Correlation between early OL picks and winning

[Orton's Arm]

Is there a relationship between taking offensive linemen early and winning? Not according to my analysis. I looked at how many offensive linemen each team took in the first two rounds of the draft, from 2001 - 2005. I then compared that number with each team's winning percentage in the 2005 season. There was no meaningful relationship between the two (R^2= -0.015).

 

This surprised me, because I'm firmly convinced a team needs a good offensive line if it's going to win games. Thoughts?

[generaLee83]

Is there a relationship between taking offensive linemen early and winning?  Not according to my analysis.  I looked at how many offensive linemen each team took in the first two rounds of the draft, from 2001 - 2005.  I then compared that number with each team's winning percentage in the 2005 season.  There was no meaningful relationship between the two (R^2= -0.015). 

 

This surprised me, because I'm firmly convinced a team needs a good offensive line if it's going to win games.  Thoughts?

753684[/snapback]

 

 

Sounds pretty interesting, if it's not too difficult and is feasible would you post some of the data.

 

I'm not doubting you at all, I like numbers and such.

[Orton's Arm]

Sounds pretty interesting, if it's not too difficult and is feasible would you post some of the data.

 

I'm not doubting you at all, I like numbers and such.

Here goes:

 

Team, #linemen, winning percentage

Arizona 1 0.312

Atlanta 0 0.5

Baltimore 1 0.375

Buffalo 1 0.312

Carolina 2 0.688

Chicago 1 0.688

Cincinnati 2 0.688

Cleveland 1 0.375

Dallas 3 0.562

Denver 1 0.812

Detroit 2 0.312

Green Bay 0 0.25

Houston 1 0.125

Indy 0 0.875

Jacksonville 3 0.75

KC 0 0.625

Miami 1 0.562

Minnesota 2 0.562

New England 2 0.625

New Orleans 3 0.188

NY Giants 1 0.688

NY Jets 0 0.25

Oakland 3 0.25

Philly 1 0.375

Pitts 1 0.688

San Diego 1 0.562

San Fran 3 0.25

Seattle 2 0.812

St. Louis 1 0.375

Tampa Bay 1 0.688

Tennessee 1 0.25

Washington 0 0.625

[Chilly]

Is there a relationship between taking offensive linemen early and winning?  Not according to my analysis.  I looked at how many offensive linemen each team took in the first two rounds of the draft, from 2001 - 2005.  I then compared that number with each team's winning percentage in the 2005 season.  There was no meaningful relationship between the two (R^2= -0.015). 

 

This surprised me, because I'm firmly convinced a team needs a good offensive line if it's going to win games.  Thoughts?

753684[/snapback]

 

Its not surprising at all.

 

If you look at most offensive lines, they're built through free agents and lower round draft picks.

 

Only one or two top picks on a line.

[MartyBall4Buffalo]

not all that suprised, I bet however the correlation between teams taking defensive linemen early and winning is higher.

[Typical TBD Guy]

The problem is that your formula, as you describe it, is based solely on quantity and not quality. Crappy teams that draft crappy OL players in the early rounds are (obviously) going to drag down the data correlation you're expecting.

 

A more relevant statistical endeavor:

 

For each draft class, make some cutoff definition of "success" - i.e. making the team, starting, or Pro Bowl appearances. Then evaluate the percentage of "successes" for each round, and see what you find (if you have the time ). My guess is that you'll find that better linemen are more likely to be found in the early rounds ... duh ... but the correlation will probably be stronger for linemen than it is for QB's, RB's, safeties, or LB's (again, if you have the time to put together all that data).

 

IMO, such statistical measurements are meaningless. The simple fact is that the OL accounts for almost 50% of your offense. If yours isn't good, then chances are your offense isn't either. And if your offense isn't good, then you know the rest.

 

At some point, a team that expects to have a good offense needs to build a good OL. And if this team is unable to build an OL through free agency or through the later rounds of the draft (i.e., our Bills dating back to the mid 90's ), then maybe it's time to switch strategies and draft them early. And if this team still can't find quality linemen in the early rounds (i.e. our Bills again w/ Mike Williams ), then the GM needs to go (i.e., Satan's son). And if quality linemen still can't be found (i.e. if Butler, Merz, and Pennington suck), maybe it's time to give the scouting department an overhaul? And if quality linemen still can't be found, then perhaps the team's fans should all just shoot themselves and end the misery right there.

[Dibs]

The problem is that your formula, as you describe it, is based solely on quantity and not quality.

753709[/snapback]

I think his formula relates well to the multitude of voices here at TBD who were(& still are) yelling "Why don't we ever take an OL early in the draft. All would be well if we did that."

Mind you, I think I would have covered the last 10 years of drafts. Impact OL in the first 2 years is rare and good ones will last generally a decade.

