Chalk Talk Discussion - BPA vs. Need

[KDawg]

We have this conversation fairly often around here, but I wanted to put this in one place.

First and foremost, no. I'm not counting this team out yet. But the inevitable "What's the proper strategy" conversation has already come up quite a few times around here. Some people are stating we need to use our top pick on a safety or corner, and use our third rounder on a right tackle. Others are saying that wide receiver and nose tackle need to be our priorities.

What a lot of people (I say a lot because there are a few) aren't saying is: Let's see what's available before we decide what we should pick.

I've flipped back and forth myself on which strategy is more important between Best Player Available and Need. It's a riddle wrapped inside an enigma. Best player available makes a whole lot of sense, but does it make sense if the best player is a quarterback and you just drafted a quarterback? The money tied up in the quarterback position alone would scare you away from that.

Most of the time, people answer that question with, "well, that's a different case, you go best player available that's not a quarterback". Which is a logical and structurally sound answer. Except the problem there becomes you're no longer drafting (or signing) a player based on BPA, you're now including need.

I see a lot of people believe it needs to be one way or the other. Draft for need, draft by what's available. I disagree with that assessment. With all things, the answer probably lies in the middle, and that's what I (in my opinion, I am not some kind of mastermind, and I don't pretend to have all the answers) firmly believe is the case here.

I think BPA should have a higher weight in the formula, but I think need has to come into play. To do that, I would find it necessary to point out the positions that I believe are the weakest, and after meeting with my staff and my general manager, asign them a ranking in order of how much we need them. I would limit myself to five positions of need. Why? You only have a limited amount of resources.

If, though, one year you have 16 draft picks and an extra $35 million to spend, you can expand that list. If you have 3 picks and $5 million in cap space, you narrow it down further.

So if I decided my top five positions of need were:

Free Safety
Cornerback
Right Tackle
Wide Receiver
Defensive End

And they were in that order, I'd assign a number to each, determining their value. Anything that isn't one of these positions gets a static value that is slightly less than the lowest value. And all the "need" positions wouldn't have to have unique values. If I determined that right tackle and safety were equally my biggest need, I'd assign them both a similar value.

I'm no mathematician, in fact, I'm fairly math stupid. However, I believe that this value shouldn't shift a player's overall value too much. If you give this value too much weight it puts entirely too much weight into team need. Team need should play a factor, but not an overwhelmingly large one. Perhaps a 1-10 number.

Next, I would need to grade all players, free agents or draftees, based on many different factors... FBI (football intelligence), scheme fit, skills, production, explosion, etc. Need would not go into my individual player evaluation. I'd likely make this scale in a 1-100 scale to allow for some variation.

After I was done grading out each prospect, I'd add my "need" value to each of their scores to create an overall BPA/Need number, and create my draft board accordingly. Maybe just add them. I don't know. Again, if I were in a position where this really needed to be figured out, I'd put some great math minds to work to help me come up with a valid formula.

Again, I don't pretend to know everything. But I think some sort of hybrid model, not necessarily my own by the way, is necessary when determining draft and free agency strategy, otherwise you wind up like the Lions under Millen.

And of course, there's always a little something called "the gut". When it's your choice and you have a player that you have graded higher than the rest and there's another guy out there who you really like that you think could contribute, sometimes or coach or GM is going to say, "We're going off the grid with this pick, please believe in me". But that can't be a regular occurence.

[Oldfan]

No reasonable mind would want to draft purely on BPA or purely on NEED. If your board shows a 75 grade for one player and a 72 for another, and you need the lower graded player more, go for him. We BPA proponents say only that stretching too far is dumb. It's hard enough to draft well. Ignoring your draft board to reach because of need only decreases the odds against picking well.

The BPA is also team-specific. He has to fit your scheme. Position value is another factor that can be team and scheme specific. The QB position is worth more in some schemes than others, for example.

We who favor BPA also argue that a surplus at one position can be intelligently handled to add value to the roster, but this requires follow through. Fred Davis was definitely was a BPA pick, but there was no follow through.

Fred could have been used with Cooley in a base two-TE set like the Patriots are doing. He also made possible the trading of Cooley, making our roster younger without a drop in quality. Neither scenario happened.

The draft and free agency are sources of talent. But they are not like Home Depot. You can't go in with a shopping list of needs. Roster building is more about recognizing opportunities and seizing them.

Bottom line: Drafting with too much emphasis on need is a win-now strategy that will hurt your draft batting average.

[KDawg]

I very much agree with your post. In essence, you said the same thing I did. The addition you made that I think is an absolutely phenomenal point is the follow through aspect. Follow through has to happen if you select based on BPNEED strategy. If Davis is the far superior player on your board to the point it trumps need, then you must utilize him appropriately or else you basically missed on the draft choice.

[Gibbs Hog Heaven]

Interesting theory Coach on the old chestnut of need vs. BPA. And there's no right or wrong way of course. It depends on a particular circumstance at a particular time. As well as great trust in your scouts, and a WHOLE heap of luck.

