Risk, ceilings, floors & draft strategy

[mayoclinic]

This is and excellent thread. I'm at the stage now that I'm pretty confident that I know who the good player are. Now comes the hard part.

That's part of the problem. We're all pretty confident that we "know who the good players are", as are most GMs. And a lot of us will be wrong.

I've had pretty decent success picking some players over the past decade. I loved Troy Polamalu, Karlos Dansby, Lofa Tatupu, DeMarcus Ware, Steven Jackson, Ryan Clady, and Greg Jennings. But I also loved David Pollack (who might have been great but got injured and didn't pan out), Michael Huff, and EJ Henderson. And I never saw Michael Roos or Owen Daniels coming. I liked Ed Reed but never thought he would become the best FS of the past decade. Allan Branch was the closest thing I had to a binky in the 2007 draft, and I thought he could be a pro bowl 3-4 DE. Oops.

So I have my favorite players, and I've been vocal about them (Aaron Curry, Jason Smith, BJ Raji, Connor Barwin, William Beatty, Robert Ayers, Louis Delmas, Hakeem Nicks, Sean Smith, Alex Mack, Eric Wood, Jarron Gilbert, James Casey, Mike Wallace, Sammie Lee Hill, Devin Moore, Kevin Huber, and Jason Phillips among them, to pick 20 players), as well as about some of the players I don't care for. I've put a lot of work into thinking about "who the good players are". But I'm sure that I'll be wrong on at least 30% of them, and perhaps 50% or more.

 

[bucky]

So are picks like #23 & #34 important to get right because they're expensive, or because they're cheap?

Good question. Let me clarify my point then. Or perhaps even re-state it now with the benefit of afterthought. It's really a 2 part answer.

Part 1:

All draft picks, with the exception of top 10, are cheap compared to experienced starting NFL players. For example, Merriweather's contract was for about $9 mil over 5 years, compared to James Sanders $9 mil over 3 years. Obviously, the higher the draft pick, the more expensive they are. So whenever a draft pick goes bust, the higher he was drafted, the higher the associated cost.


Part 2:

As Mayoclinic points out, all draft picks carry risk. And it's very difficult to predict which ones will actually reach their ceiling (Watson and Maroney have been made examples in this thread). I would argue that it's a lot easier to gage a player's floor, than to predict where their ceiling is, or how close to their ceiling they're going to get.


Part 1 + Part 2 = Team Building Strategy:

- draft solid, unspectacular players who are able to become starters within 1-2 years.
- develop these players to see how much of their potential they're able to reach
- Identify key players who are developing to an elite level and can not be easily replaced. Apply the necessary $ to retaining these key players.
- Identify players who are not progressing much past their floor. When their rookie contracts expire, replace them through the draft with similar players who are younger and cheaper.
- Apply saved $ to acquire veterans who can fill holes and/or create depth on the roster.


Which in turn leads to the following Draft Strategy:

- Evaluate players mostly for their floor - how far is this player from being a starter/contributor to my team? What role are they capable of taking on and when?
- When it's your turn to draft, draft the player with the highest floor.
- When you don't see a player with a high enough floor (i.e. later rounds), proceed to drafting players with the highest ceiling
- Sometimes, a decision has to be made: draft a player with a high floor, but a limited role (i.e. Ghost or Matt Slater) vs. one with a high ceiling, but a major project (e.g. Guyton). I do not have a rule for this situation, but suffice it to say that you will never be 100% accurate in these situations. Hopefully, they occur in the later rounds and you can make a decision based on the state of your roster at the time.


That is my thesis for the day.

 

[PATSNUTme]

That's part of the problem. We're all pretty confident that we "know who the good players are", as are most GMs. And a lot of us will be wrong.

I've had pretty decent success picking some players over the past decade. I loved Troy Polamalu, Karlos Dansby, Lofa Tatupu, DeMarcus Ware, Steven Jackson, Ryan Clady, and Greg Jennings. But I also loved David Pollack (who might have been great but got injured and didn't pan out), Michael Huff, and EJ Henderson. And I never saw Michael Roos or Owen Daniels coming. I liked Ed Reed but never thought he would become the best FS of the past decade. Allan Branch was the closest thing I had to a binky in the 2007 draft, and I thought he could be a pro bowl 3-4 DE. Oops.

