≡ Menu

When an athletic but raw player is drafted, it’s common to hear that he will succeed if he can be “coached up” in the NFL. That idea relies on the assumption that there’s going to be enough individualized coaching in the NFL for that player to reach his potential. But I’m not really sure if that is true, especially when it comes to less highly-touted prospects.

This was an interesting article about offensive linemen in the NFL who believe that the time limits on practices under the CBA “have forced NFL coaches to spend most of their time installing the offense, rather than focusing on the tricks of the trade. That’s led to sloppy play.” And Ryan Riddle, a former sixth round pick who now writers at Bleacher Report, recently tweeted something similar, saying that NFL coaches focus on the macro level rather than individual technique.

At the same time, as data takes a larger place on the national landscape, it’s often easier for NFL GMs to pick a guy with great measureables. I think GMs, like coaches, operate in a risk-averse mode where a primary goal job preservation. As a result, it may make sense to take a player with great measurables and mediocre film over someone with poor measurables and great film; if the workout warrior fails, the blame can fall on the coaches or the the player himself, while the GM is more likely to take the blame if the player with a low ceiling isn’t able to play at the next level.

This is all theory, of course. It’s possible that there is no less micro/technique coaching going on in the NFL now than there were 25 years ago, and that GMs aren’t more focused on athleticism now than they were 25 years ago. I’m not really sure how to study that. But if it is true, it would lead to a significant internal disconnect. At a time when coaches are spending less time on fundamentals, a smart GM would respond by taking more NFL-ready players and putting less emphasis on measurables and upside.

One good example of this, I think, is Germain Ifedi, the Seahawks first round offensive lineman. ProFootballFocus was highly critical of Ifedi, particularly with respect to his technique. But Ifedi shined at the combine — particularly in both jumps — and has the 36″ long arms that NFL GMs love. And Ifedi becomes an even more polarizing prospect when you consider where he landed: Seattle.

On the plus side, the Seahawks have Tom Cable, who is — correctly or incorrectly — regarded as one of the best OL coaches in the business. In theory, having a great OL coach should allow a team to take a chance on a high-upside player like Ifedi. On the other hand, Cable has been in Seattle for years, and the Seahawks have a bad track record when it comes to developing offensive linemen (James Carpenter, Justin Britt, and John Moffitt were taken in the first, second, and third rounds since 2011, the year Cable arrived in Seattle). And the team also has a track record of taking athletes and changing their position to play on the line (J.R. Sweezy, Garry Gilliam, playing Britt everywhere on the line even though he doesn’t seem to be doing well anywhere). Given that backdrop, is this a case of Seattle continuing to overthink the team’s ability to coach up offensive linemen?

But putting aside any specifics about the Seahawks, I’m curious to hear your thoughts on this tension generally. And, specifically, do you see any way to test this? Are organizations more willing now than 25 years ago to use a high pick on an athlete that needs to be coached up? And are teams worse (or, at least, given fewer resources) at coaching up players than before?

  • sacramento gold miners

    25 years ago we weren’t seeing the huge disparity between the style of college offenses and the NFL, and this is a factor in this conversation. A player with slightly worse measurables can jump ahead of another because of the offense a team runs, or off field issues. An offensive lineman can have great measurables, and still have problems with run or pass blocking. Many college offenses are helped by the wider hashmarks on the field, and utilize shorter pass plays which would be neutralized by pro defenses.

    Coaching does matter, but sometimes players like Jim Moffitt have other issues preventing success. It will be difficult to ever develop a test for this area.

    • I don’t think I agree with that — consider how many teams were still running the wishbone or option offenses 25 years ago! Yeah, I agree that the spread offense seems to make it hard to judge say, OL, but OTOH, NFL teams are copying college offenses now in a way that was not true in prior generations. It seems like the NFL steals more from the college game than vice versa, which is a change.