I've always been a need at the top end of the draft, BPA if that outweighs what's there for your need the lower end. If a BPA fell in my lap at the top end over need, but it was a position I was stocked in already; then I'm looking for willing partners to trade down.

So whilst I wouldn't put BPA higher per se, I think an amalgamation of the two dependent on individual circumstance will see you right more drafts than not.

Hail.

[DogofWar1]

Nice chalk talk, I like reading these a lot.

Ah, the age old battle. Like Dogs v. Cats, BPA v. Need has a long and storied rivalry.

Ideally, we could draft BPA every time. Unfortunately the most important consideration, IMO, is which player in the draft one expects to provide the greatest increase in W/L.

NOTE: This would be my personal ideal drafting strategy. It should be noted it would be incredibly hard and work intensive to put into action, not to mention with a fair amount of guesswork and arbitrariness.

I think you have to approach the situation from two ends. First, you need to establish a list, 1 - 300+ of the players in the draft, by skill. That's a tough job, but that's where you get BPA from. Second, you need to establish the list of positions of need, top to bottom prioritized.

For the Redskins, IMO, those positions are (in my priority order):
- Free Safety
- Right tackle and Cornerback (I seriously can't decide which to put over the other. On the one hand, I don't want Polumbus at 2013 starter. On the other hand, another year of Hall/Wilson is very unappetizing)
- **Strong Safety** (this is dependent on Meriweather)
- *ILB* (dependent on Robinson, if he develops, drop this to honorable mention)
- NT

Honorable mentions go to OG, WR, TE (dependent on Davis, if healthy, remove entirely), and DE (I think Cofield is a secret DE in waiting for once we get an NT).

Finally, having finished the BPA list and the needs list, I think you try to assign every player a "improvement to W/L record" value. Again, that's really tough to do, but probably needs to be done.

Once you have that, you can determine exactly how beneficial each pick should be, in theory, for your team.

An example scenario would be something like this:
Redskins are on the clock in the 2nd, middle of the round. Best players at the need spots are (for the sake of this accept the grade I assign them):
FS/S: Matt Elam (late 2nd)
OT: Oday Aboushi (mid-2nd)
CB: Kyle Fuller (mid-2nd)
ILB: Shayne Skov (mid-2nd)
NT: Louis Nix (late 1st)

Let's also, say, for the sake of this, Nix is the BPA.

To decide between the five, you weigh all of the players based on expected win increase, and then pick the highest. If, for example, we expect adding Elam to increase our wins by 2.5, Aboushi, Fuller, and Skov by 2, and Nix by 1.5, we'd probably pick Elam, even though Nix is BPA. If, alternatively, Nix increases wins by 2.5, then he and Elam match, and we'd have to pick between BPA and need. IMO, in cases where players have an equal expected increase in wins, I would suggest going BPA, seeking the best talent.

Why rate BPA and needs at all then, you ask? If we're going by expected increase in wins, we just rate every player by that, and pick the player with the highest number when we pick.

The reason is that BPA can serve as a tie-breaker, as can needs. In the example, Aboushi, Fuller, and Skov would all increase wins by 2. Pretend Elam isn't there, and those were the best increases, then we would eliminate Skov, as ILB is lower on needs than OT and CB. Finally, to determine the difference between Aboushi and Fuller, we go back to the BPA board. Let's say Aboushi rates 5 spots higher than Fuller (46 v. 51). For that pick, Aboushi would be the best pick.

So how do we rate expected wins increase? Well, that involves math, more math, game tape, and gut feelings. And seeing as how I'm not paid to do that, I'll leave that to Shanahan and Allen.

[Gibbs Hog Heaven]

Taking this Redskin specific for a moment, as it's come up twice already; are we saying we see Neild or to a lesser extent Baker as quality long term depth at Nose, and not potential starting material?

Hail.

[PeterMP]

Draft based on need, but trade so that your need is at least close to the BPA at the position(s) you need.

[Skins Wingman]

we drafted cousins on a bpa basis and not need, and short term that was a bad call. who knows in the long run, but as of this year, it seems like a wasted pick.

[DogofWar1]

we drafted cousins on a bpa basis and not need, and short term that was a bad call. who knows in the long run, but as of this year, it seems like a wasted pick.

Hmm, I don't know. Backup QB was a need, otherwise we'd have Grossman/Beck backing up RGIII. In hindsight, we might instead have taken a safety or CB, but backup QB was important.

[manifested]

I think there are different levels of need to take into account and need factors more into higher picks (since the lower you go, the more of a boon it is just for the player to make the team). For example, FS is a glaring need on this team. On the other hand, DL has several above average players. But the level of this need matters so roster assessment has to play a big role. I also think it's less about drafting to fill a need rather than understanding your roster to eliminate areas which you would pick. I don't think the draft should be seen as filling immediate holes but as finding players to improve the positions on the team. The bigger the gap between your assessed potential of the player and your assessment of the current player available, the more likely you should pick at that position.

