Theory: Trent Baalke's Draft Strategy is not traditional BPA

Great article.

MD has a valid point in thinking that value drafting can lead to good results -- IF you are a good talent evaluator.

If your drafting BPA, you have nearly the entire group of available players to choose from. If your only drafting for positions of need, you have a much smaller subset of those players available to you. Therefore, if you need to be a good talent evaluator to succeed using the BPA strategy, you need to be f**king amazing to succeed by going after need positions. You simply have less players to choose from.

All else being equal, the same GM will build a roster with more talent by going pure BPA, all else being equal.

The 49ers have chose option #3, which is to have a GM who simply does no wrong.

I think he just doesn't' draft who the so called "experts" say they think is the top guy. I know that mcshay and kipper spend a lot of time analyzing video tape...but, we have to remember these guys didn't play football...at least not at a high level...so what exactly makes them qualified to project anybody's talents or say who drafted well or who didn't? It's a nice show, but Baalke I think is changing the NFL. The niners are the leading team right now, with defense, draft strategy, and team unity. Everyone is wanting to be like us, just like the 80's. And we've only been good for one year...lets see how this goes.

Baalke's board assigns a value of the talent, per round. If the value isn't there in the round they are drafting in, Baalke will parlay that pick into next year's draft or trade down for talent that presents an appropriate value for the round. It's a simple, yet effective strategy. A lot better than picking a player just because it's your turn, or drafting for need. I wonder if Baalke will ever fall in love with a prospect that he deems worthy of trading up for in the 1st round.

Best Fit Available

Really wanted Cordy Glenn and Jared Crick. But Baalke deserves the benefit of the doubt after last year.

Baalke's stradegy still comes down to BPA, it's just through his eyes alone. In other words, talent is all in the eye of the beholder. Sure, he rates his board a little differently then other teams, based on attributes and certain elements but to Baalke it's still who he believes is the best player for the 49ers. Interesting way to evaluate his style though, reminds me a little of how Walsh approached the draft. Good read...

Thanks

Thoughtful, measured, analysis with intelligent responses in in Niner Talk?! I'm shocked and impressed

In all seriousness, though, I think this is a wonderful post. I'd like to introduce something to this theory that other posters mentioned, but that I think is a key part of the drafting strategy: scheme.

We can glean a lot from the Harbaugh/Roman offense from the evaluation of Andrew Luck. Greg Cosell watched a ton of film, and is very respected in the field of football and player analysis. Here's what he had to say about Luck:

Greg Cosell

The deeper throws were what we call shot plays, primarily with play action, specifically designed to attack an anticipated coverage based on field position, down and distance, personnel and formation. On those plays, the receiver was wide open. They were not difficult passes. Overall, Luck was not asked to make many tough throws at the intermediate and deeper levels. I did not see those.

Why is that important? Because that is exactly the kind of offense we run here. Short and intermediate throws, with calculated shots down the field. We have our short and intermediate receivers: Manningham and Crabtree. What we need is that player to stretch the field and catch the ball (ahem, Stephen Hill) when we take those shots.

Schematically we value certain attributes. Wide receiver is a great example of how not every player is created equal at this position. Wes Welker and Mike Wallace are both amazing wide receivers, but each brings something different to their team. We're not looking for the best "player at the position," we're identifying particular traits at positions of need.

This is a minor distinction, but an important one. I agree with OnTheClock that value is paramount when drafting. And you can ensure value by tapping undervalued resources. For example, a tweener DE in college used to go un-drafted in the NFL. The rise of the 3-4 defense meant that a player no one valued, when placed in a different scheme, produced excellent results. Closer to home, Bob McKittrick always coached up smaller more agile offensive lineman. He took another teams trash and turned them into some of the best offensive lines the NFL has seen.

If everyone is going after the same traits, then everyone goes after the same players. When you know your team identity (another word for your scheme) you can search within a draft for players who have the traits you need. And I think THIS is how the best teams draft. And I also think that, thanks to Baalke and Harbaugh, we are one of those "best teams."

