Top of the Draft Positional Budgeting Trends

[Dibs]

If you aren't feeding the key position talent for 7 straight drafts, you sure won't be up there for long ;-)

Is that a fact?

[Mikie2times]

so, you admit that your analysis is basically worthless, because you have no control to compare it to. Without the league average, and the tendencies of the rest of the teams, there is no definitive conclusion that can be reached. Anyone that has done any statistical work will agree.

Even if that was proven I still wouldn't agree that it has statistical merit. I'm not acting like the top dog, but I know I do more football statistic work then any 5 posters combined threw out the course of the year. I've had to of logged well over 7500 hours with football stats on excel. Irregardless of that, one thing I have learned beyond any reasonable doubt, is correlation does not equal causation. I find correlations all the time, some dating back for decades, but inevitably these correlations have zero impact on being able to accurately predict future statistical patterns.

 

In this case I applaud the efforts to bring statistics or tangible evidence to any discussion, as few even attempt to do that, especially to the extent the original poster did, HOWEVER, the original poster would never be able to prove that drafting at blank position, for blank percentage, of blank early picks, means anything to overall chances of success. If that was the case all GM's would follow the same pattern, because by definition, that pattern would be the most successful. Instead, as we all know getting a good player at any position means more to the overall success level of teams, and the formula in which those players are drafted for all teams, both good and bad, is completely unique to that team.

[ax4782]

Studying the drafting trends of the way Super Bowl teams approach the Top of the Draft versus the Buffalo Bills (one of only 4 teams in the NFL to have missed the playoffs this Millennium) may offer some insight into why we’ve been one of the consistently bad teams in the league for an extended period of time.

 

Using the draft records of Super Bowl teams allows a look into how those teams have “budgeted” at specific positions at the Top of the Draft. This study does not establish whether these Positional Budgeting Trends are a conscious strategy on the part of all or any of the teams in the study, but the trends do represent contrasts between the players Super Bowl teams target at the Top of the Draft versus the positions the Buffalo Bills have been drafting.

 

The methodology used for the study follows the primary trending results.

 

A comparison of Super Bowl Draft Budgets versus the Bills looks like this:

 

Super Bowl Teams: Giants, Colts, Steelers, Pats*, Bears, Seahawks, Eagles, Panthers:

 

% of Draft Top of the Budget by Position:

 

Super Bowl Teams

 

DL 23%

DB 21%

WR 14%

OL 12%

TE 9%

RB 8%

LB 7%

QB 6%

 

Bills

 

DL 16%

DB 20%

WR 18%

OL 12%

TE 0%

RB 20%

LB 6%

QB 8%

 

A few substantial differences in tendencies:

 

Buffalo has used 59% of its draft budget in the study period for Offensive players, while the Super Bowl Team Draft Budgets favor Defensive selections more often than Offensive.

 

Buffalo has “outspent” the Super Bowl teams at RB and WR while “under spending” them at TE and DL.

 

This makes the following areas those in which Buffalo most widely bucks the Top of the Draft Trends of Super Bowl teams:

 

A) Bills have a higher Top of the Draft spend on Offense than Defense, contrary to the trend with Super Bowl teams

B) Bills have no TE selection at the top, whereas all but one Super Bowl team has spent a portion of their Top of the Draft Budget on the position.

C) Bills have spent a higher ratio of their budget on WRs versus DL, bucking the Super Bowl team trend of loading up on DL at the Top of the Draft

 

Every Super Bowl team except the Seahawks has a higher DL spend than they do at WR. The DL/WR ratio favoring the DL is common among 87.5% of the Super Bowl Teams. Buffalo is already out of balance on this trend, and a selection of a WR with the #11 pick this season would put us at a nearly 1:2 DL/WR ratio, a stark contrast to the almost 2:1 ratio favored by the Super Bowl Teams on average. (The ratio favoring DL over WRs is also a trait of recent playoff teams like the Cowboys, the Chargers and the Packers).

