The Dolphins Make Me Cry.com - Forums

Just joined the forum.  Glad to be here.  Been a Dolphins fan since the early 1980s.  ;)

I'm wondering if people here would be interested in a thread devoted to discussing the Dolphins from an "analytics" point of view.

If you're unfamiliar with the concept, this article may be informative:

http://mmqb.si.com/2014/11/19/nfl-tv-broadcasts-advanced-stats

Welcome to the forums!  I work as a data analyst for our analytics product at work, so this would interest me.

It would be kind of cool.

Are you a Native Miamian?

Excellent -- thanks.  :)

Welcome please let everybody know how Jarvis Landry is worse than an average nfl wr. Thank you

We could certainly discuss that from an analytics perspective.  On what do you base that statement?

His yards per throw or something.

Data is great, but taking one piece of data to tell a story is silly.

It sure could be, and data could tell us that, as well, i.e., how silly (or not) it would be to rely on only one piece of data.

That is what is great about analytics. If you have the tool, you aren't just using one piece of data.

Stats are good, but only as long as (to paraphrase Lang...or Houseman) we don't use statistics as a drunkard uses a lamppost, for support rather than illumination.

Welcome to the boards, Dolfanalyst...get down with your funky data-crunching self. ;)

Thanks much.  ;D

There is also the converse, however, where we use no data and instead use the perceptions and opinions of others as the "lamppost."

Before you know it, you have a group of people who all believe the same thing, having had their perceptions and beliefs "validated" by each other, and the whole group is wrong.

Very true.  This is a great article in that regard:

http://archive.advancedfootballanalytics.com/2007/07/what-makes-teams-win-part-1.html

And then you have a whole group that is wrong because they pick and choose the data and feed it to others as gospel.

That can sure happen too.

Nobody has a monopoly on data, however.  If people believe there should be different data picked and chosen, they're certainly free to do that and enter it into the discussion, hopefully indicating why they've chosen that data versus whatever else was presented.

It's also good to have visibility into the source data. I know you're a fan of PFF. I am as well, and as you know, my account is no longer active and yours is based on the scheduling of renewals.

It's a shame because those premium stats were interesting to look at. They told multiple different stories from all sorts of different angles.

Yeah, use some stat like where does Landry rate in NFL History for receptions in the first two years of their career.  That will show everyone how much Landry sucks!   ::)

Welcome!  I'd be against adding more analytics to games while broadcasting.  I just want the entertainment of enjoying the game.  It's already annoying enough to listen to some of the broadcasters, and their pre scripted comments.  Just give me the yards, downs, etc. and let me watch.  Putting more data into the stream would be like adding a new military rank of Corporal Captain.

It is like a running back who has the most carries ever in the first two years of his career.

A running back that has 1157 yards and 10.5 yards per catch.  Yup, he sucks. ::)

a running back that has the most carries and is 103rd in the nfl in yards per carry.

Landry was -7% in DVOA and was ranked 56th in DYAR, one above Ted Ginn.

I think the proper question about Landry from an analytics perspective is, what skill is being measured most by the statistics of his that distinguish him from others at his position?

The next question would be, how much does that skill contribute to winning in the NFL, in comparison to other skills of players at that position?

At that point you'll have Landry's relative value.

The trap, as with all Dolphins players, is to fail to do the league-wide comparison in this regard, and stop at a surface-level appraisal of Landry, simply because he's a Dolphins player.

If the team braintrust engaged in that superficial level of analysis, it would run the risk of having a bunch of overpaid players whose skills, and whose contributions to winning in the NFL, don't warrant their salary cap hits.

The team can do better than that, and so can we.

Fortunately, the Dolphins still have two full years of Landry's rookie contract left. At $1M a year, he's clearly great value.


It's important to distinguish between "what value does a PLAY contribute towards winning" and "what value does a PLAYER have". DVOA and DYAR measure the former and while there is certainly a correlation between the two, they are not identical.

If Landry gets thrown the ball on a bunch of 5 yard hitches on third and long, that's clearly not (generally) going to contribute significantly towards winning. DVOA and other advanced "win" metrics (which I'm a big fan of, btw) are going to score those plays negatively, but that doesn't mean Landry didn't perform great on those plays. If the team never throws the ball his way in the red zone, that's also going to negatively impact his advanced (and traditional) stats.

That's why it's important to use statistics to illuminate things, not as the proverbial (and previously mentioned) lamp post aka crutch.