      This is also anecdotal, but I do think the NFL more closely resembles college football now than it did in the ’80s and ’90s. For one example, consider the recent series of posts showing that rookie and 2nd-year WRs are doing better now than they used to: http://www.footballperspective.com/receiving-yards-by-class-year-1950-to-2015/

      • sacramento gold miners

        There’s no doubt option offenses were more prevalent decades ago, but so were pro-style offenses. The NFL QBs of the 80s and 90s came from colleges where they were usually under center, and didn’t rely 100% on reading cards from the sideline. While the NFL game has more passing, the college game will continue to be different because of the hashmarks and the necessity of winning at the college level. I think the risk-reward of measurables will always be a question, and I agree with the offseason issue of practicing. We may yet see another developmental league emerge because of these issues.

        • Richie

          A QB in the spread offense is still much closer to what the NFL is doing, than a QB running the wishbone.

          It’s crazy to me to think that a guy like Irving Fryar was drafted first overall in 1984 as a WR, despite only 67 career receptions in college!

  • Corey

    I do suspect that NFL teams are worse at developing technique now, both because of CBA practice time limits and because schemes are generally much more sophisticated than they were 25 years ago, requiring a greater amount of practice time be devoted to scheme installation rather than technique.

    • Yeah, although that’s another thing that’s tough to measure: are schemes more sophisticated now? Are they more complicated to install? You used to hear about 900-page playbooks, but those days are long gone, I think.

      • Corey

        I can’t claim to be an expert on this, but my sense is that defensive schemes have gotten much more sophisticated while state-of-the-art offensive schemes have simplified a little. I think all the hybrid/exotic fronts, zone blitz packages, mixed/disguised coverages, pattern-matching, etc. were either in their infancy or didn’t exist at all in 1991.

        I suspect that the result of the huge growth in pass defense variations has been that it’s no longer feasible for an offense to pre-plan every possible response. The 900-page playbooks resulted from West Coast offense control-freak coaches wanting to have pre-planned responses to every defensive variation. So e.g. you’d have 15 plays in the playbook for one concept because you might run that concept out of 15 different formations (maybe more if you include all the motions and shifts) depending on what the defense shows or the particular matchups that week, and then the concept might have multiple route options built in that would all have be diagrammed separately for each formation, and each variation needed to be described separately in the offense’s language (leading to those self-parodying 20-word playcalls), and then installed separately at practice, etc. Teams have learned that this kind of obsession has diminishing returns, and that you need to simplify a little to keep things practical. They mostly get to a similar place now using built-in options rather than separately designated plays. I think Chris Brown has written about this.

        But recall that in 1991, while Walsh’s West Coast (as run by Holmgren at the time) was probably state of the art, a lot of teams were still using run-first schemes that were a lot less complicated. Now just about everybody has absorbed the West Coast emphasis on formation multiplicity and motions/shifts and so on. So while the most complex offense now may not be any more complex than it was in 1991 (or may even have been forced to simplify), I would guess that the median and lower quartile offense schemes are much more complicated than they were in 1991.

        • Dave Archibald

          My understanding is that the playbooks may have shrunk a little, but the plays themselves have largely become more complicated. Modern pass plays have a lot of decisions – stuff like sight adjustments, option routes, choices where the WR has to read the coverage and safety, etc. This shift has likely made the playbook smaller but probably requires more learning and teaching. “There are very few routes that are what we call ‘run-it’ routes — those are routes that stay on, no matter what. Usually, you have a ‘conversion’ of some kind.” Source: http://www.fieldgulls.com/2012/6/26/3118409/on-option-routes-conversions-sight-adjustments-hot-routes-and-the

      • Richie

        ” You used to hear about 900-page playbooks, but those days are long gone, I think.”

        Replaced by 128GB iPads. hehe

  • Adam

    I think GM’s have always been obsessed with measurables, often to their teams’ detriment, although I’m not sure this is any more prevalent today than it was decades ago. We often hear draftniks tout players who come from a “pro-style offense” (whatever that is), with the implication that said players should have an easy transition into the NFL. Of course we know that many of these players fail regardless.

    The two biggest traps GM’s routinely fall into are fixating on size and speed, while ignoring the glaringly obvious question: Can the guy play? The film tells the real story, not his height or 40 time. The vast majority of famous draft busts are comprised of players who possess NFL size and/or athleticism, but lack the skill or intelligence to play at the NFL level. This tendency is exacerbated by today’s media coverage of the combine and pro day workouts, in which foolhardy GM’s allow a brief snapshot of a player’s physical attributes to override dozens of games of film.