 

Every Super Bowl Team except the Panthers has a Top of the Draft investment in the TE position except the Carolina Panthers. The Bills have none.

 

Super Bowl teams are spending over 23%- or almost a quarter of their Top of the Draft Budget- on DL, while the Bills have committed less than 16%.

 

 

In order to compile usable information for the study, the following reasonable stipulations were adopted in order to establish a study group and time window:

1) Top of the Draft- This is represented by the first two rounds. The players selected in these two rounds represent the prospects that NFL teams have concluded are the best talent entering the league from college each season.

2) Draft Budgeting- To establish a position by position numerical score for each team, the study uses the sum of the specific draft choices in which each team selected players at each position during those first two rounds. In order to end up with a highest to lowest sum, the selections were counted inversely. Since there are 32 team picks in each round each of the first 64 picks is assigned the inverse of its position, with draft pick #1 being given a numerical score of 64 points, draft pick #2 counting for 63, etc.

3) Compensatory picks- Compensatory picks following the 64th pick of the draft were counted as 1 point in each case.

4) In establishing a window to study successful draft budgeting, the average number of years first round draft picks average playing for their original team (6-7) was used. The past 7 drafts were those considered.

5) “Super Bowl Teams” will be NFL teams who have won their Conference Championships over the past 5 seasons. This allows the Super Bowl rosters to have two mature draft classes entering the study and limits teams declining from bad contemporary drafting over the study window like the Super Bowl Raiders following the 2002 NFL Season.

6) Positions- Positions are defined by: DL, DB, WR, OL, TE, RB, LB and QB.

7) Percentages- Percentages are carried to the closest whole number.

 

This is all fascinating and very technical, however, there are a number of flaws in this analysis, as there are any time you rely purely on numbers to tell a story. First, in all but three categories, the Bills were relatively close to the league average, those three positions being RB, DL and TE. Let's take a look beyond the numbers and focus on the reality.

 

Most Super Bowl and Conference Championship teams have a solid base on offense at QB and RB. Buffalo has gone through a stable of backs since the beginning of 2000. We have drafted a number of players, including Willis, Marshawn, and Dwayne Wright. We also had Travis Henry on the roster and of course he was replaced by Willis. So we've had a lot of instability at that position. Once Willis was traded we had a screaming need at RB that had to be addressed. We did so, and had to pay for it. The revolving door at RB started with the Donahoe regime and has finally been put to a stop.

 

As for the DL, the fact is that we have drafted DL, just not as many, obviously. Buffalo, however, has not been neglectful of their needs on either line. Buffalo has brought in high quality free agents at those positions, choosing to go with a known quantity over an unknown. Buffalo has only had one top-five pick in the last ten years where you get a lot of that high end D-Line talent. Carolina may have gone to the Superbowl back in the early 2000s, but they haven't exactly been very good since then. They've had a lot of early picks. The fact is that DL hasn't been neglected as you suggest. And, the DLs they have drafted have turned out to be pretty good. McCargo looked dominant at times last year, and Kyle Williams has played very well. However, Williams doesn't fit into your calculation because he was a fifth rounder. We'll see how Stroud and Johnson do this year. Also, if you add in the free agents that we've brought in to fix the DL, I bet your numbers would change drastically.

 

The TE situation is the only one that I can't explain. Ever since Metzellars left that position has been a disaster on this team, except, arguably, for when we had Riemersma. Why they haven't drafted any TEs is beyond me. Maybe they felt they had more pressing needs in the defensive secondary or elsewhere on offense. I would say your number is off. We should not be at zero percent, as we drafted Kevin Everett in the third round of 2005. As we didn't have a first round pick that year, the third round was really our second round that year. Just saying. He was starting to come on before his horrific injury. So, saying that they have done nothing is a bit of a misstatement, but they certainly haven't done enough to fix such an important position.

