How NFL teams ignore basic economics and draft players irrationally

[PatsWickedPissah]

Actually, risk diversification is the point of the article. They are saying it is advisable to spread, rather than consolidate.

Akin to investing in the stock market, you don't put all your money in stock from one company, or your money in just one market; the idea being to spread your investment across the markets for diversified returns, to increase the likelihood that you get any return at all.

I think I see the problem now...
"So extrapolating, the best VALUE is to trade down with the 1st pick using multiple lower round picks targeting the ~4th best guy at your desired position"

What I meant to convey that their point was to trade down for draft pick value, accumulating multiple pickS targeting a much lower rated player or players at specific positions. I can see that my wording might be interpreted as using a bunch of lower picks just to target one specific guy. Problem may be in my mis-use of the English language.

 

[Ring 6]

Yeah, but what about impact? The second player might start as many games, but what is the difference in impact?

Solid starter versus playmaker or fringe starter versus solid stater?

I think that would be more interesting to know.

I don't think the study was done, or used by the OP to conclude that numbers of games started is iron clad proof of a better pick, but to be used as one verifiable criteria to illustrate the point.
Its really not possible to look at every trade and come up with an agreed upon value of the players careers.

As with most things discussed on this board, a point in an argument is not the be all and end all of the argument.

 

[PP2]

I think I see the problem now...
"So extrapolating, the best VALUE is to trade down with the 1st pick using multiple lower round picks targeting the ~4th best guy at your desired position"

What I meant to convey that their point was to trade down for draft pick value, accumulating multiple picks targeting a much lower rated player or players at specific positions. I can see that my wording might be interpreted as using a bunch of lower picks just to target one specific guy. Problem may be in my mis-use of the English language.

Yes, I guess you could put it that way. I think it's a little more aligned with boom or bust type of risk, in that when you draft more players, the risk of bust is diluted.

 

[SB39]

The problem with looking at overall, long term averages is that it ignores the times when you have one clear-cut star available, like a (Peyton) Manning or a Luck.

Besides, you can give me all the stats you want but ultimately what it comes down to is how good the people making the decisions are at making decisions.

 

[Bill Lee]

Well Wendell started every game last year.

Right, which means the Pats think he's better than any other backup they had all year, which puts him into a class of players (starters) that are better than another class of players (non-starters). Given that a game has 22 starters out of 53 players, it is a differentiating factor, one that gets amplified by 16 games per season and 32 teams per league. Yes, there will be some non-starters better than starters but that's a statistical oddity.

Also, we see that Tom Brady is a LOT better than Ryan Mallett which cancels out the Wendell Factor!

Besides, you can give me all the stats you want but ultimately what it comes down to is how good the people making the decisions are at making decisions.

Yet they'd be idiots if they ignored statistics (and also be idiots if they used statistics they didn't understand). It's clear this statistic is based on the action of those decision makers, it isn't just being pulled out of thin air.

 

[RayClay]

Actually, risk diversification is the point of the article. They are saying it is advisable to spread, rather than consolidate.

Akin to investing in the stock market, you don't put all your money in stock from one company, or your money in just one market; the idea being to spread your investment across the markets for diversified returns, to increase the likelihood that you get any return at all.

I would say this is the closest. If you're not sure, and skillful, the value is in having more picks. Teams that trade up often are sure their player is superior. There is also the intangible injury factor, bad fit etc.

The Patriots have also hit with first rounders, getting 3 of the best at position in Wilfork, Mayo (and a pick) andMankins, according to an article I posted as thread starter.

The Pats trade up sometimes and trade down to align their pick with the consensus, like they did with McCourty and Mayo. The Pats seldom spend much on a trade up, though.

 

[RayClay]

Right, which means the Pats think he's better than any other backup they had all year, which puts him into a class of players (starters) that are better than another class of players (non-starters). Given that a game has 22 starters out of 53 players, it is a differentiating factor, one that gets amplified by 16 games per season and 32 teams per league. Yes, there will be some non-starters better than starters but that's a statistical oddity.

Also, we see that Tom Brady is a LOT better than Ryan Mallett which cancels out the Wendell Factor!



Yet they'd be idiots if they ignored statistics (and also be idiots if they used statistics they didn't understand). It's clear this statistic is based on the action of those decision makers, it isn't just being pulled out of thin air.

Right. Being a starter is only one measure, that was my point.

 

[PP2]

The problem with looking at overall, long term averages is that it ignores the times when you have one clear-cut star available, like a (Peyton) Manning or a Luck.

Besides, you can give me all the stats you want but ultimately what it comes down to is how good the people making the decisions are at making decisions.

The article doesn't say you must never do that.. it's just saying that on average, it's sensible to trade down and diversify.

 

[RayClay]

I don't think the study was done, or used by the OP to conclude that numbers of games started is iron clad proof of a better pick, but to be used as one verifiable criteria to illustrate the point.
Its really not possible to look at every trade and come up with an agreed upon value of the players careers.

As with most things discussed on this board, a point in an argument is not the be all and end all of the argument.

Right, it's one factor. My take is, unless you're sure that is the guy, much better than the next pick (not necessarily next at that position either) don't spend the Capital.

I'm sure the Pats were happy to get Jones, not so much burning a 3rd and 7th to move up 10 spots for Graham (ouch), unknown for Hightower at this point.

If you're sitting at 29 and your choices look fairly even (likely in a deep draft) you're crazy not to take a good trade down.

Conversely, if you're sitting at 21 and Wilfork has dropped all the way from top 5-10 to your spot, you're running to the podium.

I think what the author is implying is falling in love with the player you just have to have is foolish if you need to trade up, or if you refuse a good value trade down.

