RB Alex Collins, SEA (2 Viewers)

[Buckna]

Oh, we're counting recovered fumbles now as well?  I bet if you took the time to count up the top 15 RB's from last year and their recovered fumbles, it would be very eye-opening and you could spin that narrative that "they were lucky that a teammate recovered it" for just about any one of them.

I don't have to take the time to count them up, fumbles are an official record keeping stat. Jalen Richards had the most for a RB at 8 total fumbles. Gurley was second with 5, the only RB's with 4 on the season are the 4 I just named in the previous post, including Alex Collins. Quite a few guys at 3 total fumbles.

And if you don't think coaches care about total fumbles as opposed to just the ones that bounced the other way when it comes to recovering them, I don't know what to tell you.

 

[Brisco54]

Think your underselling this quite a bit, he fumbled twice at the end of the season, both in close games. He just got lucky that a teammate recovered it both times. 4 total fumbles on the season isn't a huge number, but does put him tied for 3rd most for a RB last year with Freeman, Duke Johnson, and Ajayi.

4 fumbles does put you in the bottom quarter of the league (going back to 2014, the annual number of RBs with at least 4 fumbles:  8, 13, 8, 6.

As I said, its a concern, but not as significant to me as it apparently is to you.

 

[Chaka]

 I am not trying to sell a narrative, but instead show the flaws in narrative Chaka was implying. 

The theory appears to be that he is JAG who's weaknesses were exposed as the season progressed.

Nope, that's not the theory at all.

 

[Buckna]

4 fumbles does put you in the bottom quarter of the league (going back to 2014, the annual number of RBs with at least 4 fumbles:  8, 13, 8, 6.

As I said, its a concern, but not as significant to me as it apparently is to you.

As I said, 4 is not a huge number, it could be easily explained by variance or just bad luck.

However with a guy with his history, it is still pretty concerning, especially late in the season after the Ravens supposedly made it a point to fix his fumbling after the first two early in the season. Collins had 2 fumbles on 42 touches for the Seahawks in 2016, and another for the Seahawks in week two of the preseason in 2017 prior to him being cut and the Ravens picking him up. His college problems have already been mentioned. It remains to be seen what happens this year, maybe another year of focus and refinement and it will no longer be an issue.

 

[Brisco54]

taking out an outlier in statistics means taking one at both ends. not cherry picking which one you take out

Forgive me for using what appears to be a term of art.  I did not drop the CLE stats because it was a 'statistical' outlier, but a factual one.  Similar to how I would have excluded a game where the rains and mud were so heavy that no one gained more than 25 yards for either team.  

I gave a reason as to why I thought the CLE game was factually dissimilar to the others and therefore excludable.  You may disagree.

What I was not trying to do was any form of formal statistical analysis.  Again, forgive me for what apparently is a term of art.

More importantly, note that I was not proposing a narrative, but rather showing the flaws of one proposed by mirroring the fallacious errors within it.

 

[Chaka]

So, briefly, regardless of when games were played when Collins carried the ball 15 or more times (eight games) he averaged 4.4 YPC on 140 carries, which is very respectable.  Over those eight games he averaged 4.93, 5, 3.75, 6.28, 6.67, 2.83, 2.45 & 3.9 YPC which seems like a pretty even split and probably not too far in variance off of more celebrated RBs.

 

[Brisco54]

Nope, that's not the theory at all.

Please forgive my misunderstanding.  I was trying to figure out the importance of breaking the season into quartiles.

 

[Brisco54]

So to sum up, we appear to be in the following camps:

1.  Collins is an RB2 borderline RB1 talent that will perform as such once the line is healthy.

2.  Collins is a solid talent that will lose significant carries to Dixon, making both unattractive.

3.  Collins is a meh talent at best that the league has figured out and will fumble his way out of a job.

4.  Collins may be ok or even good, but Dixon is much better and the cream will rise to the top.

5.  Collins is a solid but not spectacular talent that will probably hold on to the RB1 job, but so his floor is solid, but his ceiling is pretty low.

Is that a fair representation of where we all stand?

 

[cloppbeast]

So, briefly, regardless of when games were played when Collins carried the ball 15 or more times (eight games) he averaged 4.4 YPC on 140 carries, which is very respectable.  Over those eight games he averaged 4.93, 5, 3.75, 6.28, 6.67, 2.83, 2.45 & 3.9 YPC which seems like a pretty even split and probably not too far in variance off of more celebrated RBs.

Yeah, but.... if you don't count the Pittsburgh game it's like 3.8.

