1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Take Away His 'X' Best Carries and He's Average

Discussion in 'Miami Dolphins Forum' started by DevilFin13, Apr 3, 2015.

  1. Kud_II

    Kud_II Realist Division

    3,662
    1,404
    113
    Oct 15, 2011
    Seneca, SC
    Lamar miller 1100 yards 5.1 ypc. I'd like to draft a RB in the 3rd or 4th, cause as good as Lamar was last season he's not an every down type of back.
     
  2. dolphin25

    dolphin25 Well-Known Member

    6,338
    2,400
    113
    Nov 22, 2014
    I think Miller has been good all along. His first year there often would be no holes for him to try to run through. Things have been opening up more the last couple years. He did seem to hit the hole harder and faster this past year, I love the guy.
     
    Fin4Ever likes this.
  3. Disgustipate

    Disgustipate Season Ticket Holder Club Member

    31,608
    55,632
    113
    Nov 25, 2007
    I'm not sure where you're getting these ideas from. Miami doesn't run that much read option. There are Chip Kelly influences, but it isn't what you're saying by any means. They ran a lot out of option-viable shotgun formations, but there wasn't much in the way of an "option". The runningback overwhelmingly took the ball even when the unblocked defender attacked the hand-off. The "true" option plays were a small minority. If you look at the Youtube video of all Tannehill's run plays from last year, you'd frequently watch Charles Clay arc blocking and looking for a scrape exchange that wasn't there. Teams weren't loaded up to stop the option.

    The idea that Lamar Miller got a huge benefit because he was facing a lot of 11 on 10 blocking situations just isn't correct.
     
    Tannephins likes this.
  4. Disgustipate

    Disgustipate Season Ticket Holder Club Member

    31,608
    55,632
    113
    Nov 25, 2007
    Bumping this with something I thought was interesting in the same vein:

    http://www.footballperspective.com/running-back-heat-maps/

    They produced a "heat map" for runningbacks. It determines the percentage for the distance of all of a players rushes, then color codes them. Lamar Miller looks like his quality as a rusher came largely from his strong production at producing 5-10 yard rushes.
     
    Tannephins and cbrad like this.
  5. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    That's pretty cool! It shows you one way of determining the relative influence of the O-line vs. the RB on a running play. Basically, the better the O-line, the higher the percentages for smaller gains. The better the RB, the better the percentages in the 5-10 yard range should be relative to the percentages for the small yardage gains.

    So, the more blue you see on the left, the better the O-line, while the more blue you see in the middle or right RELATIVE to the left, the better the RB. This says Lamar Miller is better than average at gaining extra yards IF the O-line opens a hole, but Miller is worse than average at making it past the 2nd level IF he gets that far.
     
    Tannephins likes this.
  6. Tannephins

    Tannephins Banned

    1,818
    572
    0
    Dec 23, 2014
    And that's partially consistent with this:

    http://www.footballoutsiders.com/stats/ol

    ...although you'll see that the Dolphins as a whole, to which Miller presumably contributed predominantly, were 9th in the league in the "open field."
     
    cbrad likes this.
  7. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Interesting...pretty much across the board, Miller had better numbers than Murray.
     
  8. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yeah, it shows the Dolphins are #1 in 2nd level yards (5-10 yards) while they drop to #9 in 10+ yards. I don't know why they don't show the data and rankings for 0-4 yards.
     
  9. jdang307

    jdang307 Season Ticket Holder Club Member

    39,159
    21,798
    113
    Nov 29, 2007
    San Diego
    Except for the 97 yard run he was less effective the more carries he got. Yes, he had the 97 yard run. But for 15 games he was worse as the game wore on. Not sure what his numbers would be if he ran it 25-30 times a game because he rarely touched 19.

    I am a fan of Miller. Huge fan. But he's never been the workhorse here.
     
    Piston Honda likes this.
  10. Tannephins

    Tannephins Banned

    1,818
    572
    0
    Dec 23, 2014
    Actually the correlation between his number of carries and his number of yards per carry, for his career, is 0.32. In 2014 that number was 0.45. So he actually performs better, in terms of yards per carry, when he gets more carries.
     
