You can only get a good QB in the top 3

[Pine Barrens Mafia]

Keep running your crusade. Mark my words: 4 QB will be gone in the first 10 picks.

 

I don't want Lamar Jackson. Ergo, I want to move up.

 

[Jay_Fixit]

56 minutes ago, joesixpack said:

Keep running your crusade. Mark my words: 4 QB will be gone in the first 10 picks.

 

I don't want Lamar Jackson. Ergo, I want to move up.

 

Your words are marked. And they’ll be incorrect.

[Pine Barrens Mafia]

Just now, Jay_Fixit said:

Your words are marked. And they’ll be incorrect.

 

Cleveland

Jets

Donkos

Bills

 

will all pick QB in the top 10.

 

So go ahead and mark them.

 

[mjt328]

3 hours ago, SoTier said:

That's all fine and dandy but drafting JP Losman or EJ Manuel at #1 wouldn't have made either significantly better (although Losman might have been better playing for somebody other than Jauron).    Being drafted #1 overall didn't prevent David Carr or JaMarcus Russell from busting.  Draft position isn't what determines a QB's success or failure.

 

Of course not.  And nobody is saying "draft order" magically makes players better or worse.

 

For the most part, NFL scouts and GMs are in pretty close agreement about which players have the best chance of success in the pros.  There are obviously some variations between teams.  One team may place a premium on arm strength, while another is more interested in accuracy.  Another may disqualify a player for off-field concerns, while another is willing to take the risk.  Doctors may have different opinions about injury history.  But if you could take a peek at all 32 draft boards, I think you would find them remarkably similar.

 

When people post stats/historical data about "draft order" - what they are really trying to say is that higher ranked prospects (based on consensus) have a MUCH higher track record of success than lower ranked prospects.  It's just simple math.  Yes, there are always going to be Tom Bradys that nobody sees coming. Yes, there are always going to be Jamarcus Russells and Ryan Leafs that are raved about as prospects, then bust horribly.  But the numbers don't lie.  Your odds are better taking your 1st or 2nd ranked guy, as opposed to the 4th, 5th, etc.

 

If you could take a look at every team's 2018 QB draft board, I would fully expect to see Sam Darnold and Josh Rosen within EVERY team's Top 2-5.  Depending on how those teams rank things like arm strength, accuracy, height, athletic ability, etc., I would fully expect Josh Allen, Baker Mayfield and Lamar Jackson sprinkled into those other spots.  Of course, someone might like Allen or Mayfield as their #1.  Others may have them at their #4-5.  But I highly doubt (for example) that anyone will have Darnold ranked as a 4th Round prospect, or someone like Kyle Lauletta as worthy of a Top 5 pick.

 

At the end of the day, if 32/32 scouts believe that Josh Rosen is going to have a better chance at success than Mason Rudolph... there is a higher probability they are going to be right.

 

[maryland-bills-fan]

1 hour ago, Hapless Bills Fan said:

 

The Buffalo Rumblings article is a good one, but the estimate/guestimate using 100/% chance is just incorrect.

First of all, in probability, you never get to 100% so 100/% chance - I have no idea where that comes from, I consulted the "better half" who is math+++ to my math+ - sorry, we think you're just wrong

 

The way to look at it is the odds of failure from repeated independent events.  If you have 64% odds of success, (1-.64) x the number of independent throws is your probability that you fail every time.  Then 1 - (probability of repeated failure) is your odds of success.

 

If you want 95% odds of success, the actual formula is log (0.05)/log (1-odds of success on each throw) = # of throws.

 

Got to run, later.

I disagree.  Look at the student T test. We have a small sample size, a murky test criteria (what is the deviation from the mean) and to talk about a 100% confidence level that there is a statistical difference between the means is just silly.    I'm reminded of Mark Twain. " there are liars, there are damn liars and there are statistics".  Ask questions about when the sample size is 2 to 10. Opps.
 

Edited March 15, 2018 by maryland-bills-fan

[Hapless Bills Fan]

3 hours ago, maryland-bills-fan said:

I disagree.  Look at the student T test. We have a small sample size, a murky test criteria (what is the deviation from the mean) and to talk about a 100% confidence level that there is a statistical difference between the means is just silly.    I'm reminded of Mark Twain. " there are liars, there are damn liars and there are statistics".  Ask questions about when the sample size is 2 to 10. Opps.
 