This would change the last 6 superbowl appearance teams to

Pit to 4

Sea to 6

NE to 4

Phil to 4

Car to 3

OAK to 5

TB to 3

And the bottom 3 teams from the last 3 drafts to

Hou N/A

NO to 6

TEN to 2

SD to 1

OAK to 5

ARI to 3

CIN to 3

DET to 6

I think this shows it is all pretty inconclusive. Mind you, I don't see any 1s or 2s in the first list.

IMO Kelso's Helmets premise that it is not how many drafted but how many good ones drafted is very true.

[eSJayDee]

I think the problem w/ your theory, is that you're not looking at enough data and a fair amount of what you're looking at is flawed &/or irrelevant.

If you drafted an OL in the 1st round in 2001 & he's no longer w/ your team (i.e. lost in FA, hurt, just sucked royal moose pud), what relevance does that have to your 2005 performance? Likewise, I'd be inclined to think that the fact that a 2005 draft choice is still riding the pine won't have much bearing on performance (One could even theorize that it's detrimental, cuz that pick could've been used elsewhere & contribute more.)

Also, cuz you're dealing w/ such minimal data (0-3) AND your that your 2 round cut-off is somewhat arbitrary AND a late 2nd round choice is virtually the same as a high 3rd round it just seems to me that there's too many varaibles to feel too confident in your conclusions. (Then again, I guess your conclusion is that drafting OL high is uncorrelated to performance.) But I think what I'm trying to say, is that all of the above

is the cause.

I do, however, commend you on your efforts to come up w/ a theory and try to quantitatively substantiate it. Now, can you tell me what the price of Gold is gonna do next week?

[Dibs]

I think the problem w/ your theory, is that you're not looking at enough data and a fair amount of what you're looking at is flawed &/or irrelevant.

If you drafted an OL in the 1st round in 2001 & he's no longer w/ your team (i.e. lost in FA, hurt, just sucked royal moose pud), what relevance does that have to your 2005 performance?  Likewise, I'd be inclined to think that the fact that a 2005 draft choice is still riding the pine won't have much bearing on performance (One could even theorize that it's detrimental, cuz that pick could've been used elsewhere & contribute more.)

Also, cuz you're dealing w/ such minimal data (0-3) AND your that your 2 round cut-off is somewhat arbitrary AND a late 2nd round choice is virtually the same as a high 3rd round it just seems to me that there's too many varaibles to feel too confident in your conclusions.  (Then again, I guess your conclusion is that drafting OL high is uncorrelated to performance.) But I think what I'm trying to say, is that all of the above

is the cause.

I do, however, commend you on your efforts to come up w/ a theory and try to quantitatively substantiate it.  Now, can you tell me what the price of Gold is gonna do next week?

753727[/snapback]

I think you're being waaaay too harsh. As I said to Kelso's Helmet, HA is addressing the concept espoused around here of "you must draft OLmen early to achieve a good OL". Going along this line, it is important to include ALL early picks, not just the successful ones. Yes, a bottom 2nd = a top 3rd but....you need a cut-off somewhere don't you? My only problem is what I stated in the earlier post....5 years is too small a sample since good OL will last for 10.

[Steven in MD]

Is there a relationship between taking offensive linemen early and winning?  Not according to my analysis.  I looked at how many offensive linemen each team took in the first two rounds of the draft, from 2001 - 2005.  I then compared that number with each team's winning percentage in the 2005 season.  There was no meaningful relationship between the two (R^2= -0.015). 

 

This surprised me, because I'm firmly convinced a team needs a good offensive line if it's going to win games.  Thoughts?

753684[/snapback]

 

First, either you meant r=-0.015 or your software is faulty, you cannot have a negative R^2.

 

Second,

 

Orlando Pace (Rams won a SB).

Jonathan Ogden (Ravens won a SB)

Walter Jones and Steve Hutchinson (Seahawks went to SB).

Except for the Patsies, most successful teams have a first round OL on the line.

[Bill from NYC]

Is there a relationship between taking offensive linemen early and winning?  Not according to my analysis.  I looked at how many offensive linemen each team took in the first two rounds of the draft, from 2001 - 2005.  I then compared that number with each team's winning percentage in the 2005 season.  There was no meaningful relationship between the two (R^2= -0.015). 

 

This surprised me, because I'm firmly convinced a team needs a good offensive line if it's going to win games.  Thoughts?

753684[/snapback]

 

Does this count?

[R. Rich]

Making snap judgments about where to take individuals is nearly as bad as just looking @ certain "indicators" (like 40 times) when making selections.

 

There are many examples of lower round offensive linemen who have excelled in the pros, just as there are @ other positions. And, people's sometimes idiotic reliance on 40 times is the reason a receiver like Jerry Rice is available later or a more recent example, S Anthony Smith of Syracuse, who had a poor 40 time and slipped in the draft, much to the delight of the Pittsburgh Steelers.