  11. Piston Honda

    Piston Honda Well-Known Member

    7,853
    8,088
    113
    Sep 23, 2014
    You missed the point entirely. Miller isn't a back who can handle 20 carries on anything approaching a regular basis w/o wearing down. Not only would that kind of work load leave him in bad shape come January it'd also shorten his career considerably.

    The staff limited his carries to prevent him from wearing down, Im glad they don't use your stats to make decisions bc Miller'd be a corpse by now.
     
  12. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    People throw out this idea that he can't carry the ball more...but what is it based on? Do you have quotes and such from the staff saying they limited his carries because he would wear down? I mean, if he's getting 15 carries a game, I'm not sure that an extra 5 a game are really going to kill him.
     
    Fin4Ever likes this.
  13. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Except that according to the link in the first post, Miller had 6 runs removed and Murray had 7, so they really weren't that far off. Biggest difference is the number of total carries, and whether or not Miller could carry it more.
     
  14. Piston Honda

    Piston Honda Well-Known Member

    7,853
    8,088
    113
    Sep 23, 2014
    I base it on there being very few backs who can handle 20 carries league wide, Miller's clear lack of enthusiasm for carrying that many times, and the staff's consistent efforts to keep him on the 12-18 range. It's not a knock on Miller as much as an acceptance of the norm.
     
  15. Tannephins

    Tannephins Banned

    1,818
    572
    0
    Dec 23, 2014
    Nonetheless, he doesn't perform worse when he does carry the ball more. He performs better.

    Whether he would wear down or not clearly isn't indicated in his yards per carry as a function of how many times he carries the ball.

    Now, if those correlations were negative, then certainly you'd have an argument. But they aren't.
     
  16. djphinfan

    djphinfan Season Ticket Holder Club Member

    111,651
    67,546
    113
    Dec 20, 2007
    not sure how anyone can rebuttal this..lol..im sure they will find a way bro.
     
  17. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    I guess the argument is that Miller may do well as long as you limit him to less than 20 carries per game, but if you consistently push it to 20+ carries per game, then his production would drop off.

    Of course, there's no way to know because Miller never carried it 20 or more times last year, and did so only once in the last 3 years, so we have no data on that. Given the numbers Tannephins is citing, coaches probably don't know what would happen either, and are probably limiting his carries based on common wisdom rather than something they've actually seen happen with Miller.
     
  18. Piston Honda

    Piston Honda Well-Known Member

    7,853
    8,088
    113
    Sep 23, 2014
    So you want the coaches to push Miller over a cliff in order to prove that pushing Miller over a cliff would hurt? LOL.

    Im looking at more than just YPC, the way he's running and finishing his runs is important in determining how he's feeling and whether or not he needs to be reigned in a bit.

    We're talking about a back who ran for 1100 on 5.1 per carry, while averaging about 13 carries a game. Preserving the kid's body is the only motive that makes sense, unless you want to say the staff was purposely trying to avoid winning games.
     
  19. Piston Honda

    Piston Honda Well-Known Member

    7,853
    8,088
    113
    Sep 23, 2014
    Watch the games. It's beyond obvious that they're limiting his touches. Given the results I'm not sure how anyone can rebuttal their approach. We need a 2nd back who eases or eliminates the drop off when Miller is out, it's the only way we'll be able to have any semblance of balance and tempo on offense.
     
  20. Tannephins

    Tannephins Banned

    1,818
    572
    0
    Dec 23, 2014
    You won't be able to find a post in which I've said that.

    What I'm saying, rather, is that the rationale for limiting Miller's carries because he performs worse when he gets more carries isn't supported by the data, and in fact it's refuted by them.

    There may be another rationale for limiting his carries that has nothing to do with how he performs when he gets more carries, however.
     