    • LightsOut85

      “whatever that is”

      When pundits say it, they could be lazy & just mean “he was taking snaps under center”. The real difference is what kind of players you have at the Y & H, and how you use them.

  • Phil

    you might be interested in this message board discussion, which sort of touches on a similar themehttp://coachhuey.com/thread/71605/sabans-recruiting-requirements (fwiw, I’m Pitt1980 in that discussion (Chris Brown of SmartFootball is spreadattack))

    —————

    on point, one of the things I said in that thread (and see no reason to disagree with now):

    “one of the things I took away from Moneyball was that it was common for scouts to think that they could take a prospect and teach him to have a patient approach at the plate, but when they went and tried to do it, they never actually had any success teaching people to do that, that maybe a patient approach at the plate was far more hardwired into a guy than it first seemed

    it seems like probably has a lot of crossover applicability to football, things like RBs doing a good job of picking wholes, or LB doing a good job of diagnosing plays and reacting to the ball, or QBs doing a good job of reading defenses

    those seem like things that can be coached up, but the degree to which players can take coaching, is probably pretty hardwired in

    or at least if I was planning recruiting strategy, I would think that working of that hypothesis would be a fruitful place to start”
    —————
    as to whether that could be tested-

    I think you could isolate positions, then see if certain measurables improve or reduce their actual AV to expected AV by draft position

    I think, what you’re actually measuring there is whether a variable is well priced in, which is what I think you actually want to measure

    example –

    is height well priced in to WR evaluations? It seems fairly straight forward that being tall is better than not being tall, I’m not sure if its straight forward whether its well priced in though (ie, were Steve Smith and Wes Welker bargain pickups because team over rate the impact of height?)

    how I would think to test that:

    you could split all the WRs into about 4 groups – 1) guys under 5’11 2) guys between 5’11 and 6’1 3) guys between 6’1 and 6’3 4) guys taller than 6’3

    then you chart where all the guys in those groups were drafted, the expected AV from those draft positions, then the AV that they wind up actually achieving

    then you see if there are any sorts of relationships, if you do that, and it all looks like noise, I think that suggests that height is relatively well priced into WR evaluations

    if it turns out that group 1 outperforms their AV draft expectations, and each group progressively does worse, that suggests that teams are overvaluing height

    or maybe group 1 and 4 do best, which suggests that being really tall, or really quick, is undervalued

    ———————

    anyway, that’s how I would think to approach that narrow question

    as to the more general question

    there are fairly plentiful draft profiles of guys (Mel Kiper as been writing draft profiles of guys for something on the order of 25 years)

    It might be interesting to do a textual analysis of all those draft profiles

    same idea as above, but now instead group them by whether something in the profile says something positive or negative about their ‘athleticism’, ‘character’, ‘motor’, ‘intelligence’, etc etc

    over the years, how draft profiles describe those things has changed

    but it’d be interesting, which of those things has a history of being well priced in, and which doesn’t

    ———————————-
    ———————————-
    ———————————-

    what the above comment glosses over, and probably shouldn’t, is the question of whether things go from being not well priced in to being well priced in (and really also vise versa)

    what makes this sort of analysis really hard, is that the NFL is a really reactive league, when one team has success with something, they get a lot of imitators

    think about how ‘tall cornerbacks’ are priced

    its possible that tall cornerbacks were underpriced, then the Seahawks drafted Richard Sherman, and every saw that a tall cornerback could play great, now everyone wants tall cornerbacks, so they may have gone from undervalued, to appropriately valued, or maybe even overvalued

    that sort of thing seems pretty common

    which makes it hard to know what the right amount of weight to give historical price (expected draft AV) is

  • Abe Froman

    If the CBA is limiting positional coaching, it suggests there should be a market for “graduate” study in position technique. It would put the onus on off-season improvement onto the player rather than the team. Basically a player who wants a second contract should be paying for a personal position coach/trainer in the off-season.

  • Andrew Healy

    I’d be very surprised if teams were worse at coaching players up than before, even with the constraints now. You have innovation on seemingly every other level (e.g., scheme, sports science). Seems hard to think backtracking is happening on technique.