So I have my favorite players, and I've been vocal about them (Aaron Curry, Jason Smith, BJ Raji, Connor Barwin, William Beatty, Robert Ayers, Louis Delmas, Hakeem Nicks, Sean Smith, Alex Mack, Eric Wood, Jarron Gilbert, James Casey, Mike Wallace, Sammie Lee Hill, Devin Moore, Kevin Huber, and Jason Phillips among them, to pick 20 players), as well as about some of the players I don't care for. I've put a lot of work into thinking about "who the good players are". But I'm sure that I'll be wrong on at least 30% of them, and perhaps 50% or more.

OK let me rephase that: I'm pretty confident who the good players are suppose to be.

As I mentioned in another thread, Klemm was suppose to be better than Brady. The draft is about turning suspects into prospects. Not all prospects turn out to be the real thing.

 

[bakes781]

Obviously we're using this argument in the great "Connor Barwin vs. Name Your Own Personal Binky", but you can also use this with the top 2 QBs in the draft.

Stafford is considered by most the "safer" choice, but Sanchez projects to have the higher ceiling. Most every draftnik believes Stafford goes ahead of Sanchez and I believe alot of that has to do with exactly what we're talking about here.

 

[patchick]

It does help to have a monopoly on the 2nd round, but if you swing and miss with your 1st round pick that "value" in the 2nd can be all cancelled out IMO.

I'd usually agree, but part of my thinking came from a funny phenomenon I notice this year. Time after time I'll read a mock draft and cringe...until I mentally rearrange the order of the picks. E.g., the mock says something like:

23 - Alphonso Smith
34 - Brian Robiskie
47 - Connor Barwin
58 - Jarron Gilbert

And I'll imagine it as:

23 - Connor Barwin
34 - Jarron Gilbert
47 - Alphonso Smith
58 - Brian Robiskie

and then think "sign me up!"

IOW, there's a big clutch of talent there, are there's just not that huge a difference between #23 and 34, or between 34 and 47. The probably that #23 will turn out to be the best of the 4 day-1 picks isn't so great.

 

[bakes781]

That's why I think the best play would be to trade the #23 pick for next season, which by all accounts is potentially supposed to be a better draft from what I heard.

 

[bucky]

IOW, there's a big clutch of talent there, are there's just not that huge a difference between #23 and 34, or between 34 and 47. The probably that #23 will turn out to be the best of the 4 day-1 picks isn't so great.

Interesting that you say that. BB said on WEEI that (I'm paraphrasing) that there isn't much difference this year between 20 and 40. It's just that at 20 you can pick whichever player you want and at 40 you just take whichever one is left. Now, having just stated that as a fact, I will also say that he was given that information by the interviewer and basically just played off it. So I'm not sure whether that's his real opinion, or he was just playing along and being theoretical about it.

 

[mayoclinic]

That's why I think the best play would be to trade the #23 pick for next season, which by all accounts is potentially supposed to be a better draft from what I heard.

In 2006 I heard that 2007 was also going to be loaded. In 2007 I heard that the draft was weaker than expected but that 2008 was going to be terrific. In 2008 I heard that the draft was weak at the top, but that 2009 looked like a dynamite class. And now I hear that 2009 is deep but not lacking in impact players, but that 2010 will be much better. And I take it all with a grain of salt. What's more, no matter how weak the draft (2005 being probably the weakest draft of the decade) and where we pick, the FO seems to do a pretty decent job of finding prospects who fill needs and fit our system. Some years better than others, but still, overall a great job. I assume they will do the same this year, and in 2010.

 

[maverick4]

I've had pretty decent success picking some players over the past decade....But I also loved David Pollack (who might have been great but got injured and didn't pan out), Michael Huff, and EJ Henderson. And I never saw Michael Roos or Owen Daniels coming. I liked Ed Reed but never thought he would become the best FS of the past decade. Allan Branch was the closest thing I had to a binky in the 2007 draft, and I thought he could be a pro bowl 3-4 DE. Oops.
...But I'm sure that I'll be wrong on at least 30% of them, and perhaps 50% or more.

That's a little different from the picture you gave in the other thread, but 50% is still pretty good, exceptional even.

 

[mayoclinic]

That's a little different from the picture you gave in the other thread, but 50% is still pretty good, exceptional even.

Good lord, please don't think that I'm claiming clairvoyance. I'm just another schmuck trying to analyze players and schemes.

I think the draft is somewhat of a crapshoot. 198 teams passed on Tom Brady, 44 on Lofa Tatupu, 30 on Nnamdi Asomugha, 21 on Ed Reed, etc. I think a 50% hit rate on their draft prospects for any team would be quite good.