 

[Interseptopus]

Forgive me for using what appears to be a term of art.  I did not drop the CLE stats because it was a 'statistical' outlier, but a factual one.  Similar to how I would have excluded a game where the rains and mud were so heavy that no one gained more than 25 yards for either team.  

I gave a reason as to why I thought the CLE game was factually dissimilar to the others and therefore excludable.  You may disagree.

What I was not trying to do was any form of formal statistical analysis.  Again, forgive me for what apparently is a term of art.

More importantly, note that I was not proposing a narrative, but rather showing the flaws of one proposed by mirroring the fallacious errors within it.

an outlier is an outlier. It doesn't matter if it's statistically significant or not. If you're going to run an average by taking out an outlier, you take out one above and below to get an accurate picture. Otherwise you are skewing your data. Stats 101

 

[Brisco54]

an outlier is an outlier. It doesn't matter if it's statistically significant or not. If you're going to run an average by taking out an outlier, you take out one above and below to get an accurate picture. Otherwise you are skewing your data. Stats 101

This rule may be true for stats (where I believe each individual stat has equal credibility and weight to all others), but it has questionable applicability to ordinary reality.  

Consider:

Player A did not play the first game of the year because he was on suspension.  He then got ten fantasy points every week except one, where he had 30 fantasy points.

I am factually excluding the first week in determining his points per game to evaluate fantasy potential because he was not eligible to play.  Because I did that, you are saying that I am "skewing the results" unless I also exclude the 30 point week?  I disagree.  In real world decisions, facts can mandate that a given piece of data lacks the credibility to be included in the data set.

I believe your top and bottom outlier rule applies when all data is treated as having equal credibility and weight... you do not have the ability to factually analyze the given piece of data, so you operate with a presumption that the top and bottom outliers are the factual anomalies because they are the statistical anomalies.  

Again, I argued the CLE game should be eliminated because of factual circumstance not because it was a low score (or high score).  If, instead, CLE had gone on strike that week so Collins ran against replacement players and got 500 yards on 10 carries, I would also eliminate that from his YPC that I would use to analyze his prospects.  I would also eliminate it if he had gained exactly the same number of yards as his career average.  The facts force the exclusion, not the that it is a statistical anomaly.

Alternatively, if you can show me another game where the CLE factual circumstances also applied, I would eliminate that game from my YPC determination regardless if the result was a low, high or average score.

 

[Brisco54]

Yeah, but.... if you don't count the Pittsburgh game it's like 3.8.

If the 4.4 for 140 carries was accurate, then if you take out the Pitt game (18 carries for 120 yards), it is 4.06 YPC, not 3.8.

Is there something factually unique about the conditions of the Steelers game that justifies it being excluded from consideration?

 

[JetMaxx]

Has anyone factored in the most injured OL and it’s poor play last year?

 

[Interseptopus]

has anyone factored in games where Collins ate scrambled eggs in the morning vs pancakes? 

 

[Chaka]

If, instead, CLE had gone on strike that week so Collins ran against replacement players and got 500 yards on 10 carries, I would also eliminate that from his YPC.

Oh, you mean the Miami game.

 

[Chaka]

This rule may be true for stats (where I believe each individual stat has equal credibility and weight to all others), but it has questionable applicability to ordinary reality.  

Consider:

Player A did not play the first game of the year because he was on suspension.  He then got ten fantasy points every week except one, where he had 30 fantasy points.

I am factually excluding the first week in determining his points per game to evaluate fantasy potential because he was not eligible to play.  Because I did that, you are saying that I am "skewing the results" unless I also exclude the 30 point week?  I disagree.  In real world decisions, facts can mandate that a given piece of data lacks the credibility to be included in the data set.

I believe your top and bottom outlier rule applies when all data is treated as having equal credibility and weight... you do not have the ability to factually analyze the given piece of data, so you operate with a presumption that the top and bottom outliers are the factual anomalies because they are the statistical anomalies.  

Again, I argued the CLE game should be eliminated because of factual circumstance not because it was a low score (or high score).  If, instead, CLE had gone on strike that week so Collins ran against replacement players and got 500 yards on 10 carries, I would also eliminate that from his YPC that I would use to analyze his prospects.  I would also eliminate it if he had gained exactly the same number of yards as his career average.  The facts force the exclusion, not the that it is a statistical anomaly.

Alternatively, if you can show me another game where the CLE factual circumstances also applied, I would eliminate that game from my YPC determination regardless if the result was a low, high or average score.