 

However, your most important oversight is, looking at those Super Bowl teams, their needs at the end of the season were ALWAYS on defense. Lets take the Patriots for example. Their offense was never their problem. They always scored points. There was no reason to mess with the system they had. You can plug any receiver in and they'll get forty catches minimum. The logical choice was D-Line. The Giants have had the same problems. Many of their top defensive players had started to retire, and they had drafted a lot of offense in 05 and 06. Thus, in 07 it made sense for them to address their biggest area of concern and take either two or three DL in the first two rounds. That also probably skewed your calculations by the way. One team taking two or three DLs in one draft would tend to throw the average off. Just saying. Then look at the Eagles. Again, it wasn't their offense lacking for production that cost them the game. It was the fact that their D just couldn't stop anyone. Most of the Super Bowl losers needed defense to try and get just that little bit better. After all, NE's defense, even with all of those D-Lineman isn't really all that great. They're just rested and playing with a 25 point lead every game. When they weren't in that situation last year, and were needed to win the SB, they just didn't have what it took.

 

So, while I think this was a good post and was well thought out, you have to realize that you can make statistics say anything, without actually being accurate. They just look that way. You can't grade the current administration in the same way you look at the Donahoe regime. They are totally different with two totally different ways of looking at the game of football. Donahoe was an offensive guy. If you look at our last two drafts, we have been defense heavy. McCargo, Youboty, Simpson, Whitner, Williams, Poz, John Wendling, C.J. Ah You and Keith Ellison. Thats nine picks in two years for the D, including three for the DL in McCargo, Williams, and Ah You. You can't say they haven't picked at those positions.

 

Buffalo has been a team that drafts in places where they have NEEDS, the past two seasons, as opposed to the glamour pick years of Donahoe. That's the smart thing to do. Right now, after the FA moves, DL is NOT A POSITION OF NEED. WR, TE, and arguably CB are the top three positions of need on this team. I won't be unhappy if they take Derrick Harvey (though Quentin Groves looks like the better player), but I would be far happier if they took Devin Thomas, particularly if they got him after trading down.

 

Nice work, but I strongly disagree.

[2020 Our Year For Sure]

We should not be at zero percent, as we drafted Kevin Everett in the third round of 2005. As we didn't have a first round pick that year, the third round was really our second round that year. Just saying. He was starting to come on before his horrific injury. So, saying that they have done nothing is a bit of a misstatement, but they certainly haven't done enough to fix such an important position.

The guy's career consisted of 2 catches for 4 yards and 0 touchdowns. My condolences to him, but to say he was coming onto anything other than the free agent pool before his injury is a stretch.

[GG]

You could have saved yourself a lot of time by simply saying that you thought Bills should have drafted Tommy Harris instead of Lee Evans, and not gone through this statistically meaningless exercise.

[Bill from NYC]

You could have saved yourself a lot of time by simply saying that you thought Bills should have drafted Tommy Harris instead of Lee Evans, and not gone through this statistically meaningless exercise.

 

Wilfork would have been OK too, no? That is who R.Rich was screaming for that year.

 

I had forgot, but Wilfork was selected only one spot before you know who.

[GG]

Wilfork would have been OK too, no? That is who R.Rich was screaming for that year.

 

I had forgot, but Wilfork was selected only one spot before you know who.

 

Which is a completely different argument than saying that Bills have ignored the DL. An equally valid beef is that Bills were stupid for taking Flowers ahead of Howard, Schobel ahead of Shawn Rogers, Denney ahead of Alex Brown, and Kelsay ahead of Umenyura. While we're at it, they were dumb in thinking that both Ted & Pat were beyond their prime.

 

But there's no reason to manufacture data to "prove" that Bills have targeted the draft in a way that isn't consistent with other teams' trends. If anything, Bills have picked the second highest number of DLs in the top 64 draft spots over the last 12 drafts.