 

[Box_O_Rocks]

Which in turn leads to the following Draft Strategy:

- Evaluate players mostly for their floor - how far is this player from being a starter/contributor to my team? What role are they capable of taking on and when?
- When it's your turn to draft, draft the player with the highest floor.
- When you don't see a player with a high enough floor (i.e. later rounds), proceed to drafting players with the highest ceiling
- Sometimes, a decision has to be made: draft a player with a high floor, but a limited role (i.e. Ghost or Matt Slater) vs. one with a high ceiling, but a major project (e.g. Guyton). I do not have a rule for this situation, but suffice it to say that you will never be 100% accurate in these situations. Hopefully, they occur in the later rounds and you can make a decision based on the state of your roster at the time.

Nice thesis. :rocker: I view your thinking here to be more "tactical," seeking immediate results, as opposed to a "strategic" approach which makes some allowance for tactical needs, but applies a longer term thought process to the draft.

Tactically, it's better to draft a Clint Sintim at #23 because he would be more likely to contribute immediately. The question I have, does he "upgrade" the Linebacker corps?
-- When I look at past first round picks, the theme seems to involve the player in his entirety (floor & ceiling), plus an aspect of the team that can be "upgraded" by that player.
---- I can't say that Sintim is an improvement on projected starters Adalius Thomas and Pierre Woods, at best he's the equivalent of Woods. Which is why I no longer consider him a good first round value for NE, he doesn't upgrade anything.
-- Now, if he was intended to develop at SILB, you'd be looking at an arguable upgrade, but then his floor is much lower in the new position.
-- If you combine the two, high floor & equivalency at OLB, plus high ceiling & low floor at ILB, you mitigate the risk, making him more palatable at #23.
---- Yet there are later round options who bring higher floors to ILB and lower risk as late round prospects.

I submit you have the "floor" & "ceiling" perspective reversed. The higher ceiling should be part of your early round consideration and the higher floor should be more of a later round consideration. For example: Vince Wilfork drafted at #21 had a "solid" floor and a high ceiling, for him to have a high floor he'd have to have been playing more of a 2-gap defense at Miami, instead of their 1-gap system. Willie Andrews in the seventh round had a high floor for Special Teams, his ceiling was less defined. Bo Ruud has a high floor at 4-3 OLB and Special Teams, his ceiling is a little more defined with his athleticism and near prototype measurables, combined with excellent intangibles.

 

[mayoclinic]

Nice thesis. :rocker: I view your thinking here to be more "tactical," seeking immediate results, as opposed to a "strategic" approach which makes some allowance for tactical needs, but applies a longer term thought process to the draft.

I submit you have the "floor" & "ceiling" perspective reversed. The higher ceiling should be part of your early round consideration and the higher floor should be more of a later round consideration. For example: Vince Wilfork drafted at #21 had a "solid" floor and a high ceiling, for him to have a high floor he'd have to have been playing more of a 2-gap defense at Miami, instead of their 1-gap system. Willie Andrews in the seventh round had a high floor for Special Teams, his ceiling was less defined. Bo Ruud has a high floor at 4-3 OLB and Special Teams, his ceiling is a little more defined with his athleticism and near prototype measurables, combined with excellent intangibles.

I have to agree about about the higher ceiling being a major part of early round consideration. We want impact players out of the draft, not just solid low-risk role players. BB goes the FA route for those to plug holes at reasonable cost. The early rounds of the draft are for potential team leaders and impact players - Seymour, Warren, Wilfork, Mayo, Meriweather, Mankins, Watson, Maroney in the 1st, Light, Wilson and Jackson in the 2nd. They may not always hit their potential, but the potential is there. The first 6 are pro-bowl or potential pro-bowl caliber players, with Watson and Maroney clearly having similar talent even though they haven't put it together.

 

[patchick]

I submit you have the "floor" & "ceiling" perspective reversed. The higher ceiling should be part of your early round consideration and the higher floor should be more of a later round consideration. For example: Vince Wilfork drafted at #21 had a "solid" floor and a high ceiling, for him to have a high floor he'd have to have been playing more of a 2-gap defense at Miami, instead of their 1-gap system. Willie Andrews in the seventh round had a high floor for Special Teams, his ceiling was less defined. Bo Ruud has a high floor at 4-3 OLB and Special Teams, his ceiling is a little more defined with his athleticism and near prototype measurables, combined with excellent intangibles.