Apologies in advance for this but not only do I still think you don't quite understand how statistics work but I am starting think you also don't quite understand how facts work.

For example in your example you are not excluding the week the player did not play as it was already excluded by definition because he did not play. What sports stat (or statistician) includes DNPs in PPG numbers? None that I am aware of. With I believe FBGs includes "targets per team game" in their target data, so there's that.

Another example is that there is no factual evidence that warrants excluding the Cleveland game in the first place. What? They stacked the box? So you're saying they game planned against the Ravens strength and it worked. Why would you exclude that other than to satisfy a pre-established conclusion?

 

[Chaka]

Again, I apologise if I come across as mean @Brisco54 but I am really having difficulty understanding some of your arguments. 

 

[JetMaxx]

has anyone factored in games where Collins ate scrambled eggs in the morning vs pancakes? 

It’s ok to still like a guy you traded away. A post like this after trading someone comes across as a little insecure as a seller. 

 

[Arodin]

an outlier is an outlier. It doesn't matter if it's statistically significant or not. If you're going to run an average by taking out an outlier, you take out one above and below to get an accurate picture. Otherwise you are skewing your data. Stats 101

Actually no.  Stats 101 has a formal definition of outliers which defines boundaries above and below the mean.  Any data beyond these bounds can be considered an outlier, making it acceptable to exclude all such and observe any changes in stats that result.

There is no requirement that equal data be removed from both ends, nor should this be done arbitrarily.

A prior poster gave a football example, but a pure numerical one would be: 5,4,5,6,4,38,5,4,3.  It is entirely justifiable to consider the effect of excluding the 38.  It is not necessary, and in fact improper to also drop the 3.

Lastly, analyzing without outliers should not replace the original analysis.  The two should be compared, and both included in a thorough analysis.

 

[Interseptopus]

It’s ok to still like a guy you traded away. A post like this after trading someone comes across as a little insecure as a seller. 

Not at all. I got a 1st 2nd and 3rd for him. cant sell any higher than that. I wont regret it unless he finishes top 10 for the next 5 seasons. I'm built for the long haul. I wasnt trying to trade him, but I find it hard to believe the majority of his owners would pass up on that deal. 

I will admit to this day I'm 50/50 on him. in fact I think I said that yesterday in the dynasty trade thread. I said I'm 50/50 and leaning toward the sell side lately. I can definitely see both sides. how he finished the year on a ypc concerns me, but watching him play I see a guy who has "it" to be a lead back, and I'm surprised Seattle cut him over their other options. so do I believe the mystical eye test or the trend of stats. I think the last 50 pages or so is likely hashing out just that. as someone who reads evidence every week, I tend to side with the statistics more. 

My post was pointing out how ridiculous and over analyzed people have gotten with Collins. As if it's going to help predict this season. 

at this point you either like him or you dont. and lucky for everyone we will find out sooner rather than later who is right, as if it matters, but to some it does. and then in 5 months we will be able to debate a different player for 7 months on end. rinse and repeat. 

 

[Bracie Smathers]

Collins was a mediocre college back, I’d say he’s a backup by mid season. He has some horrible hands with meh vision. 

Ravens are finding new ways to use RB Alex Collins in the passing game
 

...Entering the 2018 season, the Ravens are looking to build off of what they saw from Collins last season. When asked by a reporter about the possibility of using him as a slot receiver, Baltimore offensive coordinator Marty Mornhinweg liked the idea.

“That’s a great point,” Mornhinweg said in a Sunday post-practice press conference “When we first got him, [running backs coach] Thomas [Hammock] really, really did a great job, and he was sort of morphing into a really good pass receiver – both from the backfield and split. Then he had a little setback there later in the year, and now he’s coming on again. So yeah, that’s a good point. If he can do that, that will really add to our offense.”

Allowing Collins to play as a slot receiver could certainly create problems for opposing defenses. Collins profiles as a dangerous mismatch for defenders — he’s too fast for linebackers to keep up with and he’s too big for defensive backs to get physical with. As a result, sticking Collins out in the slot creates the potential for him to get out in open space and expose defenses.

In addition, keeping Collins in motion pre-snap as a running back can often help identify what kind of coverage the opposing defense is running, whether it be man or zone coverage.

Regardless of the utility that lining him up as a receiver offers the offense, it’s a good idea to add another dimension to Collins’ game. Opposing defenses now have tape on Collins after his breakout season, allowing them an opportunity to identify his tendencies and running style. Turning Collins into a receiver will go a long way towards keeping defensive coordinators around the league on their feet this season.