Originally posted by Oscar:

Schematically we value certain attributes. Wide receiver is a great example of how not every player is created equal at this position. Wes Welker and Mike Wallace are both amazing wide receivers, but each brings something different to their team. We're not looking for the best "player at the position," we're identifying particular traits at positions of need.

Great post! Having that sense of identity gives the GM direction during the draft and helps him hone in on those players that would fit and be most coachable in the system. That's why Baalke likes to look for guys who are downgraded because they came from small schools, are character risks or have an injury history. He doesn't gamble without having other alternatives though.

They picked up Miller and moved him to FB...and I still have not read a good article on just how Baalke figured that one out. Did he really trust the HS coach's assessment? Wasn't a high pick so maybe.

The niners pick up a CB FA with questionable character because they made sure they have three other guys capable of starting.

Baalke can make these picks because he knows what kind of player the team needs and what skills that player has to have, as opposed to the qualities that would be nice to have. I think James is a good example of a guy who was just too much to pass on...has the necessary tools but also the elite set. Harbaugh knows he can get the short yards they could not get last year. He might have been a pure BPA that point...not a position of need but a needed skill.

"Thoughtful, measured, analysis with intelligent responses"

I guess the kids are still watching Saturday morning cartoons... Give it time.

Good post!!! And like the name, "mebemused" The Draft, is it Science or Art? Or is it how we are able to blend the Science and Art together?
I thinking that all of us back seat draft fans only get to see a small peek hole of the whole picture. My hats off to the great job that Harbaugh and Baakle Are doing. I am very happy with how both Harbaugh and Baakle are blending Science and Art into their work. A+ for each. I for one am very pleased with this off season, and everything the 49ers have done since the day Baakle was able to hire Harbaugh. And my hats off also to to all the hard work guys like OTC and Maddog do for us fair-weather back seat draft fans each and every year. Mahalo!

Originally posted by mebemused:

OTC, how do you explain the LaMichael James pick in terms of value, in this case, Baakle's valuing speed. James is almost a duplicate of Kendall Hunter. Neither of your two theories explains that pick. The one that best explains LaMichael James' pick is pure value, BPA regardless of need.

In contrast, your own example of a BPA draft, you take Gordy Glenn in round one. Clearly, a BPA at a position of need.

In the second, you take a wonderful football player, Ronnell Lewis, who fell to the 4th round I believe because he is a tweener, All the teams recognized the talent and the desire in Lewis, but apparently they just didn't know where to play him. That would be 32 GMs telling you and all the public value boards that you overvalued Lewis.

I love all your picks, cause in each case, they are players that were rated higher pre-draft and they fell on draft day. Based on a pre-draft, consensus/averaging estimate of value, your draft is bountiful indeed. And really, what other set of criteria would we have as fans? We don't watch Kilgore in practice. We don't know how fast he is developing. But we can watch all this season's college players and form opinions on who is our favorites. And we know that we have a hole at RG that we have to fill in the draft cause we haven't signed a FA.

But your big value draft begs the question, why did all these players fall not only past Baalke, but all the other GMs as well? And the only answer we can give is to infer it was because they weren't as high on 32 team value boards as they were for Kiper, Mayock, Rang, OTC, or MD's public value boards.

Last year, you could have made the same argument about Aldon Smith and Chris Culliver, that they were drafted before their consensus ranking, hence were reaches. But now in hindsight, we marvel at how well those two selections turned out.

And there is the rub. After all that investigative work, 250+ player resumes, watching the combine, risers and fallers, Baalke and Harbaugh didn't do what we would have done based on consensus value boards, or our own variants. Now we have to sit back and watch how the players we took, and the players we could have taken perform next season, and in all fairness, the season after that.

Beauty is in the eye of the beholder. What you are advocating is, all things being equal, that consensus valuing is more reliable than Baalke's, or anybody else's, idiosyncratic valuing. And statistically, that would be correct. Yet the talent raters that we most value, are the guys who see beauty where no one else sees beauty. Baalke has made a huge gamble by going against consensus value boards. Last year's gambles and his free agent moves, brought us to within a fumble of the SB, and garned him Best Exec. All I can do, is hope he knows something that I don't know, which is definitely true, so it's not much of a reach.