 

DUDE!  This has nothing to do with small sample size, murky test criteria, or anything else. What you did is simply mathematically incorrect.

Now I make mistakes, but when my mistake is pointed out, I acknowledge it, I don't start obfuscating about 'statistical difference between the means' and '100% confidence levels' and toss in a Mark Twain quote about liars and statistics, none of which are relevant here and the latter of which is pretty damn insulting.

 

This isn't a quibble; correct math leads one to EXACTLY THE OPPOSITE CONCLUSION as your incorrect statistical calculation

 

To recap, this is what you did:

From Maryland-Bills-Fan post:

working from the OP data you can estimate/guesttimate the following

How many times to you have to draft a QB in those spots to get a GOLD QB? (= 100 /  %chance)
top 3  ..............1.56 picks
rest of 1st ......3.03 picks
2n3 & 3rd ........4.54 picks

 

To calculate the number of picks you need, you are using a value of 100%.  Nothing in statistics reaches a point of certainty, so that's wrong up front.

Then to calculate the number of picks needed for certainty, you are dividing 100% by the % chance of GOLD at different draft positions.

This is simply incorrect statistical calculation.  I'm frankly unclear where you even got this.

 

Here is how it should be done, using your posted values:

top 3: 64% "Gold".   Therefore probability of not-Gold in any one draft choice: (1-0.64) = 0.36

let's say you draft 3x in the top 3 picks.  Your chance of "not-Gold" in all 3 picks is: (0.36)x(0.36)x(0.36) = 0.05. 

This means you have (1 - 0.05) = 95% chance of success at getting a GOLD qb if you draft in the first 3 picks, 3 times.  Yea!  Success!  But 3 picks - not "1.56 picks.

 

(0.36)x(0.36)x(0.36) is often written (0.36)^3, so the equation is: (0.36)^3 = 0.05

The general equation will be (1 - %chance GOLD)^n = 0.05, where "n" is the number of picks needed to achieve a 95% chance of success.

To simplify, you take log of both sides, use the identity log (x^n) = n log (x), then solve for n to get:

n = log (0.05)/log (1 - % chance GOLD).

 

The correct values for 95% chance of success are:

top 3 (64% GOLD)  ..............3 picks
rest of 1st (33% GOLD)......7.5 picks (round up to 8 picks)
2nd & 3rd (22% GOLD)........12 picks

 

So we would need (theoretically) to pick a QB in the 2nd and 3rd round 12 times, to achieve the same odds of success as drafting 3 QB in the top 3 picks of the first.

 

When calculated correctly, it actually makes sense to "burn draft capital" and trade up to the top 3 picks vs. taking a shot at a late 1st or 2nd and 3rd round guy.

3 years of trying (assuming there's a QB we like that high each year) vs 12 years of trying?   Each guy needs a couple years to evaluate/develop, that's the difference between as many 6-9 yrs vs 24-36 years for a "GOLD" QB.

 

Hell to the Yeah, Trade Up

Pox and Pestilence to the years of GMs who put us in this abysmal hole by failing to either trade up or pull the trigger often enough at the lower rounds in the last 25 years.

 

And not incidentally, I reiterated this just in case there are any impressionable young minds reading this: we don't want them to fail AP Statistics

Or even the statistics my kid learnt in 6th grade.

 

And oh yeah: significant figures and common sense, too - 3.03 picks? 4.54 picks?  C'mon Man. 

 

 

 

 

 

Edited March 15, 2018 by Hapless Bills Fan

[Hapless Bills Fan]

13 hours ago, maryland-bills-fan said:

I am thinking that all the shrill demands to move up to the top 3 are panic driven.

 

More like data driven.  The chance of a good QB in the top 2 picks is significantly better than the chance lower down.

 

Here are the names the Buffalo Rumblings article you quote cite as 33% "GOLD" in the rest of the 1st round:

Ryan Tannehill, Joe Flacco, Jay Cutler, Aaron Rodgers, Philip Rivers, Ben Roethlisberger, Chad Pennington, Daunte Culpepper, Kerry Collins, Trent Dilfer, Chris Miller, Jim Harbaugh, Jim Kelly, Ken O'Brien, and Dan Marino.