 

[Cobbler1]

I’ve seen just brief flashes where it looks like Collins may have the ability to run some option type routes out of the backfield. I haven’t seen enough of it to really have a ton of confidence in that being more than just coach speak at this time but definetly something to look for in preseason. I’ll say that he catches the ball just fine on releases out of the backfield and checkdowns. He even made a couple nice diving catches several yards down the field one vs the Colts in the rain and the other I think at Cleveland. His hands as far as catching the ball are actually pretty good, a far cry from being horrible as the one poster described them. It would take a leap from what he showed last year in his route running ability to actually line up in the slot, regularly a LB and make the catch several yards down the field though IMO.

 

[Gottabesweet]

I've been trying to sell him all offseason.  I'm finally getting closer to a RB needy team.  I would have easily taken a 1,2 and 3 for him.   

 

[Brisco54]

For example in your example you are not excluding the week the player did not play as it was already excluded by definition because he did not play. What sports stat (or statistician) includes DNPs in PPG numbers? None that I am aware of. With I believe FBGs includes "targets per team game" in their target data, so there's that.

Ok, so change the DNP to a game where he played but only returned a kick.  I would not include that in my analysis of points per game played even if it technically met the definition.  This issue was responded to far better than I ever could by Arodin, who shows he probably knows more about statistics than the rest of us combined.

Another example is that there is no factual evidence that warrants excluding the Cleveland game in the first place. What? They stacked the box? So you're saying they game planned against the Ravens strength and it worked. Why would you exclude that other than to satisfy a pre-established conclusion?

Here is the argument that I expected folks to make... what facts was I relying on to exclude the CLE game?

I think we are really in a discussion of how probative a given fact is.  If you remember, when I was excluding the CLE game I was doing to so rebut an argument that Collins was exposed as JAG by defenses as the season progressed (an argument it turns out you were not making).

I felt that his individual performance did not support this conclusion... that the manner in which you broke the season statistics into quartiles was misleading.  I watched the game and to me the CLE game was less probative of his individual performance because:

1.   Collins had 13 yards on his first three carries, the Browns changed their defense to sell-out against the run for the rest of the first half daring Flacco to beat them.  Collins went 3 for 13 followed by a 9 rushes for 6 yards.  

3.  On his 4th snap of the second half for B-More, Collins got knocked out and had to be examined on the side line.  He got only one more snap the rest of the game on the final drive.  

4.  After it appeared Collins was knocked out, the Browns went back to standard D.  They feared Collins more than Flacco, but they did not fear Allen.

4.  Allen went 10-59 after that (9 for 60 in the 4th quarter). 

To me, this game is less probative than the others.   To me, game script and injury were more responsible for his low YPC that day than his personal talent/ability.  It was more an indictment of Flacco than Collins and thus less probative for the purposes of evaluating whether there was a downward trend as the season progressed that evinced he was JAG and so therefore I excluded it.  

Finally you said, "I am starting think you also don't quite understand how facts work."

Fair point. : )  In fact, many an opposing counsel would probably agree with you.  

 

[Brisco54]

On the subject of buying or selling Collins for Dynasty, I think it comes down to team status.  If I am not in contention I am selling.  He is far more valuable to a team that will use his points this year, and there are too many questions about B-Mores future to count on him for more than the next two years.  Further, his current market price is probably almost as high as it would be if the Dixon issue disappears... whereas his market will start to plummet if Dixon has a strong preseason.

If I am in contention, I would not move him unless I am offered an overpay or if he is surplus and the trade fills a gaping hole elsewhere in the lineup.  Dixon doing well in the preseason probably speaks more to Collins long term value.  I feel that barring injury or complete ineptitude, Collins has earned the benefit of the doubt from the B-more coaches for the first half of the season, even if he fumbles a couple of times.

As far as his current value, I stand by my earlier evaluation and say in the range of 2019 pick 1.4-1.6.

2019 picks are a year away and I always devalue picks about half a round per year in the future. 

If Collins were on the waiver wire for a dynasty rookie draft at this time, I think he would be selected no later than pick 1.8.  Personally I would not trade him straight up for Royce Freeman, Sony Michel, Calvin Ridley or Christian Kirk. (ranked as rookies 7-10 currently on this site.)

 

[kutta]

Ok, so change the DNP to a game where he played but only returned a kick.  I would not include that in my analysis of points per game played even if it technically met the definition.  This issue was responded to far better than I ever could by Arodin, who shows he probably knows more about statistics than the rest of us combined.