As it would apply to this roster, here's my strategy:

Young Elite Level Players = would not draft position highly
This level means drafting a replacement who would play at a higher level is unlikely and therefore not worth the resource investment. Mid-lower rounds possible in order to improve depth.
QB
LT

Above Average Starters: only draft highly if a potentially elite prospect is available, adds an element/role current starter doesn't possess, or you plan to rotate (Giants DL). Mid-lower rounds possible in order to improve depth.
RB
OLB
DE
FB

Solid players: Could upgrade so if you think player could reach above levels probably worth it. But if equal potential to a greater need choose the greater need at higher picks. May want to target in later rounds.
ILB (only bc Fletcher is getting old and we need to develop replacement)
LG
CB
C
WR
NT
CB (Wilson's spot)

Below Average: upgrade needed. If there's a player at a position above and a player here at about the same level of "potential" you take the player at the position here. Want to be on the lookout for mid-late round steals who could develop in a year or two to replace current starter (similar to the Perry Riley draft pick). Worth targeting at all rounds.
CB (DHall's spot)
RT
RG

Terrible: It might be worth reaching for a player with slightly lower ceiling. All things equal, absolutely, positively take the player at this position.
FS
SS

This doesn't take into account the importance of positions which is a whole other factor (like QB).

[stevemcqueen1]

- You rank every player into tiers.

- You draft the player who best fits what you're looking for on your roster in the highest tier.

- You never draft a player from a lower tier ahead of an available one from a higher tier.

For example, we've got a WR ranked as a tier 1 player available and we've got a safety ranked as a tier 2 player available. You always pick the WR, even though we have an immediate need at safety.

OR

Say we've got a tier 2 safety available and a tier 2 WR available. Well then, you go ahead and draft the safety because we've got an immediate need there and the players are in the same tier.

This is how most NFL teams do it.

Always keep in mind the draft is not meant for filling next year's holes. It's for building three years from now's foundation.

[DogofWar1]

Always keep in mind the draft is not meant for filling next year's holes. It's for building three years from now's foundation.

I still don't necessarily agree with this, I think the early picks are for filling immediate holes and the later ones are for long term foundation.

Griffin is the obvious one, he could lead us to a playoff bid at any time, just so long as our defense doesn't suck. Morris is another one. Watt is a good example from around the league, in his 2nd year he was so far along that they felt confident allowing Super Mario to walk, and so far they've been right. Whoever we take in the second, be it a safety, RT, NT, WR, or whatever, I expect them to be starting week one 2013.

You certainly do build for three years out, but I think you can just as easily use the draft to build for the current year, one year out, and two years out, at least on the early picks.

[Vilandil Tasardur]

- You rank every player into tiers.

- You draft the player who best fits what you're looking for on your roster in the highest tier.

- You never draft a player from a lower tier ahead of an available one from a higher tier.

For example, we've got a WR ranked as a tier 1 player available and we've got a safety ranked as a tier 2 player available. You always pick the WR, even though we have an immediate need at safety.

OR

Say we've got a tier 2 safety available and a tier 2 WR available. Well then, you go ahead and draft the safety because we've got an immediate need there and the players are in the same tier.

This is how most NFL teams do it.

Always keep in mind the draft is not meant for filling next year's holes. It's for building three years from now's foundation.

This times 1000.

The majority of your draft picks won't make significant contributions in that first year. By the time they're contributing in year 2 or 3, you may have an entirely different set of needs. You have to take the tier 1 guy ALWAYS, because you never know when some guy will tear an ACL, hold out, demand a trade, retire early, get into cocaine, or anything else that might necessitate depth.

Imagine if the Vikings had passed on Peterson because Chester Taylor had a big year before.

Some of you are saying that need matters most at the top of the draft but you go BPA later in the rounds. This is TOTALLY BACKWARDS. At the top of the draft, especially in the first round, you are picking FRANCHISE players that you expect to stick around for 5-10 years. This is where you take the best guy, period (except QB cause you never take a franchise QB if you have one.) I don't care if you have megatron on your team, you don't pass on a juggernaut wideout if you're drafting in the top 10 unless he is only marginally better than the next guys on the board.

The perfect example of this is us drafting Trent Williams. We missed out on the Ties AA talent (Suh and Mccoy) and had a bunch of ties A talent guys available. Of those, Berry was probably the best player, but he didn't grade out THAT much better than Trent, so we took Trent. But if by some miracle Suh had been there I don't doubt for a moment we would have taken him.

[Tennessee Ed]

I very much agree with your post. In essence, you said the same thing I did. The addition you made that I think is an absolutely phenomenal point is the follow through aspect. Follow through has to happen if you select based on BPNEED strategy. If Davis is the far superior player on your board to the point it trumps need, then you must utilize him appropriately or else you basically missed on the draft choice.

This.
Rookie salary cap might tip things in favor op bpa down the road.
Good thread, as always.

[mi6]

Hmm, I don't know. Backup QB was a need, otherwise we'd have Grossman/Beck backing up RGIII. In hindsight, we might instead have taken a safety or CB, but backup QB was important.

Agreed. Grossman and Beck were fodder. Beck was cut this year, and don't be surprised is Rexy is gone next year!