Boom or Bust: A Look at Draft Probabilities/NFP

[RamBill]

Boom or bust: A look at draft probabilities
What are your picks' chances for success?

By Tony Villiotti

http://www.nationalfootballpost.com/Boom-or-bust-A-look-at-draft-probabilities.html

In my last article, I laid out the case for splitting the NFL draft into seven draft choice ranges that do not correspond to present rounds. I further postulated that all draft choices within each range have about equal value in terms of future success. To refresh your memory, the seven draft choice ranges were:

1. Selections 1-13
2. Selections 14-24
3. Selections 25-46
4. Selections 47-73
5. Selections 74-114
6. Selections 115-187
7. Selections 188 and later

In this article I review the probability of achieving various milestones for each draft choice range. Ten milestones were selected for review and are listed below, along with the draft years considered in the probability calculations.

Regarding the draft years included in each analysis, the general policy was to establish a measurement period that allows one extra year to achieve a milestone For example, in determining whether a player started for five seasons or more, he was allowed six years to achieve those five years as a starter. This meant cutting off the five-year starter analysis with the 2008 draft year so as to allow draftees from 2008 six seasons to achieve both five years in the league and five years as a starter.

I acknowledge that a full analysis cannot be done until after a player’s career is over. The methodology employed in this article, for instance, does not count the player who takes longer than the one-year grace period to achieve a milestone. Take Jordy Nelson of the Packers as an example. He was drafted in 2008 but did not achieve starter status until the 2011 season, his fourth year. While he has only three years as a starter right now, it is highly likely that Nelson will ultimately achieve five-year starter status but he is not counted as a five-year starter yet.

The following table shows the probability of achieving each of the milestones for each of the draft choice ranges. For example, 91.3% of the players selected in the 1-13 draft positions have historically played five years or more and 71.3% of them were starters for at least five years.

This table also demonstrates that the biggest change in going from one draft choice range to another is in drafting players who subsequently earn post-season honors. For example, the probability of drafting a five-year starter declines modestly (10% or so) going from the 1-13 group to the 14-24 group. The decline in players winning post-season honors, though, is about 40%.

Some may prefer data by draft round instead for draft choice range. This next table reports the same information by draft round.

 

[jrry32]

Very interesting. About what you'd expect. And IMO, the top percentage for 2+ time All Pros being 6.2% helps explain my personal stance on selecting elite talent...like Clowney...when you have the chance.

 

[Alan]

What it tells me is don't waste an early pick trading up because you just lost two starters to gain one highly unlikely 2+ PB player. 6.2% is a suckers bet. It's all about the probabilities. Never ever trade when it costs you an early pick unless it's for a QB. That's smart drafting IMO.

 

[FRO]

Very interesting. About what you'd expect. And IMO, the top percentage for 2+ time All Pros being 6.2% helps explain my personal stance on selecting elite talent...like Clowney...when you have the chance.

Agreed

 

[Elmgrovegnome]

I took it differently. Trade the pick for the boom or bust players to a team willing to gamble for them, but at 6% it isn't worth it.

 

[brokeu91]

I did the math, it's kind of complicated. But to make a long story short, if we use both picks within the top 13 this year, based on these stats we have a 14.1% chance of selecting someone who will go to 2 or more pro-bowls

I don't know what that would mean to you guys, but I say go for it

 

[Selassie I]

So.... don't trade down Les.

 

[CGI_Ram]

So.... don't trade down Les.

Well, there is a point of diminishing returns.

You got to have 5 fourth rounders, for example, to equal the rookie start ratio of a first rounder.

That's why I'd be careful trading down. Picks are great. But premium picks is where it's at.

 

[Ramhusker]

The smartest thing is to trade down but not too far. Trading from 2 to 4 or 6 will keep your pick in the 1-13 category while gaining another pick in the 25-46 category and probably another pick in the 74-114 category. It's all a gamble anyways. There's always a Jason Smith or a Ryan Leaf in the crowd.

 

[LesBaker]

So.... don't trade down Les.

OK you got it.

 

[max]

I did the math, it's kind of complicated. But to make a long story short, if we use both picks within the top 13 this year, based on these stats we have a 14.1% chance of selecting someone who will go to 2 or more pro-bowls

I don't know what that would mean to you guys, but I say go for it

Check your numbers. If you have a 23.1% chance of getting a player who will go to 2 or more pro bowls if you use a top 13 pick on him, then the chances have to increase that you get one of those guys if you have more than one pick.