Here is the argument that I expected folks to make... what facts was I relying on to exclude the CLE game?

I think we are really in a discussion of how probative a given fact is.  If you remember, when I was excluding the CLE game I was doing to so rebut an argument that Collins was exposed as JAG by defenses as the season progressed (an argument it turns out you were not making).

I felt that his individual performance did not support this conclusion... that the manner in which you broke the season statistics into quartiles was misleading.  I watched the game and to me the CLE game was less probative of his individual performance because:

1.   Collins had 13 yards on his first three carries, the Browns changed their defense to sell-out against the run for the rest of the first half daring Flacco to beat them.  Collins went 3 for 13 followed by a 9 rushes for 6 yards.  

3.  On his 4th snap of the second half for B-More, Collins got knocked out and had to be examined on the side line.  He got only one more snap the rest of the game on the final drive.  

4.  After it appeared Collins was knocked out, the Browns went back to standard D.  They feared Collins more than Flacco, but they did not fear Allen.

4.  Allen went 10-59 after that (9 for 60 in the 4th quarter). 

To me, this game is less probative than the others.   To me, game script and injury were more responsible for his low YPC that day than his personal talent/ability.  It was more an indictment of Flacco than Collins and thus less probative for the purposes of evaluating whether there was a downward trend as the season progressed that evinced he was JAG and so therefore I excluded it.  

Finally you said, "I am starting think you also don't quite understand how facts work."

Fair point. : )  In fact, many an opposing counsel would probably agree with you.  

But the problem is, that stuff happens all the time. That’s part of the deal with RB’s. We can’t throw out a game because the script changed or was different. All RB’s face changing defenses all the time and it’s all part of their stats.

 

[kutta]

Ok, so change the DNP to a game where he played but only returned a kick.  I would not include that in my analysis of points per game played even if it technically met the definition.  This issue was responded to far better than I ever could by Arodin, who shows he probably knows more about statistics than the rest of us combined.

Here is the argument that I expected folks to make... what facts was I relying on to exclude the CLE game?

I think we are really in a discussion of how probative a given fact is.  If you remember, when I was excluding the CLE game I was doing to so rebut an argument that Collins was exposed as JAG by defenses as the season progressed (an argument it turns out you were not making).

I felt that his individual performance did not support this conclusion... that the manner in which you broke the season statistics into quartiles was misleading.  I watched the game and to me the CLE game was less probative of his individual performance because:

1.   Collins had 13 yards on his first three carries, the Browns changed their defense to sell-out against the run for the rest of the first half daring Flacco to beat them.  Collins went 3 for 13 followed by a 9 rushes for 6 yards.  

3.  On his 4th snap of the second half for B-More, Collins got knocked out and had to be examined on the side line.  He got only one more snap the rest of the game on the final drive.  

4.  After it appeared Collins was knocked out, the Browns went back to standard D.  They feared Collins more than Flacco, but they did not fear Allen.

4.  Allen went 10-59 after that (9 for 60 in the 4th quarter). 

To me, this game is less probative than the others.   To me, game script and injury were more responsible for his low YPC that day than his personal talent/ability.  It was more an indictment of Flacco than Collins and thus less probative for the purposes of evaluating whether there was a downward trend as the season progressed that evinced he was JAG and so therefore I excluded it.  

Finally you said, "I am starting think you also don't quite understand how facts work."

Fair point. : )  In fact, many an opposing counsel would probably agree with you.  

But the problem is, that stuff happens all the time. That’s part of the deal with RB’s. We can’t throw out a game because the script changed or was different. All RB’s face changing defenses all the time and it’s all part of their stats.

 

[Faust]

Observations from from the fourth day of Ravens training camp

Excerpt:

A run at the job

Alex Collins is the Ravens’ starting running back, but Kenneth Dixon lets you know he’s around, waiting in the wings. He broke off a 20- to 25-yard touchdown run during 11-on-11 action Monday by bouncing outside and down the left sideline.

Collins, though, hauled in a 30-yard pass when he beat rookie linebacker Kenny Young on a circle route down the left sideline. The Ravens want Collins to be more of a threat as a receiver out of the backfield this season.

Now he just has to learn how to pass-block better

 

[Brisco54]

But the problem is, that stuff happens all the time. That’s part of the deal with RB’s. We can’t throw out a game because the script changed or was different. All RB’s face changing defenses all the time and it’s all part of their stats.

I think you are actually agreeing with me.