Lastly, I want to thank MadDog and you for your herculean effort in covering this draft and many previous drafts. You two educate me each year on who are good players, allowing me to pretend I am the GM. Great fun. Many hours of entertaining debate on the zone. We all owe both of you props. Till next year!

Originally posted by mebemused:

OTC, how do you explain the LaMichael James pick in terms of value, in this case, Baakle's valuing speed. James is almost a duplicate of Kendall Hunter. Neither of your two theories explains that pick. The one that best explains LaMichael James' pick is pure value, BPA regardless of need.

In contrast, your own example of a BPA draft, you take Gordy Glenn in round one. Clearly, a BPA at a position of need.

In the second, you take a wonderful football player, Ronnell Lewis, who fell to the 4th round I believe because he is a tweener, All the teams recognized the talent and the desire in Lewis, but apparently they just didn't know where to play him. That would be 32 GMs telling you and all the public value boards that you overvalued Lewis.

I love all your picks, cause in each case, they are players that were rated higher pre-draft and they fell on draft day. Based on a pre-draft, consensus/averaging estimate of value, your draft is bountiful indeed. And really, what other set of criteria would we have as fans? We don't watch Kilgore in practice. We don't know how fast he is developing. But we can watch all this season's college players and form opinions on who is our favorites. And we know that we have a hole at RG that we have to fill in the draft cause we haven't signed a FA.

But your big value draft begs the question, why did all these players fall not only past Baalke, but all the other GMs as well? And the only answer we can give is to infer it was because they weren't as high on 32 team value boards as they were for Kiper, Mayock, Rang, OTC, or MD's public value boards.

Last year, you could have made the same argument about Aldon Smith and Chris Culliver, that they were drafted before their consensus ranking, hence were reaches. But now in hindsight, we marvel at how well those two selections turned out.

And there is the rub. After all that investigative work, 250+ player resumes, watching the combine, risers and fallers, Baalke and Harbaugh didn't do what we would have done based on consensus value boards, or our own variants. Now we have to sit back and watch how the players we took, and the players we could have taken perform next season, and in all fairness, the season after that.

Beauty is in the eye of the beholder. What you are advocating is, all things being equal, that consensus valuing is more reliable than Baalke's, or anybody else's, idiosyncratic valuing. And statistically, that would be correct. Yet the talent raters that we most value, are the guys who see beauty where no one else sees beauty. Baalke has made a huge gamble by going against consensus value boards. Last year's gambles and his free agent moves, brought us to within a fumble of the SB, and garned him Best Exec. All I can do, is hope he knows something that I don't know, which is definitely true, so it's not much of a reach.

Lastly, I want to thank MadDog and you for your herculean effort in covering this draft and many previous drafts. You two educate me each year on who are good players, allowing me to pretend I am the GM. Great fun. Many hours of entertaining debate on the zone. We all owe both of you props. Till next year!

Originally posted by mebemused:

OTC, how do you explain the LaMichael James pick in terms of value, in this case, Baakle's valuing speed. James is almost a duplicate of Kendall Hunter. Neither of your two theories explains that pick. The one that best explains LaMichael James' pick is pure value, BPA regardless of need.

In contrast, your own example of a BPA draft, you take Gordy Glenn in round one. Clearly, a BPA at a position of need.

In the second, you take a wonderful football player, Ronnell Lewis, who fell to the 4th round I believe because he is a tweener, All the teams recognized the talent and the desire in Lewis, but apparently they just didn't know where to play him. That would be 32 GMs telling you and all the public value boards that you overvalued Lewis.

I love all your picks, cause in each case, they are players that were rated higher pre-draft and they fell on draft day. Based on a pre-draft, consensus/averaging estimate of value, your draft is bountiful indeed. And really, what other set of criteria would we have as fans? We don't watch Kilgore in practice. We don't know how fast he is developing. But we can watch all this season's college players and form opinions on who is our favorites. And we know that we have a hole at RG that we have to fill in the draft cause we haven't signed a FA.