In Landry's case, if you look at the statistics that he has more direct influence over, he looks like a top 10 receiver. His catch rate is very good and his yards after catch place him among the league elite.

http://espn.go.com/nfl/statistics/player/_/stat/receiving/sort/receivingYardsAfterCatch

If you look at that list, you'll see two tight ends and three running backs in the top 10. Of the five receivers, Landry has the second most YAC and is the ONLY receiver without a fumble.

I agree that it would be very difficult to find fault with Landry at that salary, given even just the volume of his production, without getting into any of the particulars of it.


Very good points there.  Thanks! :)

Actually its nothing like that, you aren't even comparing apples to oranges, more like comparing red meat to cucumbers.

Receiving a hand off is as easy as catching a pass.... isn't it???  ???

It probably isn't, but if the percentage of drops by receivers happened to be very low across the board, then the argument would have some merit, in that there would be very little variation in catching ability across receivers in the league.  If there is such little variation in it, then it wouldn't be a whole lot different, practically speaking, from being handed the ball, when you're comparing receivers to each other.

If just about every receiver is excellent at something, then no receiver is distinctive in that area, and so that skill or ability becomes all but meaningless in comparing them to each other.

If you want to use some analytics, use team analytics, they are much more revealing.

Here's one for you, turnovers forced 16 (rank 29th). 13 INT's and 3 Fumble recoveries. Oh and by the way, one of the INT's was fumbled and given back to the opponent without the offense ever seeing the ball.

Miami's turnover margin wasn't too bad, only -3 but that's because the offense did a fairly good job of not turning the ball over with only 19 (one of those was a defensive player) tied for 8th in the league, but the defense was terrible only forcing 16 turnovers. Turnovers have a tremendous impact on a game because of the change in field position. Not giving up points is big for a defense, but creating turnovers is even more important because not only do they prevent points, but often times it means giving your offense favorable field position which leads to points.

Want another team stat? Average starting field position. I don't have the stats for it, maybe Rich can provide it, but I'd be willing to bet that Miami's average starting field position for their offense was inside their own 35 yard line and possibly inside their own 30. I did a quick look at Miami's first 3 games and the average starting position was inside their own 27 yard line. They started 19 of 33 drives either at the 20 or inside their own 20. That's primarily because Miami's defense either let the opposing team score or allowed them to drive far enough down the field to pin their offense deep in it's own territory.

Now I don't put all Miami's problems on the defense, the offense sure had it's fair share of ineptitude moving the ball resulting in good field position for the opponent, but all you people that blame Tannehill for not doing enough on offense, need to take a good look at the defense and realize that the defense has gotten worse every single year that Tannehill has been the QB while Tannehill and the offense has gotten better 3 out of 4 years. Tannehill and the offense aren't solely to blame, the defense has to take a LOT of the credit.

Good stuff.  Turnover margin is indeed hugely predictive of winning.  That's been a key ingredient in Carolina's extreme success this year.

With regard to your starting field position stuff, I've often thought that one area in which analytics could help teams is in advising them when not to return a kickoff out of the end zone.  I suspect that with the changes in the rules on kickoffs, which allow kickoffs from the 35 (formerly the 30) and only a two-man wedge for the receiving team, the statistics would show that the minority of kickoffs into the end zone are brought out to at least the 20.  It'd be therefore better off to down those kicks in the end zone and take the ball on the 20.

So often we see a returner bringing a kick out from six or eight yards deep in the end zone, aided by only a two-man wedge, against a kicking team that took off from its 35.  When that happens I'm sitting there slapping my head, wondering why they don't just down it.  I don't have the stats for it, but it sure seems like the majority of those kick returns are stopped somewhere between the 10 and 15.

As for the Dolphins' defense, it was indeed terrible this year.  The addition of Suh caused none of what was expected for the defense overall.  This is primarily why we saw a regression in the team's record, despite that Tannehill and the pass offense played largely like he did last year, as did the running game.

EVERY team had an average starting field position inside their own 35 and only the top 4 managed to crack 30 (and just barely). The Dolphins were indeed bad, averaging a start 25.54 yards (27th) from their own endzone. However, considering that places us only 6 yards out of first place, I'm not convinced there's any major effect here.

Looking only at starting field position after kickoffs, things get a lot tighter. The league average is the 22 yard line and the standard deviation is only 1.2 yards. Miami was 24th with an average start just north of the 21 yard line.

(source (http://www.footballoutsiders.com/stats/drivestatsoff))

In reality, to make any meaningful "starting field position" analysis, we need to know if the major assumption made here is actually valid (and I doubt it is):  Is starting field position linearly related to scoring?

If that is not that case, then talking about the average makes little sense if we want to know the effect on games.

