Jump to content
Sign in to follow this  
Hapless Bills Fan

QB Draft Stats: probability of finding That Guy

Recommended Posts

So better not try amirite 

  • Haha 1

Share this post


Link to post
Share on other sites
6 hours ago, Hapless Bills Fan said:

This has almost become my yearly pre-draft post (like Just Jack's preseason post, only less exciting and more informational)

The source is Pro Football Reference Draft Finder, coped into Excel for better slicing and dicing.  In previous years, I've used the 'eyeball test', this year I used hard numeric cutoffs.
Shout-out to Shady Bills Fan for help with the Excel files.  He's a Right Guy!

 

Looking at the first 5 rounds of QB drafted in prev. 20 years (2017 omitted for insufficient data, and didn't go back further due to arguments about rule changes)

I calculated three statistics I consider important for QB e v a l: completion percentage, YPA, and TD/INT ratio.

My sort criteria were: Greater than 59% completion, Greater than 6.5 YPA, Greater than 1.5 TD/INT (practically speaking, that means if a guy throws 3 TD in a game, he throws 1 and not 2 INTs)

 

Here's what the summarized data look like for success rate in picking a QB who can do these things, by draft round.  1st round broken down further.

image.png.2ed13c6e1ce9975765fa200d546fb068.png

 

Bottom line: even at the top of the 1st round, the odds of getting a good QB are something like 50-50.  At the bottom of the 1st round, it falls to about 20%, which is the same as the 2nd round.  If you play with the criteria a bit, it may rise to 1 out of 3 (30-33%) at the bottom of the 1st round.

I'll include the names of the QB these three criteria sorted below, not sure it will be legible: the bottom line is you can nitpick names and cutoffs, but the "song remains the same" overall.

Surprises to me: criteria excluded 1st round QB Eli Manning, Cutler, Culpepper, Bridgewater, Campbell.  In the later round the surprise exclusion was Schaub (likely be included without his final year).  If you add the criterion of averaging >220 ypg, you exclude Kaepernick and Tannehill

 

The data suggest a couple things.   Not shown, but with 2 exceptions, all the successful high 1st round QB were pick 1 or 2.  Therefore, it may be unwise to mortgage too much of the draft to get to the 1st round 3-5 picks: Go Hard, or stay home.  The success rate in picks 6-10 is no higher than the success rate in the 2nd round.  It rises from pick 11-20, so if we're going to trade up, trading up a few picks may be the value strategy vs trading up to pick 3-10. 

After the 3rd round, it's basically throwing darts.  This is why letting someone else throw the dart, and trying to pluck the bulls eye off the dartboard is a popular choice.

 

If you start writing a response, "You need to look at this that or the other data or recalculate everything with the moon over the formahaut and X set to Malignify", please save us all time and dial 1-800-Bite-Mee - No seriously, check your personal vals, bro.  You're welcome to take and massage the data for yourself.   If you message politely I'll even get you my excel file.

image.thumb.png.67f757ae3b0265b23e5a8648c17ec4a0.png

image.thumb.png.45ad075d645ebff916f6ed70113f4493.png

 

 

 

Interesting info.  It would be more relevant if picks by the Bills and Browns were excluded :)

Share this post


Link to post
Share on other sites
7 hours ago, Hapless Bills Fan said:

 

and TD/INT >= 1.5

 

Dont like it?  1-800....is the number *smooch*. 

 

I think it works - it produces a list of QB you can win with given a decent D, includes most QB I think it should, and excludes most QB I think it shouldn't.

I highlight the exceptions - most of the surprise exclusions failed on TD/INT.  Most of the surprise inclusions have low YPG.  Keep in mind I'm using career data so one has to allow for guys who had early bad years then the switch flipped on.

 

But you are welcome to collate the data yourself and slice it any way that pleases you.

 

Im good. You’ve done fine work here. 

 

Also tells me you have decent probability outside of the top 5 so do your homework qb scouting staff. 

Share this post


Link to post
Share on other sites
2 minutes ago, Over 29 years of fanhood said:

 

Im good. You’ve done fine work here. 

 

Also tells me you have decent probability outside of the top 5 so do your homework qb scouting staff. 

