turn out equivalent to the worst selection isn't encouraging, because it's saying, essentially, that he's already used up his slump for the year, and will now need to play significantly ABOVE average for the rest of the season to equal is average. So he'd now need to have his best stretch of 12 games he's ever had with the Patriots, just to end up with his average season for us. And that's very unlikely.

So the fact that a four game stretch is already enough to make him need to play better than he has for us in any 12 game stretch implies that it is enough to be taken seriously.

I agree that the uncharacteristic drops were a bit worrisome, but still the sample size is just too small to say that it's clearly a sign of deterioration. How many drops were in the Jets game?

I don't think anyone expected Moss' stats not to go down this year, as will Welker's.

How can the sample size be too small? Even if Moss didn't drop another pass all year, he'd still end up with more than in previous years. And the chances of him not dropping any more passes are marginal. The chance of random distribution putting five drops over 16 games all in the first four games is not quite vanishingly small, but getting there.

 

[emoney_33]

But he didn't catch 10/10 passes against the Jets, he caught 2/10. And his 58% in the other 3 games is still only around his average. In fact, out of the 4 games he's played, in only one did he equal his usual average catch rate. The problem is that, unlike in previous years, he's gone an unusual number of games without having any above-average performances, which should be expected about as often as a sub-par performance.

And looking around the Carolina game, it looks like if you jigger it so that it misses a 5/6 performance against Buffalo, it does look like you might get somewhere around 43% over four games -- so how much better is it to say that Randy's first four games are only <em>just as bad as</em> his worst four-game stretch with us, and no worse? It's better, but not significantly better.

The problem is that the four-game stretch to start the season is a random selection, whereas picking the three worst games around a bad Carolina game isn't random, but selected for poor performance. So to have the first four games turn out equivalent to the worst selection isn't encouraging, because it's saying, essentially, that he's already used up his slump for the year, and will now need to play significantly ABOVE average for the rest of the season to equal is average. So he'd now need to have his best stretch of 12 games he's ever had with the Patriots, just to end up with his average season for us. And that's very unlikely.

So the fact that a four game stretch is already enough to make him need to play better than he has for us in any 12 game stretch implies that it is enough to be taken seriously.



How can the sample size be too small? Even if Moss didn't drop another pass all year, he'd still end up with more than in previous years. And the chances of him not dropping any more passes are marginal. The chance of random distribution putting five drops over 16 games all in the first four games is not quite vanishingly small, but getting there.

I'm not sure exactly what your argument is? Fact of the matter is the completions / targets number is skewed by 1 game. And a 2nd game is pretty negligible because he was targetted only once on a fake spike.

You seemingly have gone off on a tangent to argue that Moss was going to have a statistically "down" year compared to past years with the Patriots and I don't think you'll find anyone try to argue that. Most expected that to be the case even before game 1.

 

[bucky]

Does that make Wes Welker suck as well? His targets are down 27% from last season (8.5 in 2010 vs. 11.5 in 2009). And this drop-off isn't related to Edelman, who has also seen a drop in targets.

We are shifting to a more balanced offense, and adding new targets in Hernandez and Gronk, along with a healthy Tate. Of course the target numbers to Moss and Welker should go down. If they didn't, something would be seriously wrong.

I know Moss wasn't playing great, but it was early, it was a small sample size, and people are drawing all sorts of unfair conclusions from it.

First, I never said that Moss sucked. I never mentioned one subjective thing about him. Just pointing out the numbers as compared to the rest of his career with the Pats.

Second, Welker's completion/target % is the same in 2010 as it was in 2009 and 2007. His total targets are down - which is partially related to his injury and partially to the team's balance.

And third, the sample rate here is 4 games, which is 25% of the season. And while that's not a full season, it's not 1 game either. I think 25% is certainly a significant enough sample rate to be able to draw conclusions.

 

[PatsWickedPissah]

Does that make Wes Welker suck as well? His targets are down 27% from last season (8.5 in 2010 vs. 11.5 in 2009). And this drop-off isn't related to Edelman, who has also seen a drop in targets.

We are shifting to a more balanced offense, and adding new targets in Hernandez and Gronk, along with a healthy Tate. Of course the target numbers to Moss and Welker should go down. If they didn't, something would be seriously wrong.

I know Moss wasn't playing great, but it was early, it was a small sample size, and people are drawing all sorts of unfair conclusions from it.

You posted and answered your question well.

Before dismissing the small Moss sample, compare it to other 4 game samples. Choose one with a Jets game in it to really be comparable.

 

[emoney_33]

First, I never said that Moss sucked. I never mentioned one subjective thing about him. Just pointing out the numbers as compared to the rest of his career with the Pats.

Second, Welker's completion/target % is the same in 2010 as it was in 2009 and 2007. His total targets are down - which is partially related to his injury and partially to the team's balance.

And third, the sample rate here is 4 games, which is 25% of the season. And while that's not a full season, it's not 1 game either. I think 25% is certainly a significant enough sample rate to be able to draw conclusions.