I'm 100% sure that someone has run the numbers and figured out exactly how the probability of scoring varies with field position. Googling, I came up with the following:

http://phdfootball.blogspot.com/2013/06/field-position-and-scoring.html

At first glance the result is surprisingly linear, but a closer look reveals a few issues: It appears FGs and TDs and treated equally when looking at the black graph (the close-to-linear one) representing "probability of scoring by (first down) field position". Second, safeties and the negative effect of increased chance of the other team scoring isn't included.

Weighing touchdowns and field goals by their point values would result in a significant up-tick the closer to the opponents end zone one starts. At the other end of the field, correcting for the risk of safeties and the cost associated with the risk of really bad field position, would cause a downturn likely into the negative.

The end result would be a distribution that matches conventional wisdom pretty well: Starting way back is VERY bad, starting on your opponents half of the field is VERY good.

Without an accurate distribution (and not just my hand-waving), it's not easy to know which kickoffs should be returned and which shouldn't. Hangtime and placement of kick are, of course, significant issues that cannot be ignored, but even so, it seems plausible that returning a kick from deep in ones own endzone is a valid course of action provided that the returner can reliably get some distance (perhaps 15-20 yards) out. That is going to depend in large part on hangtime and coverage and, as such, is going to be something that a returner would need to evaluate before returning a kick.

One seemingly intelligent approach would be to have the player alongside the returner take a look at how far down the field the coverage team has gotten just before the return man catches the ball (which would incorporate the hang-time of the kick), make a determination based on that and where the returner is, and communicate the decision to the return man.  We see this often, but again, my feeling is that too many of those returns in the end zone are being brought out.

If starting way back in one's own territory is indeed very bad, then obviously you want to avoid that, and go ahead and take the ball on the 20 when it makes sense to do so.  That also avoids the possibility of a fumble on the return, which, while rare, probably results in points for the opposition the vast majority of the time.

The frequency of "smashing success" on kick returns just seems so rare that the strategy instead ought to be one that minimizes risk.

^ There are teams that employ a secondary guy to basically tell the returner when not to take it out of the endzone. There was one instance in the game against the Patriots last Sunday when the (NE) returner looked like he was going to bring the ball out and was then stopped by one of his own guys visibly signalling for him not to return the ball.

6 yards difference on 184 drives. That's 1,104 extra yards for the season that Miami had to move the ball then the team in 1st place. That's about 70 yards a game. You don't think that's significant? I disagree.

You can sum up the yards, but that doesn't necessarily make for a valid comparison. Instead you'd need to look at what value 6 yards represent and see if that extra "expected scoring" would be in any way meaningful for any game. Hint: It would not.

Not the 6 yards, anyway, which was my point. Scoring isn't linearly related to field position, so while starting on the 26 or 32 isn't going to make much of a difference, starting on the 15 five times and on your opponents one yard line once WILL make a difference compared to starting on the 29 every time.

You can blow it off as nothing too, doesn't necessarily make it a valid position.

After reading the forums for the past two weeks, let me expand my prior statement above to add, PLEASE less statistical crap in forums also.  It detracts from the enjoyment of the site.

BTW, where's MikeO been hiding?  Did he go and get suspended again?

He's not coming back. He sent a farewell email to all the fantasy football participants that there are too many morons on this site for him to enjoy himself or something.

I don't think there are too many it's just that the morons (and the non dolphins fans) are rather loud. Would really help this site a ton if the "ignore" button worked. If it were me (and it's not, so I'm not complaining), it'd be the first thing that I'd get fixed.

He was browsing here as recently as today. When it comes to MikeO, I can take him or leave him. I appreciate the fact that he is a true Phins fan, and he definitely was prolific in his posting. I wasn't really a big fan of the "Gospel According to MikeO" attitude though. That's really the only thing I gave him grief over.

He couldn't come back if he wanted to. He was banned for (repeatedly) making posts about my wife and daughter.

Good riddance. That farewell letter to the fantasy football folks exceeded the FDA's recommended daily allowance of smarminess.

He's actually a great and intelligent poster ... he just needs to learn some communication skills. Opinion and fact are two different things and he has a really hard time of understanding that others do not take his opinion as fact. As well, he attacks others opinions as if they have no right to have one. What is the saying ...people who are always right end up being the loneliest people. It's sometimes better to be quiet than right. Haha ... I guess that can be said about a lot of people in here.

I must have missed that...

This about sums it up for me, like Stroke said I wasn't a fan of the "Gospel according to MikeO".  He brought a lot of good information though, and some bad.

I'm sorry I've got nothing to add.  :-[


This comment made me laugh,and has stuck in my head all morning,and I've chuckled about it all day.

Can't say I will miss him!!!!!  8)