 

[RayClay]

Draft pick value was the substance of the article
Salaries were only mentioned in a brief aside and not by me

There's a section on salaries as an important factor. There's no doubt that's true in the salary cap area.

 

[lurker1965]

Hey, nice article, but wrong metric If player A starts 104 games and player B starts 100 that is only 4% difference. If player A has 5 pro-bowls (or all pros, whatever) and player B is a journeyman starter there is more than a meager difference. The metrics do not measure it.

 

[PP2]

Hey, nice article, but wrong metric If player A starts 104 games and player B starts 100 that is only 4% difference. If player A has 5 pro-bowls (or all pros, whatever) and player B is a journeyman starter there is more than a meager difference. The metrics do not measure it.

Like Andy said, there is no one true fireproof metric. The numbers of games started is just a criteria that can be statistically measured.

 

[BradyFTW!]

While these studies are useful to a point, when you begin and end your evaluation with "games started", it tells a fraction of the story. By games started, Brandon Pettigrew (or insert TE of your choice, really) is better than Rob Gronkowski.

 

[patsfan13]

THe idea of risk diversification is very valid there are other factors that are harder to track. THe Patriot roster for example over the past 10 years is generally harder to make than many other rosters.

Also when BB get a player in camp he tends to pay less attention to their draft spot than other GM's, hence the number of UDFAs on the Pats roster over the years.

Then we have the issue of rating players and stacking the board. It seems that when the PAts see a number of players grouped with similar ratings they will frequently trade down to get a player in the same 'group' + a pick frequently a future pick.

Many (most?) teams are in a situation where they are pressured to win in a short window so GM's coaches don't lose their jobs, this will frequently lead to panic moves. BB isn't in that situation so he can be more disciplined in his approach to the draft.

 

[Nunchucks]

I would say the Patriots pick in this draft goes against this philosophy, they picked talent over value

 

[Bill Lee]

Yes, we can all think of mediocre players that are starters but how many superstars can we think of that are NOT starters? Every metric you can think of has some flukes, but that doesn't make it a meaningless metric.

 

[RayClay]

I would say the Patriots pick in this draft goes against this philosophy, they picked talent over value

That's why this study only indicates one factor. The Patriots will go after a player when they consider him superior for what they want. With a pick in the bottom of round two (and reportedly Seattle ready to pull the trigger) they went after their man. Also, by the fact a QB wasn't picked for a few turns, it seems their wasn't a trade down available.

It isn't a "philosophy", it's just a statistical analysis. The Patriots have no problem trading up for the right guy (Chandler Jones, Hightower), but they are more disciplined about falling in love with and overrating players. Also, they rarely make huge tradeups.

Good drafters consider many factors and change as the league changes.

 

[Urgent]

Another good commentary on the Cade Massey analysis of the economics of the NFL draft.

At it's simplest:
Projecting player performance is highly uncertain. Highly-drafted players fail to perform, and undrafted players emerge as starters.

Imagine that players in the first round, on average, have a 60% chance of starting for five years.
Players in the second round have a 50% chance, and players in the third round have a 40% chance.

Three scenarios:
Keep the first round pick - expected value of starts = 48 starts (5 years * 16 games * 60%)
Trade the first round pick for two seconds - expected value of starts = 80 starts (5 years * 16 games * 50% *2)
Trade the first round pick for three thirds - expected value of starts = 96 starts (5 years * 16 games * 40% * 3)
The actual research from Massey and Thaler shows that the percentage differences are not as sharp as 60% - 50% - 40%, and using total starts simplifies the analysis.

But it's basic probability theory - teams consistently significantly overestimate their ability to project the performance of players. They absolutely cannot do it accurately. They absolutely believe they do.

Thus the team that recognizes this flaw can consistently improve performance. By trading down.

One can find any individual player, and point out that a player drafted earlier turned out better.
One can point out that all starters are not of the same quality.
But over time and over large numbers, these trends are supported.

If you could only draft one single player in a single draft, you would rather have a higher draft pick.
If you can draft multiple players over multiple years, you are better off optimizing your expected value.

 

[Rusty Coupe]

I would say the Patriots pick in this draft goes against this philosophy, they picked talent over value

I think Easley fits into one of BB main ideas which is one player fitting multiple roles, which also lends itself to value.

• Here’s what Bill Belichick had to say about Easley last night:

“He’s a versatile guy that’s played different positions along the defensive line. … You see him against guys that are either in the league from previous years or guys that you have on your draft board this year or you’re going to have on there next year. I think he’s a very disruptive player and he showed up against good competition.”

“He’s an explosive player. Very explosive. … He’s played everywhere along the defensive line. It depends on where you put him. I’ll say you don’t see a lot of guys who do that – who play, I mean, he lines up on the nose, he lines up on the guard, he lines up on the tackle, he lines up out wide at times. You can see him playing all those spots.”

http://www.bostonglobe.com/sports/2014/05/09/easley/Oykrdgn3D8acVC5Sn7h2aL/story.html

 

[RobAllan]

The problem with looking at overall, long term averages is that it ignores the times when you have one clear-cut star available, like a (Peyton) Manning or a Luck.

Besides, you can give me all the stats you want but ultimately what it comes down to is how good the people making the decisions are at making decisions.

Actually, in a way, you just made the point of the article. I hate to bring this up, but there was a genuine debate about whether Manning or Leaf was a better pick in that draft. Both were relatively clear-cut stars, which means that the boom/bust effect was 50% even for that position in that draft. In hindsight, I'm ashamed to admit I actually liked Leaf better at the time. (I already had a hate on for Manning even before he was a Colt.)