That is true, but keep this in the context that we were discussing.  The point I was rebutting (apparently I was arguing with myself) was that the stats reflected Collins trended down as the season progressed and that this was evidence that he was JAG.  I was pointing out that the last quartile included a game that skewed the stats.

I support using his YPC for the entire year so that we are including all the ups and downs of the entire season.  If you break the season into quartiles and one of those happens to include a situation like the CLE game, then those stats for the quartile will be lower, but that is not evidence that the players performance was trending downward.

It would be like concluding age is catching up to Joey Votto because he only had an OBP of .300 for a three week period... but during that 3 week period he faced 4 or 5 Cy Young candidates.  Over the course of a season batters will face the whole range of pitchers... RBs will face the whole range of game scripts in a season.  Picking out a four week window and drawing a conclusion based on YPC seems foolish unless you look and consider each individual situation/game script.

 

[Brisco54]

Observations from from the fourth day of Ravens training camp

Excerpt:

I am loving all the talk of including him more in the passing game.

 

[kutta]

I think you are actually agreeing with me.

That is true, but keep this in the context that we were discussing.  The point I was rebutting (apparently I was arguing with myself) was that the stats reflected Collins trended down as the season progressed and that this was evidence that he was JAG.  I was pointing out that the last quartile included a game that skewed the stats.

I support using his YPC for the entire year so that we are including all the ups and downs of the entire season.  If you break the season into quartiles and one of those happens to include a situation like the CLE game, then those stats for the quartile will be lower, but that is not evidence that the players performance was trending downward.

It would be like concluding age is catching up to Joey Votto because he only had an OBP of .300 for a three week period... but during that 3 week period he faced 4 or 5 Cy Young candidates.  Over the course of a season batters will face the whole range of pitchers... RBs will face the whole range of game scripts in a season.  Picking out a four week window and drawing a conclusion based on YPC seems foolish unless you look and consider each individual situation/game script.

Not all game scripts are created equal. The original point, I believe, was that before he became an every down back, he had an outrageous YPC. That tailed off once he became an every down back.

I think most people agree that a third down, COP back will have a higher YPC than an every down back. But once you are an every down back, you can’t start cherry picking games based upon the defense or game plan. 

 

[Bracie Smathers]

His hands as far as catching the ball are actually pretty good,...

It would take a leap from what he showed last year in his route running ability to actually line up in the slot, regularly a LB and make the catch several yards down the field though IMO.

Fair enough observation but it sounds as though Morninweg is saying that he started to make progress in that direction last year before he hit a bump but this year he is not only picking up where he left off but that he has made progress.

” Mornhinweg said in a Sunday post-practice press conference “When we first got him, [running backs coach] Thomas [Hammock] really, really did a great job, and he was sort of morphing into a really good pass receiver – both from the backfield and split. Then he had a little setback there later in the year, and now he’s coming on again. So yeah, that’s a good point. If he can do that, that will really add to our offense.”

Allowing Collins to play as a slot receiver could certainly create problems for opposing defenses. Collins profiles as a dangerous mismatch for defenders — he’s too fast for linebackers to keep up with and he’s too big for defensive backs to get physical with. As a result, sticking Collins out in the slot creates the potential for him to get out in open space and expose defenses.

All points of Alex Collins having 'good hands' as a receiver with the speed to outrun LBers and the size to create a physical mismatch with DBs could all come together if his route running improved to the point where he can expose defenses.

Take another look at what Morninweg said, it could be just coachspeak or it could be a fair observation.

... he was sort of morphing into a really good pass receiver – both from the backfield and split.​

... now he’s coming on again​

 

[Chaka]

I think you are actually agreeing with me.

That is true, but keep this in the context that we were discussing.  The point I was rebutting (apparently I was arguing with myself) was that the stats reflected Collins trended down as the season progressed and that this was evidence that he was JAG.  I was pointing out that the last quartile included a game that skewed the stats.

It also included the game against Pittsburgh where he averaged 6.67 YPC.

The only reason I presented quartiles is because that's how PFR presents the data.

If context of game script is truly important then I would think you would want to give less weight to his first three games (25 carries for 206 yards) where not one defense gave him any thought in game preparation.

I already posted his numbers for games with 15+ carries, which is much more favorable than using game# splits.

I never called him JAG but I was around here for his hot start where people were going bonkers over his"talent" and pretty much making every excuse in the world for, or outright ignoring, everything that followed. 

The truth is not much has changed in Baltimore this year. All the problems with coaching and surrounding talent are still there. Teams game planning for Collins is still going to happen. Everything that made him go from 8.24 YPC on his first 25 carries to 4.1 YPC on his next 187 carries is still in play.