But your big value draft begs the question, why did all these players fall not only past Baalke, but all the other GMs as well? And the only answer we can give is to infer it was because they weren't as high on 32 team value boards as they were for Kiper, Mayock, Rang, OTC, or MD's public value boards.

Last year, you could have made the same argument about Aldon Smith and Chris Culliver, that they were drafted before their consensus ranking, hence were reaches. But now in hindsight, we marvel at how well those two selections turned out.

And there is the rub. After all that investigative work, 250+ player resumes, watching the combine, risers and fallers, Baalke and Harbaugh didn't do what we would have done based on consensus value boards, or our own variants. Now we have to sit back and watch how the players we took, and the players we could have taken perform next season, and in all fairness, the season after that.

Beauty is in the eye of the beholder. What you are advocating is, all things being equal, that consensus valuing is more reliable than Baalke's, or anybody else's, idiosyncratic valuing. And statistically, that would be correct. Yet the talent raters that we most value, are the guys who see beauty where no one else sees beauty. Baalke has made a huge gamble by going against consensus value boards. Last year's gambles and his free agent moves, brought us to within a fumble of the SB, and garned him Best Exec. All I can do, is hope he knows something that I don't know, which is definitely true, so it's not much of a reach.

Lastly, I want to thank MadDog and you for your herculean effort in covering this draft and many previous drafts. You two educate me each year on who are good players, allowing me to pretend I am the GM. Great fun. Many hours of entertaining debate on the zone. We all owe both of you props. Till next year!

Originally posted by mebemused:

OTC, how do you explain the LaMichael James pick in terms of value, in this case, Baakle's valuing speed. James is almost a duplicate of Kendall Hunter. Neither of your two theories explains that pick. The one that best explains LaMichael James' pick is pure value, BPA regardless of need.

In contrast, your own example of a BPA draft, you take Gordy Glenn in round one. Clearly, a BPA at a position of need.

In the second, you take a wonderful football player, Ronnell Lewis, who fell to the 4th round I believe because he is a tweener, All the teams recognized the talent and the desire in Lewis, but apparently they just didn't know where to play him. That would be 32 GMs telling you and all the public value boards that you overvalued Lewis.

I love all your picks, cause in each case, they are players that were rated higher pre-draft and they fell on draft day. Based on a pre-draft, consensus/averaging estimate of value, your draft is bountiful indeed. And really, what other set of criteria would we have as fans? We don't watch Kilgore in practice. We don't know how fast he is developing. But we can watch all this season's college players and form opinions on who is our favorites. And we know that we have a hole at RG that we have to fill in the draft cause we haven't signed a FA.

But your big value draft begs the question, why did all these players fall not only past Baalke, but all the other GMs as well? And the only answer we can give is to infer it was because they weren't as high on 32 team value boards as they were for Kiper, Mayock, Rang, OTC, or MD's public value boards.

Last year, you could have made the same argument about Aldon Smith and Chris Culliver, that they were drafted before their consensus ranking, hence were reaches. But now in hindsight, we marvel at how well those two selections turned out.

And there is the rub. After all that investigative work, 250+ player resumes, watching the combine, risers and fallers, Baalke and Harbaugh didn't do what we would have done based on consensus value boards, or our own variants. Now we have to sit back and watch how the players we took, and the players we could have taken perform next season, and in all fairness, the season after that.

Beauty is in the eye of the beholder. What you are advocating is, all things being equal, that consensus valuing is more reliable than Baalke's, or anybody else's, idiosyncratic valuing. And statistically, that would be correct. Yet the talent raters that we most value, are the guys who see beauty where no one else sees beauty. Baalke has made a huge gamble by going against consensus value boards. Last year's gambles and his free agent moves, brought us to within a fumble of the SB, and garned him Best Exec. All I can do, is hope he knows something that I don't know, which is definitely true, so it's not much of a reach.