 

Yes, I think that was one of the key observations for me - good scouting in the late 1st, 2nd, or 3rd could really pay off

1 hour ago, joesixpack said:

So better not try amirite 

 

That's not my take-home but I put the data out there - spin it as you wish

  • Like 1

Share this post


Link to post
Share on other sites
3 hours ago, PolishDave said:

 

So using your criteria and stats posted.....

 

Percentage of times teams found a decent Qb in 

 

1) Top 10 draft picks                                        45%             (teams decided they absolutely MUST have this qb)           

2) Round 1 - picks 11-32                                  18%             (teams decided they want the guy but won't trade up hardcore with first picks for him)

3) All round 1 picks combined                        34%             

4) Rounds 2 and 3 combined                          18%             (teams think guy might make a franchise QB, but serious question marks)

5) Rounds 4-5                                                       8%             (unlikely to become Franchise guy soon but take a shot because he has enough upside chance)

 

 

Fun interpretation:

Based on these odds it makes sense for the Bills to trade down with both of their first round picks so that they have at least 3 first round picks next year.    Odds are they get a guy because 34% each guy drafted.

 

Or:

 

Take both first round QB's this year.   If you don't hit, take one in the first next year.

 

Or

 

Take one first rounder this year.

 

If you don't hit, take one first rounder next year.

 

If you don't hit there, take one first rounder the following year.

 

Whoever came up with the idea of taking a QB in the first every year until you hit on one is a genius.

 

 Unfortunately, statistics don’t work that way. 

Share this post


Link to post
Share on other sites

Mcdermot and Beane have a plan for the QB position and they will implement it this off- season....

  • Like 1

Share this post


Link to post
Share on other sites
20 minutes ago, Schmuggs said:

Mcdermot and Beane have a plan for the QB position and they will implement it this off- season....

 

As long as it's not a plan where they lose their hat.  Any plan where you lose your hat is a Bad Plan

Share this post


Link to post
Share on other sites

So maybe we shouldnt even try right, like the Bills have done the last 17 years. 0% chance to win Super bowl with Tyrod.

  • Like 2

Share this post


Link to post
Share on other sites

I think you should draft a QB every year , with the 1st, 2nd, 3rd or possibly the 4th ( think of Dak Prescott). If you hit on one you are golden. If you hit on two You can recoup draft picks up the ying yang.

Share this post


Link to post
Share on other sites

Thank you for this post.  It should be brought back every year at draft time.  Really good QBs are very hard to find.  It isn’t as easy as just trading “whatever it takes to go get your guy”.

Share this post


Link to post
Share on other sites

I find it prophetic, hilarious and sad that the Bills First Round Pick falls exactly into the WORST percentage slot on this chart.  Now THAT is Billsy!

Share this post


Link to post
Share on other sites
2 hours ago, Tatonka68 said:

So maybe we shouldnt even try right, like the Bills have done the last 17 years. 0% chance to win Super bowl with Tyrod.

 

Tatonka68, I'd like to introduce you to your new BFF Joesixpack.

Joesixpack, Tatonka68.

 

You guys have a lot to say to one another, I feel eternally sure

Share this post


Link to post
Share on other sites

Any list that has Manning, Newton and Vick in the no category is an epic fail IMO

Edited by matter2003
  • Like 1

Share this post


Link to post
Share on other sites
7 hours ago, PolishDave said:

Whoever came up with the idea of taking a QB in the first every year until you hit on one is a genius.

 

I have a question about this strategy.  If you do that, how are you going to know if any of these guys are really any good before you have to shuffle one off to get the next one?

 

Let's play this out.

 

2018 - Bills draft QB B in i round 1. He plays.  I assume you have a vet and Nasty Nate and QB A.  

2019 - Bills draft QB B in round 1.  Who plays?  Can you really judge QB A on one year?  So you dump the vet.  Now you have QB A, QB B, and Nasty Nate.

2020 - Bills draft QB C in round 1 QB A has played 2 years so so .  QB B has hardly played. Cut Nasty Nate.

2021 - Bills draft QB D in round 1. Now you have cutting one of your first round pick QBs.  

 

If you don't hit soon, then you got a bunch of 1st round picks that haven't played much.  The rest of your team suffers because all your first round picks go to one position that only 1 can play at a time.  The other first rounders are  riding the pine. Each QB gets one, maybe 2 years. Can you really judge a guy that fast?  

 

I don't know.  Could work I suppose.