Actually, I did the math. The number you should get is 40.9%. That means if we use both our picks in the top 13, then according to the NFP table we have a 40.9% chance of getting at least one player who will go to 2 or more pro bowls. N = 1- ((1-23.1%)^2)

The chances of getting 2 pro bowl studs is included in the 40.9%, however, there is only a 5.3% chance of that specific outcome. Much simpler...
N = 23.1%^2

And even better is the odds of getting at least one guy who makes at least one pro bowl. = 68%

 

[brokeu91]

Check your numbers. If you have a 23.1% chance of getting a player who will go to 2 or more pro bowls if you use a top 13 pick on him, then the chances have to increase that you get one of those guys if you have more than one pick.

Actually, I did the math. The number you should get is 40.9%. That means if we use both our picks in the top 13, then according to the NFP table we have a 40.9% chance of getting at least one player who will go to 2 or more pro bowls. N = 1- ((1-23.1%)^2)

The chances of getting 2 pro bowl studs is included in the 40.9%, however, there is only a 5.3% chance of that specific outcome. Much simpler...
N = 23.1%^2

The number was 10.9%. In this scenario to get at lease one person who will go to 2 all pros, the math would be 10.9 +10.9-(10.9)^2. I made an error on the calculation, it comes out to 21.8.

 

[max]

The number was 10.9%. In this scenario to get at lease one person who will go to 2 all pros, the math would be 10.9 +10.9-(10.9)^2. I made an error on the calculation, it comes out to 21.8.

Check again. I think you're mixing up your All Pros with your Pro Bowlers.

 

[Rams and Gators]

What it tells me is don't waste an early pick trading up because you just lost two starters to gain one highly unlikely 2+ PB player. 6.2% is a suckers bet. It's all about the probabilities. Never ever trade when it costs you an early pick unless it's for a QB. That's smart drafting IMO.

The aim should always to be to find the elite player, you should be able to find a starter from UDFA/Low cost FAs, I'd happily trade Barksdale and Dunbar for example for Quinn.

 

[Elmgrovegnome]

Anyone would make that trade.

One thing being ignored is overall team make up. How many dire needs are there?

The other thing is available prospects. In some years the entire top ten is lousy. I would have been estatic if the Rams had been able to trade back instead of drafting Jason Smith for example.

 

[max]

The aim should always to be to find the elite player, you should be able to find a starter from UDFA/Low cost FAs, I'd happily trade Barksdale and Dunbar for example for Quinn.

A better example would be trading Lauranitas and Chirs Long for Kuechly. And then drafting Clowney.

 

[Alan]

Rams and Gators having trouble counting with his shoes on:
The aim should always to be to find the elite player, you should be able to find a starter from UDFA/Low cost FAs, I'd happily trade Barksdale and Dunbar for example for Quinn.

Name me 5 low cost UDFAs/FAs that the Rams acquired that were quality starters. Dunbar wouldn't fit that category and Barksdale is the exception that proves the rule. You're advocating a drafting strategy that is successful 10.9% of the time versus my strategy that is successful at a vastly superior rate for everything except 2+ PB appearances.

Let's look at at some recent history.
Atlanta traded away the farm for Julio Jones because they thought they were only a player away and Julio Jones was elite. Drafting 6th only 3 years later.
Washington traded away the farm for RGIII because they thought he was their future franchise QB. Drafting 2nd only 2 years later.
St. Louis traded up to draft Tavon Austin because we thought he was elite. Drafting 3 picks later this year because our record got worse.

Care to give me your examples of recent success stories?

So yeah, I disagree with that drafting philosophy. The great thing about math is it's black and white. The great thing about forums is you frequently have multiple conflicting views that give everyone something to talk about.

 

[Rams and Gators]

Name me 5 low cost UDFAs/FA that the Rams acquired that were quality starters. Dunbar wouldn't fit that category and Barksdale is the exception that proves the rule. You're advocating a strategy that is successful 10.9% of the time versus my strategy that is successful at a vastly superior rate for everything.

Let's look at at some recent history.
Atlanta traded away the farm for Julio Jones because they thought they were only a player away and Julio Jones was elite. Drafting 6th only 3 years later.
Washington traded away the farm for RGIII because they thought he was their future franchise QB. Drafting 2nd only 2 years later.
St. Louis drafted Tavon Austin because we thought he was elite. Drafting 3 picks later this year because our record got worse.

Care to give me your history of recent success stories?