 

I'm going to go out on a limb here and say Old Timers here raise their eyebrows at Dilfer, Miller, and Harbaugh as examples of QB GOLD, while newer chaps aren't so persuaded about Tannehill, Flacco, and Cutler.  I would say yes to T and F but not Cutler.  That would mean 11/45 or 24% (not 15/45 or 33%).

 

The overall point remains - at best, it's 2x as likely to get a good QB in the top 3 picks than in the rest of the round.  At worst, more like 3x as likely.

 

If there's someone you like, and you can, TRADE UP

 

[maryland-bills-fan]

1 hour ago, Hapless Bills Fan said:

 

DUDE!  This has nothing to do with small sample size, murky test criteria, or anything else. What you did is simply mathematically incorrect.

Now I make mistakes, but when my mistake is pointed out, I acknowledge it, I don't start obfuscating about 'statistical difference between the means' and '100% confidence levels' and toss in a Mark Twain quote about liars and statistics, none of which are relevant here and the latter of which is pretty damn insulting.

 

This isn't a quibble; correct math leads one to EXACTLY THE OPPOSITE CONCLUSION as your incorrect statistical calculation

 

To recap, this is what you did:

From Maryland-Bills-Fan post:

working from the OP data you can estimate/guesttimate the following

How many times to you have to draft a QB in those spots to get a GOLD QB? (= 100 /  %chance)
top 3  ..............1.56 picks
rest of 1st ......3.03 picks
2n3 & 3rd ........4.54 picks

 

To calculate the number of picks you need, you are using a value of 100%.  Nothing in statistics reaches a point of certainty, so that's wrong up front.

Then to calculate the number of picks needed for certainty, you are dividing 100% by the % chance of GOLD at different draft positions.

This is simply incorrect statistical calculation.  I'm frankly unclear where you even got this.

 

Here is how it should be done, using your posted values:

top 3: 64% "Gold".   Therefore probability of not-Gold in any one draft choice: (1-0.64) = 0.36

let's say you draft 3x in the top 3 picks.  Your chance of "not-Gold" in all 3 picks is: (0.36)x(0.36)x(0.36) = 0.05. 

This means you have (1 - 0.05) = 95% chance of success at getting a GOLD qb if you draft in the first 3 picks, 3 times.  Yea!  Success!  But 3 picks - not "1.56 picks.

 

(0.36)x(0.36)x(0.36) is often written (0.36)^3, so the equation is: (0.36)^3 = 0.05

The general equation will be (1 - %chance GOLD)^n = 0.05, where "n" is the number of picks needed to achieve a 95% chance of success.

To simplify, you take log of both sides, use the identity log (x^n) = n log (x), then solve for n to get:

n = log (0.05)/log (1 - % chance GOLD).

 

The correct values for 95% chance of success are:

top 3 (64% GOLD)  ..............3 picks
rest of 1st (33% GOLD)......7.5 picks (round up to 8 picks)
2nd & 3rd (22% GOLD)........12 picks

 

So we would need (theoretically) to pick a QB in the 2nd and 3rd round 12 times, to achieve the same odds of success as drafting 3 QB in the top 3 picks of the first.

 

When calculated correctly, it actually makes sense to "burn draft capital" and trade up to the top 3 picks vs. taking a shot at a late 1st or 2nd and 3rd round guy.

3 years of trying (assuming there's a QB we like that high each year) vs 12 years of trying?   Each guy needs a couple years to evaluate/develop, that's the difference between as many 6-9 yrs vs 24-36 years for a "GOLD" QB.

 

Hell to the Yeah, Trade Up

Pox and Pestilence to the years of GMs who put us in this abysmal hole by failing to either trade up or pull the trigger often enough at the lower rounds in the last 25 years.

 

And not incidentally, I reiterated this just in case there are any impressionable young minds reading this: we don't want them to fail AP Statistics

Or even the statistics my kid learnt in 6th grade.

 

And oh yeah: significant figures and common sense, too - 3.03 picks? 4.54 picks?  C'mon Man. 

 

 

 

Sorry, but you don't understand the difference between independent and  coupled variables. Did you ever think that once you had a franchise QB,  you wouldn't keep trying to draft one?   What does "taking the log of xxxx" have to do with this simple thing?   Ok, here it is simple.  Once you got the franchise guy, you stop. The chance to get "the guy" in the first draft is 64%.   So you only try for him in the 2nd draft 36% of the time.  You chance of getting him in the 2nd try is (36% {the chance you are still looking) X 64% (your chance of getting him in the second swing).  Your chance of not getting him that timer and having to try again is 36%.     0.64 + (.36x .64) + (.32x.32x.64) for three years.