Would you like to go on record as saying that Moss will likely be close to 41% this season? Yes or no.

It's really simple here, the sample size is tiny with a 1 game skew. The stats prove NOTHING at this point in time.

 

[convertedpatsfan]

It's important to remember that, whatever the clever quip says, statistics don't really lie themselves -- it's people that add the lies into their motivated interpretation of the statistics. Much of the time, this comes in the form of people skipping ahead several steps and reaching conclusions of their choosing among a wider array of possible interpretations of the stats.

But that doesn't mean you should ignore the statistics altogether.

Fair enough. I'm certainly not ignoring the numbers, just a bit alarmed by what people are concluding from them.

Randy Moss' catch rate through the first four games is 42%. While this is a sample size of only 1/4 of the season, a decline that steep from his average with Brady (61%) and his career average outside of Oakland (59%) should not be written off to standard variation without justifying that conclusion.

In other words, writing off Moss' decline in catch rate to sample size without first verifying that he has had similar 4-game stretches of under-average performance in the past is just as much of an unfounded conclusion as people who are assuming that his catch rate will stay around 40% this season. Admittedly, I've only given his game logs a cursory glance, but no similar four game slump in his career w/ Brady has jumped out at me.

Regarding similar sample sizes, I'd first argue it's really a 3-game stretch since he caught 5 of 8 passes in week 1, good for 63% of targets. The number of targets and catches also project to around his career averages (based on the small sample size of 1 game!). But if we look at 4-game stretches, there are a few.

2009 was a pretty solid year for Moss, but between NFL weeks 11 to 14 (starting with the second Jets game), Moss caught only 11 passes out of 26 targets, 42%. He did separate a shoulder at some point, but not sure when or how long it lasted.

2008 might be unfair since Cassel was at the helm, but weeks 11 to 14 again saw Moss catch 11 passes out of 25 targets, 44%.

Even the record-setting 2007 featured a similar period in terms of catches per targets near the end of the year, with a 6-game stretch resulting in 32 passes out of 67 targets, 48%.

None of this confirms or denies that Moss may be done or Moss may have a lot left. But I feel 4 games are not enough to conclude either way.

A further reason to suspect a factor beyond statistical variance is the number of dropped passes that have been credited to Moss so far -- an NFL-leading 5. Now, admittedly, dropped passes is a partially subjective stat, but, in the interest of fairness, the STATS Inc. game charters make a point of giving receivers the benefit of the doubt, and only awarding drops on the most obviously catchable passes. Moss 5 obvious drops through four games is, I believe, more than he had through 16 games in 2007 and 2009... and we haven't even played a game in bad weather yet.

This was the most disturbing thing about Moss this season, and it really annoyed me. Drop numbers seem more difficult to find for past seasons, though I read Moss had 6 drops in 2007 and last year had 7 (though I also read 9 for 2009). Based on so few targets, 5 drops is alarming, but then again, Welker is tied for 4th in drops with 3.

Added to the drop in general catch rate, even though the sample size is only four games, Moss would already need to have an unprecedented -- even for him -- final 12 games of the season in order to finish the year tied with his averages with Brady. This, again, argues against Moss' statistical downturn being attributable to variance.

I don't think he would ever catch up to his averages with Brady simply because the offense has changed. The integration of the tight ends as well as more emphasis on the running game and the emergence of Tate meant fewer targets for Moss and Welker, and the numbers reflect that even for Welker, though he's caught a similar percentage of passes.

Also, I've seen arguments that a few bad passes from Brady could enough to account for Moss' sub-par showing. First, let's remember that since we're mostly comparing these four games in 2010 to Moss' 32 prior games with Tom Brady, the quarterbacks relative talent for throwing the deep ball is a constant. Let's also note that the quarterback in question is completing passes above his already-impressive completion percentage in his years with Moss, and also, most importantly, is not targeting Moss for a higher percentage of "bomb" passes than in years past, but simply completing fewer of them. Let's also remember that just as a receiver's ability to catch the ball is affected by the quarterbacks passes, the quarterback's passes are equally influenced by the receiver's ability to get separation. If a receiver has good separation at the time of the pass, a QB can hit him on the shoulder. If the receiver has only half a step, the QB needs to overthrow him, and hope he can catch up to the pass.

I would keep the sample size in mind. Week 2 and week 4 were awful performances, no argument there. Week 1 wasn't that bad, and in week 3, he only received 3 targets, but he caught 2, both for touchdowns.

While he hasn't had a huge game this season, it's also worth noting he never caught more than 5 balls or gained more than 75 yards in the last 7 games of the 2009 season.

So basically, what all of this adds up to is that Moss' sub-par play through four games, while not enough to dismiss him altogether, is certainly cause to be cautiously worried if you're relying on him for the rest of the season, and something that would clearly factor into BB's decision to trade him.

I agree with this conclusion 100%.