Trade Value Charts

[nycdan]

As the talk about draft trades amps up, it's probably a good time to discuss everyone's favorite NFL Draft Trade Value chart.  

I'm going to blow past the history of it and focus on what it is, and how to distinguish between the growing number of versions.

What is it?  The value chart is at its heart, a curve function showing a declining value for each successive pick.  It is generally meant to follow a logarithmic scale where the difference in value decreases with each successive pick.  So the difference in value between picks 1 and 2 is greater than between 2 and 3, and so on.  This makes perfect sense from a mathematical perspective.  

The challenge is in determining what those differences in value should be.  In the absence of any other data (e.g. if you didn't know the pool of prospects) you would likely use some version of a normal log function.  But each year, we do know something about the prospect pool so that changes things.

Now let's look at the usual go-to chart that we are all used to - Draftek.  https://www.drafttek.com/NFL-Trade-Value-Chart.asp  This is the so-called Jimmy Johnson chart in unmodified form.

Now look at the top 10 rows:

1  JAX	3000
2  DET	2600
3  HOU	2200
4  NYJ	1800
5  NYG	1700
6  CAR	1600
7  NYG	1500
8  ATL	1400
9  SEA	1350
10  NYJ	1300

Big problem.  They show a gap of 400 pts from 1 to 2, 400 pts from 2 to 3, 400 pts from 3 to 4, and then 100 pts for each between 4 and 8.  They are essentially saying the value of the draft is massive for picks 1 to 3 and then declines to nearly flat from 4 and beyond.  But when you look at the actual players, there's no logic to that.  In fact, the value from picks 1 through 10 is probably as flat as its been in years.  Certainly there is no way DET or HOU would pay the massive price it would take to move up 1 position according to this chart.  It suggest that we would get a low 3rd round pick to slide down to #5, but would pay a mid 2nd round pick to move up to #3.  Meanwhile, DET would pay the same mid 2nd round pick to move up from #2 to #1.  Put another way, it would cost DET the same to go from 2 to 1 as it would cost ATL to go from 8 to 4.  Hogwash.

There is another chart that has a much smoother curve called the Rich Hill chart.  It looks like this: https://www.drafttek.com/NFL-Trade-Value-Chart-Rich-Hill.asp

1  JAX	1000
2  DET	717
3  HOU	514
4  NYJ	491
5  NYG	468
6  CAR	446
7  NYG	426
8  ATL	406
9  SEA	387
10  NYJ	369

The gaps are tighter once you get past HOU at 3 but I would argue the gaps from 1-2 and 2-3 are still too big for this year.  This is because this chart is still not based on this year's prospects.  The other problem is the gaps are really too small when you compare to the value of the gaps.  This chart says the value of swapping picks 6 and 3 is 68 pts.  That's the value of the 71st pick (high 3rd round).  Clearly that's not going to get it done.

Jason, at overthecap has another chart, but while the distribution of the top-10 is the smoothest of all, the gaps are insanely small.  This is because they use salaries to calculate value.  As a result, their gap to trade from #6 to #3 is the equivalent of the 186th pick in the draft, a mid-6th rounder.  Come on, Jason.  You can do better.

My only point here is to point out, as has been done before, that the Jimmy Johnson chart is very flawed.  In a draft with one prospect that stands out above all others (e.g. John Elway, Trevor Lawrence), the #1 pick is worth a great deal more in comparison to the #2 than it is this year, when there is no consensus #1 pick (and when it's not a QB). 

So when thinking about trade value, take that into consideration.

 

[jetstream23]

I use this one from Rich Hill...

[nycdan]

28 minutes ago, jetstream23 said:

I use this one from Rich Hill...

 

Yeah, that's the second one I mentioned.  Probably as good as any but some of the gaps are weird.

PIT moving up from 20 to 10 would cost them pretty much just their 2nd round pick.

CAR moving up from 6 to 4 would cost them a low 3rd rounder (#90 pick).

On the flip side, us swapping #4 for #3 with HOU would only cost us our 2nd 4th round pick #117.

Maybe that's close in this year's draft, but those numbers do feel very low to me.

 

[section314]

4 hours ago, jetstream23 said:

I use this one from Rich Hill...

Same here

[chirorob]

On 3/30/2022 at 1:10 PM, jetstream23 said:

I use this one from Rich Hill...

Even that.

Would you trade the #4, #10, and a 2nd rounder for #1 this year?   No way I'd do that.

[jetstream23]

2 hours ago, chirorob said:

Even that.

Would you trade the #4, #10, and a 2nd rounder for #1 this year?   No way I'd do that.

Wouldn't do it either.

WAAAAY too much to give up and although Hutch is probably the best Edge there are plenty more to choose from.

[nycdan]

11 hours ago, chirorob said:

Even that.

Would you trade the #4, #10, and a 2nd rounder for #1 this year?   No way I'd do that.

That's the point.  Last year, a team would have traded far more than that to move up for Golden Boy.  These charts have to be recalculated every year and even then, each team should assign different values.  If you didn't need a QB last year, the value of #1 would be far less than if you did.

Trades can happen when teams have different values on the same picks.  Honestly, in a perfect world, trades should be much more common as teams shuffle their picks to optimize value.  But there is an inherent resistance based on fear of screwing up.  Call it trade friction.  If every team had a trade chart perfectly customized to their view and needs, there would be dozens of trades in the first round alone.  Stick and Pick is actually a really dumb strategy that only makes sense because it's basically impossible to quantify pick value well enough.

[chirorob]

1 hour ago, nycdan said:

That's the point.  Last year, a team would have traded far more than that to move up for Golden Boy.  These charts have to be recalculated every year and even then, each team should assign different values.  If you didn't need a QB last year, the value of #1 would be far less than if you did.

Trades can happen when teams have different values on the same picks.  Honestly, in a perfect world, trades should be much more common as teams shuffle their picks to optimize value.  But there is an inherent resistance based on fear of screwing up.  Call it trade friction.  If every team had a trade chart perfectly customized to their view and needs, there would be dozens of trades in the first round alone.  Stick and Pick is actually a really dumb strategy that only makes sense because it's basically impossible to quantify pick value well enough.

100% agree.

People always use Jimmy Johnson's chart.   That was 1 arbitrary chart that he made for that specific draft.

[nycdan]

28 minutes ago, chirorob said:

100% agree.

People always use Jimmy Johnson's chart.   That was 1 arbitrary chart that he made for that specific draft.

I'll go a step further.  If I was trying to design a truly optimized chart for the draft, the values would reset with every pick.  Every time the top player on your board isn't picked, the next pick becomes a little bit more valuable because part of that original value incorporated the uncertainty of who would be available.  If Player X was worth moving up to the 20th pick for example, and he's still on the board at 21, there's really no good reason you shouldn't be willing to offer just as much to move up to 21.  In fact, you'll get a slightly cheaper contract so it would actually be a little more favorable at that point.

I also don't think the charts take into account the changes in contracts in the 1st round and between rounds 1 and 2 (i.e. the length and option years).  It's possible, even probable, that teams do these things internally but just don't publicize them.  

[Beerfish]

The chart takes a lot of flack but it is and has been a very useful tool.  From NFL teams to us amateur schmucks it lays some though imperfect, still useful parameters to work from.