[AKC]

HOWEVER, the original poster would never be able to prove that drafting at blank position, for blank percentage, of blank early picks, means anything to overall chances of success. If that was the case all GM's would follow the same pattern, because by definition, that pattern would be the most successful.

 

Now you're adding subjectivity to the math- you're assuming that there's a "light on" in every front office. Look how hard it is to get fans to recognize that better teams use more early draft equity for defensive players- what's to say 12 team's front offices aren't just as oblivious?

[AKC]

you have to realize that you can make statistics say anything, without actually being accurate. They just look that way.

 

 

 

I would agree if we were talking about a statistical study- you can increase and decrease the blinders and end up discovering little if anything.

 

But the study here is simply a trending study- it uses objective draft history and points out clear, substantial differences between the way the best teams have done it for 10 years versus the Bills. How much that trend means is up to the person considering it- I offer it for consideration and you have done that. Thanks for the reasoned response.

[GG]

Now you're adding subjectivity to the math- you're assuming that there's a "light on" in every front office. Look how hard it is to get fans to recognize that better teams use more early draft equity for defensive players- what's to say 12 team's front offices aren't just as oblivious?

 

Now we're supposed to believe that multi-billion $$ enterprises who tape and study tendencies on how players line up before a snap to guess the direction of a play, devote 1,000s of man hours evaluating each draft prospect are too stupid to hire an intern to put numbers into a computer to see if there's a statistical trend on draft success?

[Bill from NYC]

Now we're supposed to believe that multi-billion $$ enterprises who tape and study tendencies on how players line up before a snap to guess the direction of a play, devote 1,000s of man hours evaluating each draft prospect are too stupid to hire an intern to put numbers into a computer to see if there's a statistical trend on draft success?

 

He might be on to something.

Seriously Gerry, how can the Bills go more than a decade without having a good tight end? How/why did they draft first round defensive backs when their OL sucked? How can the Lions spend untold early draft picks on wideouts.

 

The losses are right there in plain sight, yet patterns like this kept going.

[Ramius]

He might be on to something.

Seriously Gerry, how can the Bills go more than a decade without having a good tight end? How/why did they draft first round defensive backs when their OL sucked? How can the Lions spend untold early draft picks on wideouts.

 

The losses are right there in plain sight, yet patterns like this kept going.

 

It could be that they thought their FA acquisitions addressed the OL problems, and needed secondary help.

 

It also could be that the FO realizes that 1st CBs usually give you more bang for the buck than a high-priced FA CB. Just because you dont see this, doesnt mean that they dont.

[GG]

He might be on to something.

Seriously Gerry, how can the Bills go more than a decade without having a good tight end? How/why did they draft first round defensive backs when their OL sucked? How can the Lions spend untold early draft picks on wideouts.

 

The losses are right there in plain sight, yet patterns like this kept going.

 

My beef is that he's making up random arguments to complain that the Bills have sucked for nearly a decade.

 

I'd venture to guess that the fact that Bills have gone through 5 coaching staffs and different offensive/defensive philosophies in 10 years as the greater culprit than deviating from a simplistic drafting formula that has zero meaning.

[Bill from NYC]

It could be that they thought their FA acquisitions addressed the OL problems, and needed secondary help.

 

It also could be that the FO realizes that 1st CBs usually give you more bang for the buck than a high-priced FA CB. Just because you dont see this, doesnt mean that they dont.

 

Sure my friend, but the w/l records don't lie.

One thing I know from reading your posts is that you want the Buffalo Bills Football Team to win as much as I do. Bang for the buck is only good when you are winning, no?

Another talented, strong, physical player up front would help this team more than a fast, agile corner who almost picks off or breaks up passes, especially when playing in the elements where passing can be something of a moot point.

[Ramius]

Sure my friend, but the w/l records don't lie.

One thing I know from reading your posts is that you want the Buffalo Bills Football Team to win as much as I do. Bang for the buck is only good when you are winning, no?