    One interesting thing to look at with all of this is differences across position. At some positions (e.g. WR), college production predicts NFL performance much better than combine stats (with the caveat that those combine stats still matter, but they’re just already picked up by college production). In other roles (e.g. pass rusher), combine numbers are actually predictive. In general, I think a rough rule of thumb is that combine numbers (and physical measurables) fill the gaps when college production is hard to assign to a player. Speed and size are useful for running backs b/c RB stats are so much a product of blocking. CBs are, at least for now, hard to measure well and so teams will tend to rely more on physical skills than college performance.

  • Phil

    you might be interested in this message board discussion, which sort of touches on a similar a theme http://coachhuey.com/thread/71… (fwiw, I’m Pitt1980 in that discussion (Chris Brown of SmartFootball is spreadattack))

    —————

    on point, one of the things I said in that thread (and see no reason to disagree with now):

    “one of the things I took away from Moneyball was that it was common for scouts to think that they could take a prospect and teach him to have a patient approach at the plate, but when they went and tried to do it, they never actually had any success teaching people to do that, that maybe a patient approach at the plate was far more hardwired into a guy than it first seemed

    it seems like probably has a lot of crossover applicability to football, things like RBs doing a good job of picking holes, or LB doing a good job of diagnosing plays and reacting to the ball, or QBs doing a good job of reading defenses

    those seem like things that can be coached up, but the degree to which players can take coaching, is probably pretty hardwired in

    or at least if I was planning recruiting strategy, I would think that working of that hypothesis would be a fruitful place to start”

    —————

    as to whether that could be tested-

    I think you could isolate positions, then see if certain measurables improve or reduce their actual AV to expected AV by draft position

    I think, what you’re actually measuring there is whether a variable is well priced in, which is what I think you actually want to measure

    example –

    is height well priced in to WR evaluations? It seems fairly straight forward that being tall is better than not being tall, I’m not sure if its straight forward whether its well priced in though (ie, were Steve Smith and Wes Welker bargain pickups because team over rate the impact of height?)

    how I would think to test that:

    you could split all the WRs into about 4 groups – 1) guys under 5’11 2) guys between 5’11 and 6’1 3) guys between 6’1 and 6’3 4) guys taller than 6’3

    then you chart where all the guys in those groups were drafted, the expected AV from those draft positions, then the AV that they wind up actually achieving

    then you see if there are any sorts of relationships, if you do that, and it all looks like noise, I think that suggests that height is relatively well priced into WR evaluations

    if it turns out that group 1 outperforms their AV draft expectations, and each group progressively does worse, that suggests that teams are overvaluing height

    or maybe group 1 and 4 do best, which suggests that being really tall, or really quick, is undervalued

    ———————

    anyway, that’s how I would think to approach that narrow question

    as to the more general question

    there are fairly plentiful draft profiles of guys (Mel Kiper as been writing draft profiles of guys for something on the order of 25 years)

    It might be interesting to do a textual analysis of all those draft profiles

    same idea as above, but now instead group them by whether something in the profile says something positive or negative about their ‘athleticism’, ‘character’, ‘motor’, ‘intelligence’, etc etc

    over the years, how draft profiles describe those things has changed

    but it’d be interesting, which of those things has a history of being well priced in, and which doesn’t

    (that also seems like a fairly giant project, that I wouldn’t expect someone without a research assistant to do, but if I were an NFL team, with cheap labor to burn, I think that’d be interesting to know (for all I know, there are teams that have already done that project and are well on to more interesting ones))

    ———————————-
    ———————————-
    ———————————-

    what the above comment glosses over, and probably shouldn’t, is the question of whether things go from being not well priced in to being well priced in (and really also vise versa)

    what makes this sort of analysis really hard, is that the NFL is a really reactive league, when one team has success with something, they get a lot of imitators

    think about how ‘tall cornerbacks’ are priced

    its possible that tall cornerbacks were underpriced, then the Seahawks drafted Richard Sherman, and everyone saw that a tall cornerback could play great, now everyone wants tall cornerbacks, so they may have gone from undervalued, to appropriately valued, or maybe even overvalued

    that sort of thing seems pretty common

    which makes it hard to know what the right amount of weight to give historical price (expected draft AV) data is