Lastly, I want to thank MadDog and you for your herculean effort in covering this draft and many previous drafts. You two educate me each year on who are good players, allowing me to pretend I am the GM. Great fun. Many hours of entertaining debate on the zone. We all owe both of you props. Till next year!

Originally posted by mebemused:

OTC, how do you explain the LaMichael James pick in terms of value, in this case, Baakle's valuing speed. James is almost a duplicate of Kendall Hunter. Neither of your two theories explains that pick. The one that best explains LaMichael James' pick is pure value, BPA regardless of need.

In contrast, your own example of a BPA draft, you take Gordy Glenn in round one. Clearly, a BPA at a position of need.

In the second, you take a wonderful football player, Ronnell Lewis, who fell to the 4th round I believe because he is a tweener, All the teams recognized the talent and the desire in Lewis, but apparently they just didn't know where to play him. That would be 32 GMs telling you and all the public value boards that you overvalued Lewis.

I love all your picks, cause in each case, they are players that were rated higher pre-draft and they fell on draft day. Based on a pre-draft, consensus/averaging estimate of value, your draft is bountiful indeed. And really, what other set of criteria would we have as fans? We don't watch Kilgore in practice. We don't know how fast he is developing. But we can watch all this season's college players and form opinions on who is our favorites. And we know that we have a hole at RG that we have to fill in the draft cause we haven't signed a FA.

But your big value draft begs the question, why did all these players fall not only past Baalke, but all the other GMs as well? And the only answer we can give is to infer it was because they weren't as high on 32 team value boards as they were for Kiper, Mayock, Rang, OTC, or MD's public value boards.

Last year, you could have made the same argument about Aldon Smith and Chris Culliver, that they were drafted before their consensus ranking, hence were reaches. But now in hindsight, we marvel at how well those two selections turned out.

And there is the rub. After all that investigative work, 250+ player resumes, watching the combine, risers and fallers, Baalke and Harbaugh didn't do what we would have done based on consensus value boards, or our own variants. Now we have to sit back and watch how the players we took, and the players we could have taken perform next season, and in all fairness, the season after that.

Beauty is in the eye of the beholder. What you are advocating is, all things being equal, that consensus valuing is more reliable than Baalke's, or anybody else's, idiosyncratic valuing. And statistically, that would be correct. Yet the talent raters that we most value, are the guys who see beauty where no one else sees beauty. Baalke has made a huge gamble by going against consensus value boards. Last year's gambles and his free agent moves, brought us to within a fumble of the SB, and garned him Best Exec. All I can do, is hope he knows something that I don't know, which is definitely true, so it's not much of a reach.

Lastly, I want to thank MadDog and you for your herculean effort in covering this draft and many previous drafts. You two educate me each year on who are good players, allowing me to pretend I am the GM. Great fun. Many hours of entertaining debate on the zone. We all owe both of you props. Till next year!

Originally posted by BrianGO:

Bill Walsh used to talk about how he would draft players who "fit" on the football team well. "What specific things can this player do?" Dwight Clark was not a great athlete, but Walsh fit him in by looking at what he could do well, but physically and mentally.

Another thing is, once a team has lots of talent, you have to become more cognizant of the fact that you want your draft picks to have a chance at making the team better, not just providing a slight upgrade by beating out another good player.

Most importantly though, they liked the players they took.

I recall in one of his books saying that they really liked "player A" but he had so few positive plays his sr season with in 5 yards of the line of scrimmage they took him off thier board altogether

So yea they went from love to don't touch even though the guy has awesome mesurables and fantastic college production

Originally posted by Oscar:

Thoughtful, measured, analysis with intelligent responses in in Niner Talk?! I'm shocked and impressed

In all seriousness, though, I think this is a wonderful post. I'd like to introduce something to this theory that other posters mentioned, but that I think is a key part of the drafting strategy: scheme.