Another talented, strong, physical player up front would help this team more than a fast, agile corner who almost picks off or breaks up passes, especially when playing in the elements where passing can be something of a moot point.

 

The difference is that i think you win by getting talented players at any position, including CB. The Bills woes this decade go far beyond simply drafting CBs*. It was more along the lines of bad overall drafting from 2002-2005, and a failure to upgrade the OL as well during that time. TD tried and tried to get by using 2nd rate upgrades along the OL.

 

*just a quick note: we've spent more 1st/2nd round picks on the DL (4) since 2001 than we have on the secondary (2) in that same span. The problems have not come from drafting CBs, but the overall neglect of drafting the OL in the early rounds (1-3), not just the 1st, and the refusal (until 2006) to spend FA bucks on the really good OLs.

[AKC]

Sure my friend, but the w/l records don't lie.

One thing I know from reading your posts is that you want the Buffalo Bills Football Team to win as much as I do. Bang for the buck is only good when you are winning, no?

Another talented, strong, physical player up front would help this team more than a fast, agile corner who almost picks off or breaks up passes, especially when playing in the elements where passing can be something of a moot point.

 

Without a doubt, it's always interesting to look at the way the best do it versus the worst. I should qualify that- for some of us it's interesting. In my opinion the best team's tendencies have some reveals in the draft- and it'll be fun to watch it play out over the years. One thing that is for sure- there's a difference in the ways the Bills have been doing it and the way the best teams have. Maybe our friend Ramius will take the time to expand the study to the whole league, and we can see some even more revealing patterns as every team's early investment is on display.

[Mikie2times]

Without a doubt, it's always interesting to look at the way the best do it versus the worst. I should qualify that- for some of us it's interesting. In my opinion the best team's tendencies have some reveals in the draft- and it'll be fun to watch it play out over the years. One thing that is for sure- there's a difference in the ways the Bills have been doing it and the way the best teams have. Maybe our friend Ramius will take the time to expand the study to the whole league, and we can see some even more revealing patterns as every team's early investment is on display.

It still wouldn't make a difference, correlation does not equal causation. ANYBODY who works with numbers will tell you that. I guarantee if you input overall wins over the last decade, with frequency drafted early at any position, side of the ball, group of players (like OL or DL) the correlation will not be stronger then .35 or -.35 for any of the positions you use. Your evidence when left without the outside variable of human manipulation will erode to nothing of statistical merit. Hence the reason no GM uses formulas such as this or ever will.

 

All teams value certain positions more then others, so GM's balance the risk of taking the 4th best CB or DT, and having him bust, then taking the first TE or OLB and having more potential for him to pan out. Some teams run systems that require more players at certain positions. Some teams fill needs by plugging in FA's. Each team is different. Over the long run you don't just draft a position, and magically become successful. It's far more important to draft good players, period. Any GM in football would agree with that.

[AKC]

Hence the reason no GM uses formulas such as this or ever will.

 

I don't really find any substantial disagreement with that at all- I'd be surprised if any great GM had a static positional formula they applied on draft day.

 

At the same time, I have no doubt all the good managers do have- if not a formal positional formula- an innate understanding of the more likely positional landmines. For instance, in this season where the WR position is so weak that there's not even a consensus top 15 or 20 pick, I'm guessing that no top GM would even think about reaching out for one of these likely early round busts early even if they had a great need at WR.

 

I never intended the study to support the existence of hard or static rules- but there is the chance it does offer a little bit of a look into the understanding that the best managers have about which positions are the more likely to make your team better in the long haul versus those that are risky enough to find a different means of addressing them unless you're so rooster-sure about a guy at a premium pick level like 11 that you know he can't miss and you take him because you'd take him even if you were picking at the 5 spot.

[Orton's Arm]

It still wouldn't make a difference, correlation does not equal causation. ANYBODY who works with numbers will tell you that.