All I have been doing is cautioning against using 4.6 YPC as the baseline for his production.

 

[Cobbler1]

It also included the game against Pittsburgh where he averaged 6.67 YPC.

The only reason I presented quartiles is because that's how PFR presents the data.

If context of game script is truly important then I would think you would want to give less weight to his first three games (25 carries for 206 yards) where not one defense gave him any thought in game preparation.

I already posted his numbers for games with 15+ carries, which is much more favorable than using game# splits.

I never called him JAG but I was around here for his hot start where people were going bonkers over his"talent" and pretty much making every excuse in the world for, or outright ignoring, everything that followed. 

The truth is not much has changed in Baltimore this year. All the problems with coaching and surrounding talent are still there. Teams game planning for Collins is still going to happen. Everything that made him go from 8.24 YPC on his first 25 carries to 4.1 YPC on his next 187 carries is still in play.

All I have been doing is cautioning against using 4.6 YPC as the baseline for his production.

While I don’t necessarily disagree that Collins will still be the focus of opposing defense’s game plans, the offense will have 7 new starters. The majority of those I think most would consider upgrades.

 

[Chaka]

How do we contextualize Collins performance against Pittsburgh in light of the fact that the Steelers were playing on a short week after losing Shazier on MNF? Heck they didn't even know if Shazier would ever walk again at that point. I think you could argue that had an impact on Pittsburgh's preparation and performance.

And what can we say about that Miami game? I am not sure what happened there but it sure looked like Miami quit quick, fast and in a hurry in that one.

 

[Dr. Octopus]

How do we contextualize Collins performance against Pittsburgh in light of the fact that the Steelers were playing on a short week after losing Shazier on MNF? Heck they didn't even know if Shazier would ever walk again at that point. I think you could argue that had an impact on Pittsburgh's preparation and performance.

And what can we say about that Miami game? I am not sure what happened there but it sure looked like Miami quit quick, fast and in a hurry in that one.

You can nitpick positive factors and/or negative factors that affected certain games ad nauseam for every player in the league - at a certain point you have to rely on big picture data.

Collins is far from a sure thing of course, but he looked good running the ball, his o-line is healthy, and the Ravens have upgraded their WRs and TEs. On the flip side Kenneth Dixon (who many people like a lot) returns from injury and suspension to add more competition for carries.

Weighing offseason "coach-speak" (that Collins is the favorite to start) and the other positives, it looks like Collins should have a pretty decent season and be a solid RB2. His ceiling is likely low, and his floor, while not necessarily worry free, seems to be pretty stable.

It's up to individuals to determine what that is worth and whether he presents good value during the draft. For those in dynasty leagues it's a little more tricky, whether to buy or sell, but that's the nature of the beast with "fringe" type players.

 

[Brisco54]

All I have been doing is cautioning against using 4.6 YPC as the baseline for his production.

I can agree with that.  However the difference between a 4.6 and a 4.2 or even a 4.0 YPC does not have as significant an effect on how I value the player as others.

For example, I still rate Mixon pretty high despite him having a comparatively abysmal 3.51 YPC last year.

As someone pointed out earlier in the thread, the difference in a half yard in YPC over a season is not tremendously significant for fantasy points.

Where the 1/2 yard change in YPC does matter is if it threatens the players total touches.  Enter the threat of Kenneth Dixon.  However, comparing the rushing statistics, I am not sure Dixon looks in any way superior Collins as a rusher.

First, the complaint about using 4.6 YPC for Collins is because he had a great YPC as a Change of pace/rotational back that dropped once he was the feature back.  As pointed out multiple times in this thread, it is expected that change of pace/rotational backs will have a higher YPC.

The only stats we have on Dixon seem to be from when he was a change of pace/rotational back.  If we are being consistent, his 88 for 382 career (all from 2016) 4.34 YPC should be compared only to Collins 37 for 261 (the four games when he was a change of pace/rotational back before he became the feature back) that yields a 7.0 YPC.  At worst, if your are comparing results in Baltimore you have to use Collins 4.6 YPC.

HOWEVER, there is one big wrinkle and one small wrinkle.  The big wrinkle is that Dixon got his 4.34 behind a healthy Yanda and Baltimore line while Collins got his behind a line missing up to three starters including Yanda.  The smaller wrinkle is that Dixon's stats were as a rookie whereas Collins has had a couple of years of sideline seasoning.  I saw this is a small wrinkle because unlike other positions, the RB learning curve for rookies as far as rushing the ball is much less steep.