First year:  64% chance.   2nd year 87%  3rd year 95.3%.  .  So on average, 1.5 tries is close enough for government work.   Really do you have to be so insulting in you tone?  Don't you think it is childish to quibble about whether to round up or not round up the numbers. Gessh !

 

 

 

Quote

 

 

Edited March 16, 2018 by maryland-bills-fan

[BillsCelticsAngelsBama]

14 hours ago, maryland-bills-fan said:

Steve Billieve

Steve Billieve wrote   <q> "You never know what will happen.  Top 3 overall QBs have a much higher success rate than those late in the first or after.  We've set the board to draft a QB high, which I guess is why this is a thread at all.

Do we change the game plan now?  </q>

 

I don't know what the game plan was, although a lot of people think they do know.   A good manager will have an up-front plan, but should be able to take advantage of changes in the situation. A military axiom is " No plan survives contact with the enemy".  What IS the gameplan?

 

Version #1 (most obvious):  The Bills were planning on tanking in 2017, so they would draft high and  get a franchsie QB in the 2018 draft.  They screwed up and got pick #21 and have been trying to claw their way back to the top of the draft.  They traded away good players (Watkins) to get more draft picks. They desperately traded away their (often) starting left offensive tackle to get a #12 so they might trade up to a top 3 (5?) pick.

 

Alternate Version #2.   The Bills worked in the off season to clear high salary cap, fairly good players in order to stockpile picks to rebuild the team. This includes a shot at a superior QB as well.   They have been wheeling and dealing and the last guy out was the LOF to bump up one of the draft picks to #12. They will draft a QB with that draft and will get a good one.  The rest of the draft (5 more picks in the first three rounds) will be used to restock the offensive line (center!),  linebackers and wide receivers.   They got McCarron as a two year stopgap, to give time for that rookie QB to develop.  They don't have to draft an immediate NFL ready guy who will start in game 1.

 

I think Alternate Version #2 is what they are going with.   By, the way, if we go with plan Version #1,  I will HATE seeing the 5 players we could have drafted beating the Bills in games for the next 7 years.

 

 

 

 

There's no way Version #2 is the plan. IMO they have already discussed trade plans with picks #1-#4 and will know who they are getting by the morning of the draft. The league is different than 35 years ago. You can't breathe on receivers so the value of an accurate QB is more important than back then. This Version  lends itself to management thinking, "what the hell, we'll take the best quarterback left". They are too focused IMO to just let the chips fall where they may.  Unless they go thru workouts and have 5 QB's they like, not likely. then they could trade to #7 Tampa Bay. All these moves were not made to praying and hoping their guy is there at #12.

11 hours ago, wppete said:

Have a feeling we are either trading for #2 or #3 this year. Rosen or Darnold. 

You are on the right track for sure. the Bills were aggressive in trades last year; and that was for a receiver and O-Lineman. If they can find a partner; and I think they probably already have, a tentative one(or two)deals are already in the works.They ARE are trading for our next QB.

[maryland-bills-fan]

9 minutes ago, MOVALLEYRANDY said:

There's no way Version #2 is the plan. IMO they have already discussed trade plans with picks #1-#4 and will know who they are getting by the morning of the draft. The league is different than 35 years ago. You can't breathe on receivers so the value of an accurate QB is more important than back then. This Version  lends itself to management thinking, "what the hell, we'll take the best quarterback left". They are too focused IMO to just let the chips fall where they may.  Unless they go thru workouts and have 5 QB's they like, not likely. then they could trade to #7 Tampa Bay. All these moves were not made to praying and hoping their guy is there at #12.

You are on the right track for sure. the Bills were aggressive in trades last year; and that was for a receiver and O-Lineman. If they can find a partner; and I think they probably already have, a tentative one(or two)deals are already in the works.They ARE are trading for our next QB.

Uh,   I know I drink a lot and am happy to remember why I came into a room, but I don't recall writing that first statement.   Anyway, reality is just an illusion.

 

[BillsCelticsAngelsBama]

8 hours ago, Boca BIlls said:

How about the Bills go get the QB they want... Instead of settling for whoever is left. 

We have a winner !!!!