  21. Tannephins

    Tannephins Banned

    1,818
    572
    0
    Dec 23, 2014
    If they're basing their beliefs about that on his past performance, as they should, then they should believe his production would increase if they ran him 20+ times a game.

    However, as I said, there are other rationales for limiting the carries of a running back in the NFL, and I suspect they're basing their decision with Miller on those, rather than on his performance when he gets more carries.
     
  22. jdang307

    jdang307 Season Ticket Holder Club Member

    39,159
    21,798
    113
    Nov 29, 2007
    San Diego
    Not sure what you mean. Written before his 97 yard carry

    http://www.miamiherald.com/sports/nfl/miami-dolphins/article4706007.html

    So you are saying he doesn't perform worse when he carried it more? He did until week 17.

    He even admits the problem, being asked why this was the case:
    So you're saying this problem doesn't exist but there is an article about it and he was asked about it and admitted not knowing why that problem exists.

    But it doesn't exist.

    They talk about his weight going from 212 up to 220. I wonder if the extra weight is wearing him down. If he's at 225 now, we'll see very soon whether this is true.
     
  23. Tannephins

    Tannephins Banned

    1,818
    572
    0
    Dec 23, 2014
    The problem there is that you had six games in 2014 in which he had only 11 or 12 carries total in a game, and in those games his number of yards per carry on carries 11 and 12 (9 of them total) was 0.67. Excluding his 97-yard run, his number of yards per carry on carries 13 through 19 was 3.58.

    So, clearly there isn't a linear relationship between Miller's number of carries and his number of yards per carry, whereby his number of yards per carry decreases as he carries the ball more.

    Instead you have a non-linear relationship in which we see this "bouncing around" in his part, and when we pan back and simply look at the overall correlation between number of carries and yards per carry, we see the 0.45 figure I mentioned earlier.
     
  24. Piston Honda

    Piston Honda Well-Known Member

    7,853
    8,088
    113
    Sep 23, 2014
    Your data is weak. You're using a single metric, YPC as a substitute for overall performance.

    Miller averaged 13 carries per with a high of 19, IIRC he went over 15 only a handful of times. On such a small sample size of carries 15 and above a 97 yd carry would skew the data, making it appear that he had a higher YPC as he got more carries.

    Again, watch the games. See how Lazor used Miller in his game plans, usually running him to get the O off to a good start or limiting him early so he'd be available late. There was never a game where he rode Miller throughought. Instead Lazor used the short pass to augment the run and Williams/Thomas as ineffective backups to spell Miller when needed. Yet your data is suggesting that all he had to do was keep running Miller. It's silly to imply he'd use Miller's piss poor backups and/or expose Tannehill to the punishment he took behind a suspect OL/pass pro when all he had to do was give Miller 20-25 carries week in and out. But that's what your data implies.
     
  25. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    They might be basing it on past performances of RB's in general. What do the stats look like for YPC from the 20th carry onward vs. up to the 19th carry for the average RB in the league? If it gets worse from 20+, then it makes sense why they wouldn't let Miller carry more.

    So just because he gets better up to some point doesn't mean coaches should think that trend will continue if they go beyond that point.
     
  26. Tannephins

    Tannephins Banned

    1,818
    572
    0
    Dec 23, 2014
    The original statement was that Miller "was less effective the more carries he got." That isn't true. He was actually more effective the more carries he got, and if you're aware of a better measure of "effectiveness" for a running back than yards per carry, let's hear it.

    If you remove the 97-yard carry, the correlation between his number of carries and his number of yards per carry, on a game-by-game basis in 2014, is 0.24. Again a positive relationship, meaning that as he carries the ball more, his number of yards per carry increases.
     
  27. Tannephins

    Tannephins Banned

    1,818
    572
    0
    Dec 23, 2014
    http://www.footballoutsiders.com/stat-analysis/2007/370-carries-revisited

    http://www.nationalfootballpost.com/what-happens-after-300-carries/

    http://www.cheatsheet.com/sports/ho...-work-for-an-nfl-running-back.html/?a=viewall
     
    cbrad likes this.
  28. jdang307

    jdang307 Season Ticket Holder Club Member

    39,159
    21,798
    113
    Nov 29, 2007
    San Diego
    What is the correlation if you cut it off at 10
     
  29. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Good info.. confirms through stats what most would say intuitively: use a RB too much and wear and tear accumulates, hurting future performance.