I have been worried about Dixon primarily for his pass catching skills, but if they are truly happy with Collins skills as a receiver, I think Collins has a very safe lock on 15-20 touches a game for at least the first 8 games in 2018.  I would be fine if those yield a YPC of 4.0 if he holds on to his red zone role he assumed in the later part of last season.  In a given game, the difference between 20 rushes at 4.0 and at 4.6 is 80 yards vs 92 yards... or just over one fantasy point.

Finally, I think the decision to pass on an RB in the draft shows Baltimore confidence in the entire stable of RBs.  The have a starter, back up and change of pace they are pleased with.  You can decide for yourself who they currently envision in each role.

 

[Cobbler1]

Dixon left practice today with a hammy injury.

 

[Interseptopus]

Dixon left practice today with a hammy injury.

That's what happens when you go off the juice. Your tendons, ligaments, muscles are more prone to injury. I always felt like that's how you could tell who was on something; those who got big and some performed really good over a short period of time, followed by a lot of pulled muscles etc. 

 

[hardcoredx]

Didn’t Gillislee have a great DVOA 2 years back, only to be proven a bum? I’ll buy low on Dixon. I’m not impressed with most 5th round talents. 

 

[Young 8]

Didn’t Gillislee have a great DVOA 2 years back, only to be proven a bum? I’ll buy low on Dixon. I’m not impressed with most 5th round talents. 

Well, then I suppose you're impressed with 4th round talents.

 

[Interseptopus]

Neither back is going to work out, and Baltimore will be drafting Damien Harris round 2 next NFL draft after taking AJ Brown round 1. 

 

[Chaka]

Didn’t Gillislee have a great DVOA 2 years back, only to be proven a bum? I’ll buy low on Dixon. I’m not impressed with most 5th round talents. 

IDK but even if true what does Gillislee have to do with Collins?

And you don't have to be a bum to fall out of favor with Belichick.

 

[JoeJoe88]

IDK but even if true what does Gillislee have to do with Collins?

And you don't have to be a bum to fall out of favor with Belichick.

I guess he's saying because they're both 5th round backs?  Kinda lost me too.  Howard and Ajayi were also 5th round picks, so...

Also, Gillislee did that being a chance of pace guy running behind one of the better backs in the league in McCoy.  He barely had over 100 carries.  Collins had over 200, and was shouldering the load for this offense a good chunk of last season.

 

[Brisco54]

We have reached that time of the year where you dread seeing a given player's thread make it back to the Shark pool front page cause it is normally bad news.

Thankfully, it appears this bump was no new news.

No news is good news for established starters for the next month..

 

[Interseptopus]

We have reached that time of the year where you dread seeing a given player's thread make it back to the Shark pool front page cause it is normally bad news.

Thankfully, it appears this bump was no new news.

No news is good news for established starters for the next month..

Alex Collins will be on page 1 until the Alex Collins question is answered once and for all.

We could only be so lucky to have Collins get hurt within the first 3 weeks, prompting further speculation while he is out on how good he could have been

 

[Chaka]

We have reached that time of the year where you dread seeing a given player's thread make it back to the Shark pool front page cause it is normally bad news.

Or, if you're a big fan, worrying that his ADP is skyrocketing prior to your draft.

 

[Brisco54]

Alex Collins will be on page 1 until the Alex Collins question is answered once and for all.

We could only be so lucky to have Collins get hurt within the first 3 weeks, prompting further speculation while he is out on how good he could have been

Which question is that?

At this point I am pretty convinced that, barring injury, he will be the starting RB for Baltimore week 1.  That's the only question I was hoping to get answered this offseason.

If you are asking how good he will actually be during the season... that's a question that exists for every single player in fantasy, and one that will not be answered during the preseason except by an injury...  so not one that merits constant discussion/life on the front page.

Until Dixon makes it back to the practice field or Collins gets injured, I am not sure what else there is to say about this guy in the next month

 

[shadrap]

Collins saved my season last year in two leagues, but it was maddening to see him replaced at times close to the goal line.  His stat line could have been off the charts coming down the stretch.  I worry about the same thing this year.

 

[cloppbeast]

We have reached that time of the year where you dread seeing a given player's thread make it back to the Shark pool front page cause it is normally bad news.

Thankfully, it appears this bump was no new news.

No news is good news for established starters for the next month..

A bump of the Alex Collins Thread means the universe is in working order. You should worry if it's not on the front page.