[MadBuffaloDisease]

It has been said that high-round picks get through college relying mostly on natural ability and not the fundamentals and hard work. Probably true to some extent.

[R. Rich]

To me, it's about having the ability to evaluate talent. Sometimes, you're going to get really good players in the high rounds. Other times, especially if your staff is good, you'll find guys who are able to play and more than willing to work for their shot. If you think there aren't many players out there like Kurt Warner (meaning they'd bust their tail for a chance to prove they can play pro ball), then you're kidding yourself. The key is having a staff that can accurately evaluate which players can make the transition and which ones are just dreamers.

[ROSCOE P. COE TRAIN]

Is there a relationship between taking offensive linemen early and winning?  Not according to my analysis.  I looked at how many offensive linemen each team took in the first two rounds of the draft, from 2001 - 2005.  I then compared that number with each team's winning percentage in the 2005 season.  There was no meaningful relationship between the two (R^2= -0.015). 

 

This surprised me, because I'm firmly convinced a team needs a good offensive line if it's going to win games.  Thoughts?

753684[/snapback]

 

 

 

 

Why only use 1st adn 2nd roudners. Expand to include ANY DRAFTED OL give us your information. Teams that win -- will have drafted OL...

[KRC]

There are too many variables to make this statistical analysis meaningful. What is the relative strength of the rest of your offensive unit? How is your defensive unit? You might be able to put a lot of points on the board, but if your defense cannot stop the opposition, it does not matter when it comes to winning. What about the strength of the other teams in your division/conference? How do injuries to your skill players affect the outcome of games?

 

While your premise of "a team needs a good offensive line if it's going to win games" may be good, the proof is not conclusive as it does not take into account other factors towards the success of a team (in fact, it just ignores them).

[Alaska Darin]

I think the paradigm is just too simple. Over the past few seasons, our team has suffered from mediocre QB play, poor in-game adjustments, and a loser mentality. It's easy to blame it all on the offensive line, but it's just not reality.

[R. Rich]

You might be able to put a lot of points on the board, but if your defense cannot stop the opposition, it does not matter when it comes to winning.

753815[/snapback]

 

 

Good one, Kenneth. I believe that may have something to do w/ why the Texans went after a defensive player w/ their first pick instead of a running back. After all, they're getting production out of the RB position in Domanick Davis. What would Reggie Bush add to that? A few more TDs? How does that help if you have a defense ranked last statistically in just about every category? You're still going to lose unless you address that. Even though it wasn't the sexy pick, I applaud Houston for going that route.

[Bill from NYC]

Good one, Kenneth.  I believe that may have something to do w/ why the Texans went after a defensive player w/ their first pick instead of a running back.  After all, they're getting production out of the RB position in Domanick Davis.  What would Reggie Bush add to that?  A few more TDs?  How does that help if you have a defense ranked last statistically in just about every category?  You're still going to lose unless you address that.  Even though it wasn't the sexy pick, I applaud Houston for going that route.

753852[/snapback]

 

It works both ways. In the mid to late 90s, the Bills had a star studded defense, especially on the front 7. There were names like Bruce, Speilman, Paup, Cowart, Hansen, Big Ted, etc., and you know what? Teams could eventually score because we had a lousy offense, with names like Panos, Nails, Ostroski, Spriggs, etc. A defense simply gets worn out when they atr on the field for too long.

Simply put, teams need an element of balance, but neglecting, or failing in attempts to build an offensive line are two tickets to long term losing.

[RJ (not THAT RJ)]

It works both ways. In the mid to late 90s, the Bills had a star studded defense, especially on the front 7. There were names like Bruce, Speilman, Paup, Cowart, Hansen, Big Ted, etc., and you know what? Teams could eventually score because we had a lousy offense, with names like Panos, Nails, Ostroski, Spriggs, etc. A defense simply gets worn out when they atr on the field for too long.

Simply put, teams need an element of balance, but neglecting, or failing in attempts to build an offensive line are two tickets to long term losing.

753861[/snapback]

 

 

Your last sentence is spot on, Bill, though it also contains a philosophical problem. Clearly there is a big difference between "neglecting" and "failing." To neglect would mean to do nothing, while it is quite possible to do many things and have them not work. The Bills' problem has been less that the O-Line has been neglected, and more that things have not worked out properly. That is not a defense of TD or JB or Marv, but rather an indication of the degree of chance that goes into sports success. One can draft OL men high, and find out one has drafted Tony Mandarich or Mike Williams just as easily as Anthony Munoz or Orlando Pace. One can sign FAs and sometimes get LeCharles Bentley and sometimes get Joe Panos--both of whom then suffered injuries after signing. You know what I mean.

 

To answer the original post in this thread, it appears that there is no single formula for creating a good OL, or a good team, beyond saying that one needs to find good players who play well together. How to do that is a constant challenge, and chance plays a huge role....