You are correct about correlation/causation. A correlation between A and B can exist for any of the four reasons: 1) A causes B, 2) B causes A, 3) C causes A and B, 4) coincidence. With a large enough sample size, statistical tools can be used to reduce the probability of #4 to where it can safely be ignored.

 

Nothing in the above paragraph necessarily dooms all statistically-based efforts to analyze draft patterns. A while back, I looked at the composition of each team's starting offensive line. Each starting offensive lineman a team had acquired with a first round pick was correlated with two additional wins. Each starting offensive lineman a team had acquired with a second round pick was correlated with 0.5 of an additional win.

 

What do those correlations mean? There are several possible explanations: 1) Coincidence is always possible, even though what I found was statistically significant at a 95% confidence level. 2) Using a first round pick to successfully fill any position of need probably correlates with more wins. 3) The teams with the most draft picks on their lines might be better/more successful at building through the draft in general. 4) It might make sense to use high draft picks to build your offensive line.

 

Can statistical tools shed light on which of these explanations is correct? Indeed they can. Had I been feeling more motivated, I would have run this same kind of labor-intensive comparison for a few other comparison positions--LB, DB, etc. If the same correlation existed for those other positions, explanations 2 or 3 are most likely correct. If the correlation was weaker or non-existent for those other positions, explanations 1 or 4 are most likely true.

 

Supposing, for the sake of argument, that explanation 4 is true, what would it mean in terms of how teams should be run? Certainly it wouldn't provide teams with any "formula" to determine an ideal course of action. At most, it would inform teams that, given a choice between drafting two players of roughly equal caliber, they should choose the offensive lineman over a LB or DB. (This assumes, of course, that the players in question are at positions of need.)

 

As you have noted, the data AKC has gathered do not prove anything in the statistical sense. But those data provide an interesting basis for discussion. For example, the Bills have spent heavily (in terms of draft picks) on the RB position; while ignoring the TE position in the first two rounds of the draft. To me, that datum is indicative of TD's short-sightedness and desire for a quick fix. He used a 2nd round pick on Travis Henry, even though he already had Antowain Smith. A more far-sighted man might have used the pick on some other position--such as OL--rather than trying to achieve immediate gratification at the RB position.

 

Over the long run you don't just draft a position, and magically become successful. It's far more important to draft good players, period. Any GM in football would agree with that.

I'll agree with you, with one qualification. It's important to draft good players, and to keep them on your team for the duration of their careers. "First contract and out" isn't a very effective way to build a winning franchise.

[Dibs]

......But there's no reason to manufacture data to "prove" that Bills have targeted the draft in a way that isn't consistent with other teams' trends. If anything, Bills have picked the second highest number of DLs in the top 64 draft spots over the last 12 drafts.

You've inspired me....

Here are 1st/2nd/3rd DLmen choices for all teams over the last 8 drafts......10-6 records or better in bold.....5-11 records or worse in red.

 

Vikings 4/2/0

Eagles 4/1/3

Rams 4/0/2

Texans 4/0/1

 

Saints 3/1/1

Patriots 3/1/0

Cowboys 3/0/1

Jets 3/0/0

 

Chiefs 2/3/2

***Bills 2/3/2***

Ravens 2/3/0

Cardinals 2/2/3

Giants 2/2/2

Bears 2/2/1

Packers 2/0/3

Bengals 2/0/1

Browns 2/0/0

Jaguars 2/0/0

 

Broncos 1/2/2

Titans 1/2/2

49ers 1/2/1

Steelers 1/2/1

Colts 1/2/1

Seahawks 1/2/1

Panthers 1/1/2

Falcons 1/1/1

Buccaneers 1/1/0

Chargers 1/1/0

Raiders 1/0/2

 

Lions 0/4/1

Dolphins 0/1/0

Redskins 0/0/0

 

I conclusively conclude that any conclusion is inconclusive.