 

Other teams have come close.  Jets and Clowns come to mind.  Jets drafted Gino, Petty, and Hackensack in the period of about 5 years and they're still playing rental vet QBs like Fitz and McCown.  The draftees have hardly played.  Clowns, well they go through QBs like a person with the flu goes through kleenex.

  • Like 1

Share this post


Link to post
Share on other sites
10 hours ago, PolishDave said:

 

So using your criteria and stats posted.....

 

Percentage of times teams found a decent Qb in 

 

1) Top 10 draft picks                                        45%             (teams decided they absolutely MUST have this qb)           

2) Round 1 - picks 11-32                                  18%             (teams decided they want the guy but won't trade up hardcore with first picks for him)

3) All round 1 picks combined                        34%             

4) Rounds 2 and 3 combined                          18%             (teams think guy might make a franchise QB, but serious question marks)

5) Rounds 4-5                                                       8%             (unlikely to become Franchise guy soon but take a shot because he has enough upside chance)

 

 

Fun interpretation:

Based on these odds it makes sense for the Bills to trade down with both of their first round picks so that they have at least 3 first round picks next year.    Odds are they get a guy because 34% each guy drafted.

 

Or:

 

Take both first round QB's this year.   If you don't hit, take one in the first next year.

 

Or

 

Take one first rounder this year.

 

If you don't hit, take one first rounder next year.

 

If you don't hit there, take one first rounder the following year.

 

Whoever came up with the idea of taking a QB in the first every year until you hit on one is a genius.

 

Nope. It takes you at least 3 years to know if you have a good QB or not, so if you take a bunch year after year how do you evaluate them?

 

It's not Madden. You can't see their stats right after the draft.

Share this post


Link to post
Share on other sites

I was shocked when the Bills did not draft Mahomes or Watson without having to trade up. Tre White is a great player if they can land a good QB without trading up.

Share this post


Link to post
Share on other sites

The top 5 guys are 50/50 and I think that's because they are expected to be the 'guy' and carry a franchise, often with poor rosters. The early 2nd round have a much better success rate than I expected. Is it because they are going to a franchise where they are not expected to carry the franchise and be the 'guy' right away? Or is it because they are going to teams with a better roster around them? Does anybody really think Jimmy Garropolo would be the QB he is today had he went to the Browns and not New England?

 

Having a solid prospect is important but drafting a prospect with a talented team around him is just as equally important to them not busting. I'm not big on expecting a 20 something year old to be a savior, they need to be handled properly. Sit them down behind a professional, build a roster around him and take the pressure off of him. That is just as important as which prospect you draft or where in the draft you take him.

 

  • Like 1

Share this post


Link to post
Share on other sites
9 hours ago, reddogblitz said:

 

I have a question about this strategy.  If you do that, how are you going to know if any of these guys are really any good before you have to shuffle one off to get the next one?

 

Let's play this out.

 

2018 - Bills draft QB B in i round 1. He plays.  I assume you have a vet and Nasty Nate and QB A.  

2019 - Bills draft QB B in round 1.  Who plays?  Can you really judge QB A on one year?  So you dump the vet.  Now you have QB A, QB B, and Nasty Nate.

2020 - Bills draft QB C in round 1 QB A has played 2 years so so .  QB B has hardly played. Cut Nasty Nate.

2021 - Bills draft QB D in round 1. Now you have cutting one of your first round pick QBs.  

 

If you don't hit soon, then you got a bunch of 1st round picks that haven't played much.  The rest of your team suffers because all your first round picks go to one position that only 1 can play at a time.  The other first rounders are  riding the pine. Each QB gets one, maybe 2 years. Can you really judge a guy that fast?  

 

I don't know.  Could work I suppose.

 

Other teams have come close.  Jets and Clowns come to mind.  Jets drafted Gino, Petty, and Hackensack in the period of about 5 years and they're still playing rental vet QBs like Fitz and McCown.  The draftees have hardly played.  Clowns, well they go through QBs like a person with the flu goes through kleenex.

 

6 hours ago, MJS said:

 

Nope. It takes you at least 3 years to know if you have a good QB or not, so if you take a bunch year after year how do you evaluate them?

 

It's not Madden. You can't see their stats right after the draft.