Interesting take. But even among the round 7/UDFA crowd, it seems to me the Pats have had their best successes with the low floor/high ceiling guys like Cassel, Woods, Redd, Gutierrez & Guyton who were loaded with physical potential and football/aptitude question marks, as opposed to the more polished but less athletic Lua, Ruud, Elgin, Roach, etc.

 

[mayoclinic]

Interesting take. But even among the round/UDFA crowd, it seems to me the Pats have had their best successes with the low floor/high ceiling guys like Cassel, Woods, Redd, Gutierrez & Guyton who were loaded with physical potential and football/aptitude question marks, as opposed to the more polished but less athletic Lua, Ruud, Elgin, Roach, etc.

Perhaps there is a "bimodal" distribution of low floor/high ceiling guys going in the first 2 rounds or very late in the draft, with the middle rounds going for less physically gifted but solid players like Dan Koppen, James Sanders, Ellis Hobbs, Asante Samuel, Gerald Wilhite, etc.

 

[maverick4]

Strategy-wise, it depends on the franchise. Al Davis loves using high value picks on low floor, high ceiling guys, especially those with speed. Belichick's recent Day 1 picks on high ceiling but raw players haven't panned out.

We have had success spending low value picks on players with low floors and high ceilings. The high value picks go on players with ideally high floors and high ceilings. It is up to a franchise's style whether for mid-rounds they value higher floors or ceilings.

 

[patchick]

Perhaps there is a "bimodal" distribution of low floor/high ceiling guys going in the first 2 rounds or very late in the draft, with the middle rounds going for less physically gifted but solid players like Dan Koppen, James Sanders, Ellis Hobbs, Asante Samuel, Gerald Wilhite, etc.

OK, to geek out fully...so maybe something like this, where the x-axis is selection round, the pink line represents the minimum acceptable ceiling and the y-axis the minimum acceptable floor?

 

[Box_O_Rocks]

Interesting take. But even among the round 7/UDFA crowd, it seems to me the Pats have had their best successes with the low floor/high ceiling guys like Cassel, Woods, Redd, Gutierrez & Guyton who were loaded with physical potential and football/aptitude question marks, as opposed to the more polished but less athletic Lua, Ruud, Elgin, Roach, etc.

We might disagree a little here...
- Cassel's floor appeared low because he hadn't played, but he clearly had the fundamentals and tools or he'd never have been of interest. You can argue that his ceiling was high, but again I think that's more a matter of 20/20 hindsight. At the time he was drafted he was a physical prototype NFL QB with a solid, if unexceptional floor and the intangibles to warrant a 7th round flyer. I don't think you could even make a case for his ceiling being high enough to project him as a #3 QB. Certainly as recently as last September there were very, very few people who felt he could be a servicable NFL reserve, let alone a skilled performer. Happily most of those few were Patriots' coaches.
- The four UDFA you list wouldn't even be worth bringing into camp if they had low floors, it was their ceilings that were hazy and the only way to determine if they had any more potential was to bring them into camp.
- Lua to me was a low floor, he had played decently in limited reps between Tatupu's departure and Maualuga's assension, he was limited athletically as you noted, certainly once he got into camp he did nothing to encourage any hopes I might have had for him. I don't know how you could project a high ceiling for him.
- Ruud I've discussed.
- Elgin had a good solid floor and good athleticism, but as an undersized OC you couldn't project his ceiling to be very high. His technique helped raise his ceiling a smidge, but we pretty much had him pegged for the Practice Squad and further development.
- Roach I'm drawing a blank.

I don't see anything in these late round/UDFA kids to say, "wow, he's got a ton of potential." Heck, that Brady kid they brought in was in camp because he had a high floor, but he looked like he was bumping his head on the ceiling.

 

[mayoclinic]

OK, to geek out fully...so maybe something like this, where the x-axis is selection round, the pink line represents the minimum acceptable ceiling and the y-axis the minimum acceptable floor?

We can call it the "Patchick-Mayo distribution" (or P-M for short). Pretty cool, though according to your graph the ceiling is about as high in the 7th round as it in in the 1st. I hope you're right.

 

[WhiZa]

After reading this thread, and desiring to take an OLB at 23, now I'm leaning toward trading out!

 

[maverick4]

OK, to geek out fully...so maybe something like this, where the x-axis is selection round, the pink line represents the minimum acceptable ceiling and the y-axis the minimum acceptable floor?

I think after rd3 of your chart, it is up to team preference/style.