We can glean a lot from the Harbaugh/Roman offense from the evaluation of Andrew Luck. Greg Cosell watched a ton of film, and is very respected in the field of football and player analysis. Here's what he had to say about Luck:

Quote:

Greg Cosell

The deeper throws were what we call shot plays, primarily with play action, specifically designed to attack an anticipated coverage based on field position, down and distance, personnel and formation. On those plays, the receiver was wide open. They were not difficult passes. Overall, Luck was not asked to make many tough throws at the intermediate and deeper levels. I did not see those.

Why is that important? Because that is exactly the kind of offense we run here. Short and intermediate throws, with calculated shots down the field. We have our short and intermediate receivers: Manningham and Crabtree. What we need is that player to stretch the field and catch the ball (ahem, Stephen Hill) when we take those shots.

Schematically we value certain attributes. Wide receiver is a great example of how not every player is created equal at this position. Wes Welker and Mike Wallace are both amazing wide receivers, but each brings something different to their team. We're not looking for the best "player at the position," we're identifying particular traits at positions of need.

This is a minor distinction, but an important one. I agree with OnTheClock that value is paramount when drafting. And you can ensure value by tapping undervalued resources. For example, a tweener DE in college used to go un-drafted in the NFL. The rise of the 3-4 defense meant that a player no one valued, when placed in a different scheme, produced excellent results. Closer to home, Bob McKittrick always coached up smaller more agile offensive lineman. He took another teams trash and turned them into some of the best offensive lines the NFL has seen.

If everyone is going after the same traits, then everyone goes after the same players. When you know your team identity (another word for your scheme) you can search within a draft for players who have the traits you need. And I think THIS is how the best teams draft. And I also think that, thanks to Baalke and Harbaugh, we are one of those "best teams."

I think you just about EXACTLY summarized and put into other words what I am getting at here. Bravo. That is excellent.

Instead of elements, you use the word "traits" and I think that's a great synonym for what I'm referring to in this discussion. Our drafts look very, very different from the rest of the NFL because we implement this trait-specific board stacking technique. That is not traditional BPA and isn't really even similar to it. Traditional BPA weighs overall positional needs and takes the best player of the entire pool, as long as the value of a non-need player does not exceed the weighted value of a player at a needed position, since the player at a needed position would get at least a slight boost in "attributed value" because of the fact that they are a position of need.

Our strategy is much more complex than traditional BPA. We know our positional needs, but we target specific players with specific traits in those positions, which decreases the player pool, and that is exemplified in Dshearn's post below.

Originally posted by Dshearn:

I recall in one of his books saying that they really liked "player A" but he had so few positive plays his sr season with in 5 yards of the line of scrimmage they took him off thier board altogether

So yea they went from love to don't touch even though the guy has awesome mesurables and fantastic college production

The draft strategy we employ typically will have inconsistent yields across entire drafts in terms of player success. More often then not, almost all players contribute, or almost none do from a given draft. But while riskier, however, it has the potential to create a lasting dominant roster -- a dynasty, if you will -- if done correctly . While traditional BPA can be that way too once in a while, it typically is a smoother curve or line of success. Teams like the Colts will experience long-term moderate success, and may even win a Super Bowl, but their overall team will mostly be in that "almost dominant but not quite" category.

Now that I see what Trent is doing, I don't mind his strategy so much. However, he's going to have to establish that he can get the players right more often than not, because this strategy has greater risk of creating brief stretches of sub-par to very bad drafts.

As long as we can mold the players into what we expect to be able to make them into, we should experience long term success

I think Balke just defines "need" differently. Instead of saying "i need an OLB", he says "i need to improve my pass rush". Instead of saying "i need a WR", he says "i need more speed to stretch the field".

So its BPA to fill a need.... like what everybody else does. He is just smarter about identifying the needs.

Originally posted by blm7754:

I think Balke just defines "need" differently. Instead of saying "i need an OLB", he says "i need to improve my pass rush". Instead of saying "i need a WR", he says "i need more speed to stretch the field".

So its BPA to fill a need.... like what everybody else does. He is just smarter about identifying the needs.

This seems the most concise explanation of a very complicated process.