    I'm still wondering though whether the performance itself on average drops off from the 20th carry onwards. It's asking the question of how wear and tear affect a RB during the game. Not expecting you to have the stats handy, but if you ever find out, please post.
     
  30. Tannephins

    Tannephins Banned

    1,818
    572
    0
    Dec 23, 2014
  31. jdang307

    jdang307 Season Ticket Holder Club Member

    39,159
    21,798
    113
    Nov 29, 2007
    San Diego
    No amount of spin will hide the fact on carries 11-20 he had less than half the YPC. That even HE, Mr. Lamar Miller, admitted it was a problem but has no idea why it was happening.

    It's a non-linear relationship that happened the whole year. You have to get through 11 and 12 to get to 13-19 and even then, there was a dip down to 3.58 which flies in the face of your argument that he gets better the more he carries. He's a full 2 ypc less 13-19 than 1-10.
     
  32. jdang307

    jdang307 Season Ticket Holder Club Member

    39,159
    21,798
    113
    Nov 29, 2007
    San Diego
    I may have missed it, please put your correlation formula up so I can look. I suspect I know what the formula is, and why it's faulty, but I may be wrong since I didn't see the formula yet.
     
  33. Tannephins

    Tannephins Banned

    1,818
    572
    0
    Dec 23, 2014
    Hypothetically speaking, what if on carries 18 and 19, he averaged more yards per carry than on carries 1 through 10, but it wasn't sufficient to alter the average on carries 11 through 20? What would that do to the argument that he "is less effective as he carries the ball more"?

    Surely you can see that with that example, the better way of measuring whether he "is less effective as he carries the ball more" is with a correlation between number of carries and number of yards per carry, and not with the artificial and arbitrary data split you're using (carries 1 through 10, and carries 11 through 20).
     
    resnor likes this.
  34. Tannephins

    Tannephins Banned

    1,818
    572
    0
    Dec 23, 2014
  35. jdang307

    jdang307 Season Ticket Holder Club Member

    39,159
    21,798
    113
    Nov 29, 2007
    San Diego
    ^^^ Not what I was looking for. Looking for the actual formula with his numbers plugged in.
     
  36. jdang307

    jdang307 Season Ticket Holder Club Member

    39,159
    21,798
    113
    Nov 29, 2007
    San Diego
    To get to 18-19, again, he always has to get through carries 11-12. Throughout a whole season I'm comfortable with that. If it happens against next season, then what?

    Again, I think I've pinpointed the fault with this correlation number but I need to see where you got the numbers.
     
  37. Tannephins

    Tannephins Banned

    1,818
    572
    0
    Dec 23, 2014
  38. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Yeah, I meant to post earlier, why split at 1-10? Why not look at 1-15? Or 1-5, 6-10, 11-15, and 15-20? Does 10 carries have some sort of weight?
     
  39. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    There's in principle nothing wrong with partitioning data into 2,3 or any number of parts (it's what histograms basically do) to see if there's some structure to the data that's being missed by not partitioning it that way.

    The problem is the data hasn't been presented in an easy-to-read format so it's hard to tell if there's some structure that suggests using (Pearson) correlation to describe the data is misleading, or that correlation is actually a decent stat to use in this case. The only other concern would be whether each bin has sufficient data points, but that doesn't look like a problem with 1-10 vs. 11-20 carries.
     
  40. jdang307

    jdang307 Season Ticket Holder Club Member

    39,159
    21,798
    113
    Nov 29, 2007
    San Diego
    Well it's obvious why the split is at 10, a round number. If you look at 1-15 Miller only has 12 carries all year that qualify as 16-30.

    1-5 is perhaps too small a sample.
     

Share This Page