So yeah, I disagree with that drafting philosophy. The great thing about math is it's black and white. The great thing about forums is you frequently have multiple conflicting views that give everyone something to talk about.

Who said that they had to be quality starters, Brandon Gibson has a chance to fit into the 5 year starter rule, Dunbar is a 4 year starter, and I'd be surprised if he doesn't start the required 8 games this year.

I'm assuming the low cost guys have to be current so I can't have London Fletcher and Kurt Warner . Can I at least have Danny Amendola? No? OK I'll give 5 from last year who fit into the started 8 games criteria, Kellen Clemens, Cory Harkey, Jo-Lonn Dunbar, Joe Barksdale, Rodney McLeod, didn't we claim Chris Williams off the waiver wire claim? I'll count him. Can I have Hekker? technically as a punter he doesn't qualify as a starter but he qualifies as a Pro Bowler. Now Clemens will never start for 5 seasons but Harkey has every chance, Dunbar is a season away, Barksdale will, McLeod may not depending what we do in the draft, and Williams will most likely start at least 8 games next season.

Surely for the Jones/RGIII/Austin argument to work then the other side of the trade needs to pay off:
Cleveland went from a 5 win team and won 4, 5 and 4.
The Rams went from 2 wins to 7 and 7.
And the Titans went from a 6 win team to a 7 win team.
The only team it payed off for was the Rams, and as you point out they went a blew it by trading up themselves.

 

[Alan]

Rams and Gators asking to borrow my hearing aids:
Who said that they had to be quality starters, Brandon Gibson has a chance to fit into the 5 year starter rule, Dunbar is a 4 year starter, and I'd be surprised if he doesn't start the required 8 games this year.

I'm assuming the low cost guys have to be current so I can't have London Fletcher and Kurt Warner . Can I at least have Danny Amendola? No? OK I'll give 5 from last year who fit into the started 8 games criteria, Kellen Clemens, Cory Harkey, Jo-Lonn Dunbar, Joe Barksdale, Rodney McLeod, didn't we claim Chris Williams off the waiver wire claim? I'll count him. Can I have Hekker? technically as a punter he doesn't qualify as a starter but he qualifies as a Pro Bowler. Now Clemens will never start for 5 seasons but Harkey has every chance, Dunbar is a season away, Barksdale will, McLeod may not depending what we do in the draft, and Williams will most likely start at least 8 games next season.

Surely for the Jones/RGIII/Austin argument to work then the other side of the trade needs to pay off:
Cleveland went from a 5 win team and won 4, 5 and 4.
The Rams went from 2 wins to 7 and 7.
And the Titans went from a 6 win team to a 7 win team.
The only team it payed off for was the Rams, and as you point out they went a blew it by trading up themselves.

I just did didn't I? So if a team has a hole and they pick someone off the street and put a uniform on him that makes him a starter? Hell, I could fit that description. I look really good in a uniform BTW. That's why I I said quality starters. I assume you're trying to build a good team through FA and the draft. Placeholders are only that, placeholders. How does Hekker fit into your UDFA/FA strategy? I'm pretty sure we drafted him. Clemens wasn't a starter, he was an injury replacement. McLeod sucked. How did Amendola turn out for the Cheatriots? You're adding Dunbar who we cut? Twice? In a row.

How does Hekker fit into your UDFA/FA strategy? I'm pretty sure we drafted him. Clemens wasn't a starter, he was an injury replacement. McLeod sucked. How did Amendola turn out for the Cheatriots? You aren't serious mentioning Chris Williams are you? So as I look at your 5 players I see Barksdale. Would it make it easier for you if I changed quality to average or slightly below average? Then you can include Dunbar's first year with us.

So what's your overall strategy? If it's adding guys like Chris Williams, who we jettisoned (a description that pretty much fits every single player you mentioned) even though we have huge holes at his position, then I for sure don't like your plan.

So you didn't address my math.

I'm not saying your strategy can't work. It has worked occasionally in the past but it hasn't worked very often and the math that RamBill posted is why. You can argue with me but it's hard to argue with math.

Did I mention that I love math? :seizure: :rockon: :cheers:

 

[jrry32]

What it tells me is don't waste an early pick trading up because you just lost two starters to gain one highly unlikely 2+ PB player. 6.2% is a suckers bet. It's all about the probabilities. Never ever trade when it costs you an early pick unless it's for a QB. That's smart drafting IMO.

Well, no, it's all about talent evaluation. And the 6.2% is a 2+ time All Pro. And that "suckers bet" is worth making if you're very good at what you do.