 

Oh yeah.   There's that.  :thumbsup: :lol:

 

Share this post


Link to post
Share on other sites
On 1/13/2018 at 9:03 AM, Hapless Bills Fan said:

I calculated three statistics I consider important for QB e v a l: completion percentage, YPA, and TD/INT ratio.

My sort criteria were: Greater than 59% completion, Greater than 6.5 YPA, Greater than 1.5 TD/INT (practically speaking, that means if a guy throws 3 TD in a game, he throws 1 and not 2 INTs)

Taylor stats for 2017:

 

14TD/4INT = 3.5:1 TD to INT

Completion %: 62.6

YPA: 6.7

 

By your above criteria, wouldn't that make Tyrod "That Guy"?

 

Share this post


Link to post
Share on other sites

Why even field a QB if they are so hard to find, might as well just play in the wildcat all game.

Share this post


Link to post
Share on other sites
On 1/14/2018 at 2:42 PM, Young34 said:

Taylor stats for 2017:

 

14TD/4INT = 3.5:1 TD to INT

Completion %: 62.6

YPA: 6.7

 

On 1/14/2018 at 2:42 PM, Young34 said:

By your above criteria, wouldn't that make Tyrod "That Guy"?

 

 

If we'd drafted him high, it would certainly make him a QB that we or others would still be interested in.  (If TT were a 3rd year guy who had started as a rookie and been drafted high, some teams arguably would have more patience because of the above).

 

I considered adding a 4th criteria of YPG, something pretty mild like >220 ypg. 

 

I didn't, because I didn't have the data to justify it as a QB criterion with good correlation to team wins and I didnt want to play "massage the data until the QB sort exactly as I think they should sort".  But if I had, it would have taken out three quarterbacks.  Two of them were Colin Kaepernick and Tyrod Taylor.

 

I chose my three criteria as QB stats correlated to team wins, to shed light on the odds of just picking a QB who can play in the NFL, vs an out-and-out bust and they seem to have done that job.  Keep in mind what others have alluded to:  If we tightened criteria to choose only elite QB, then the odds of finding one plummet.

On 1/14/2018 at 2:44 PM, Tatonka68 said:

Why even field a QB if they are so hard to find, might as well just play in the wildcat all game.

 

If you could actually win consistently in the NFL playing wildcat all game, I'm sure someone would do it.

 

That's also missing the point of the exercise as I see it, which isn't to say "it's hard, so don't try". 

 

As I see it, what the data say are:

1) the draft, even high in the draft, is a percentage, not a certainty, even for experts who make this their job.  If we're going to trade up to the top of the draft, mortgaging all other positions, we better be sure.

2) the Bills historical approach over the last 20 years, which is to draft or acquire 1 guy and give him 3 years (without exploring other avenues during that time), is a Fail per the probabilities

3) there's a case to be made that signing a good FA, even at a high price, is a value strategy (again, if the experts are sure) because it spares high draft picks for other positions

Share this post


Link to post
Share on other sites
On 1/14/2018 at 1:27 AM, reddogblitz said:

 

I have a question about this strategy.  If you do that, how are you going to know if any of these guys are really any good before you have to shuffle one off to get the next one?

 

Let's play this out.

 

2018 - Bills draft QB B in i round 1. He plays.  I assume you have a vet and Nasty Nate and QB A.  

2019 - Bills draft QB B in round 1.  Who plays?  Can you really judge QB A on one year?  So you dump the vet.  Now you have QB A, QB B, and Nasty Nate.

2020 - Bills draft QB C in round 1 QB A has played 2 years so so .  QB B has hardly played. Cut Nasty Nate.

2021 - Bills draft QB D in round 1. Now you have cutting one of your first round pick QBs.  

 

If you don't hit soon, then you got a bunch of 1st round picks that haven't played much.  The rest of your team suffers because all your first round picks go to one position that only 1 can play at a time.  The other first rounders are  riding the pine. Each QB gets one, maybe 2 years. Can you really judge a guy that fast?  

 

I don't know.  Could work I suppose.

 

Other teams have come close.  Jets and Clowns come to mind.  Jets drafted Gino, Petty, and Hackensack in the period of about 5 years and they're still playing rental vet QBs like Fitz and McCown.  The draftees have hardly played.  Clowns, well they go through QBs like a person with the flu goes through kleenex.

QB A plays linebacker.  Solved.  Next?

Share this post


Link to post
Share on other sites
Sign in to follow this  

×