Trading Up: The Desperation Markup

Last night’s first round of the draft was pretty crazy with a lot of unexpected developments.  My mock mock draft correctly picked three out of the thirty-two selections, which is two more correct than I was expecting.  Of the other mock drafts in my sample, the results ranged from zero picks right to six picks right.  All in all, that’s pretty weak but around what I expected.

One of the three picks that I got right was Deshaun Watson going at #12, which is all thanks to stupid teams overpaying for meh quarterbacks.  Three of the six trades made last night involved trades for quarterbacks and two included the Browns (this is their Super Bowl).  So I bet you’re curious to know if these trades were rational, right?  If so, you’re in luck (I may have given away my conclusion already…).  Below is a quick analysis of these trades using my chart and the Jimmy Johnson chart for reference.  Chase beat me to the punch and posted something very similar which I would encourage you to read as well.

Quick Note on my Methodology

Three of these trades, those involving the quarterbacks, included future 2018 draft picks.  There are a few ways to value future picks relative to current picks.  The first is to assume the pick falls in the middle of the respective round.  The other is to use an indicator to predict where the pick is likely to fall based on how the team is expected to perform in the coming year.

Bill Barnwell, an ESPN football writer, claims that point differential (or Pythagorean wins) “is historically a better predictor of [a team’s] future win-loss record than [its] actual win-loss record from the previous season.” On its face, I agree with him.  If teams lose a bunch of close games, it may not accurately signal the strength of that team.  In baseball, this has shown to be true, though it’s easier to derive a conclusion when there are 162 games in the sample each year rather than just 16.

I decided to test this idea by calculating the correlation between win percentage the following year and each current year win percentage metric: Pythagorean win percentage, estimated win percentage (click the link if you want to learn more about this), and actual win percentage.

The correlation between the following year win percentage and Pythagorean win percentage as well as with estimated win percentage is very close to zero, which indicates that there is little relationship between these metrics and a team’s performance the following year.  That is, they’re useless for predicting how a team will perform next season.  Note that since I do not include other variables (another way to approach this is to run a regression with the following year win percentage as the dependent variable, the selected win percentage metric as an independent variable, and include other independent variables as control variables), this shouldn’t be taken to mean more than it shows here.  However, it does tell us that we can’t reliably predict next year’s win percentage from these metrics, as Barnwell claims.

The correlation between win percentage and next year’s win percentage is still fairly low but it is higher than the other two.  Based on this, win percentage the previous year is a more reliable indicator of next year’s performance.  Therefore, I used the 2017 draft order as the hypothetical 2018 draft order.  I also applied a completely arbitrary (since apparently no one can figure out a reasonable range for this rate) 15 percent discount rate to future picks to account for the time value of draft picks.

Now that I bored you with these details, let’s analyze each of these trades including the “winner” of each trade.  The “winner” of each trade is the team that acquired more capital today from the trade.  In the long run, though, the winner of the trade will be the team that translates that capital into actual value.  Teams may pay a premium (the “desperation markup”) to acquire a pick to select a player whose value, according to the team, is equal to or higher than the desperation markup.

Applying the Desperation Markup

1.  49ers trade #2 to the Bears for #3, #67, #111, and a 2018 3rd round pick

The 49ers selected Mitchell Trubisky here with the assumption that if they didn’t make this trade, the 49ers would sell the pick to another team interested in Trubisky.  We’ll never know who the other interested team(s) were but the 49ers used their leverage to acquire a significant amount of draft capital from the Bears.

According to the Jimmy Johnson chart, which significantly overvalues the first half of the first round, the difference between what the Bears gave up to acquire Trubisky is equal to the 89th pick (the 25th pick of the third round). However, my chart, which is empirically derived unlike the Jimmy Johnson chart, shows that the Bears sacrificed value equal to the 14th pick in the draft! That’s a large desperation markup to pay to move up one spot.  That’s betting a lot on Trubisky’s marginal value compared to a different player who the Bears could have selected with the third pick.

To clarify, this doesn’t mean that the 49ers should have included a first round pick in the deal. That doesn’t make any conceptual sense.  What it does mean is that a more equitable trade would replace the two third rounders and the 4th rounder the Bears lost with a fifth round pick.  Removing those picks and adding the fifth rounder is approximately equal to the value of the 14th pick.

2.  Bills trade #10 to the Chiefs for #27, #91, and a 2018 1st round pick

The Chiefs didn’t pay as high of a desperation markup for their quarterback of the future as the Bears did, in terms of relative draft capital.  Moving up 17 spots in the first round is a costly endeavor, but it shouldn’t cost as much as the Chiefs paid.  According to the Jimmy Johnson chart, the Chiefs should have demanded a fourth round pick in addition to #10 to make this more equitable (though the Bills don’t own a fourth round pick this year).  According to my chart, however, the variance is equal to the 37th pick.  Overall, the Chiefs could have made this bold move but paid a little less (or asked for a little more).  The Chiefs expect to compete next year with Alex Smith as their quarterback and this analysis assumes that they will.  Any decline in play or unlucky injuries (Teddy Bridgewater anyone?) could make this trade even worse before factoring in whether Pat Mahomes will live up to his potential.

3.  Browns trade #12 to the Texans for #25 and a 2018 1st round pick.

Similar to the Chiefs, the Texans expect to compete next year for the Super Bowl, especially if Deshaun Watson or Tom Savage (or just anyone not named Brock Osweiler) can provide solid quarterback play and J.J. Watt and co. on the defense stay healthy.

I’ve seen arguments that what the Texans really gave up for Watson should include the Brock Osweiler trade to the Browns (Osweiler, 2017 sixth round pick, and 2018 second round pick for 2017 fourth round pick).  Sorry folks, but that violates some economic principles and I just can’t have that.  That trade (and signing Osweiler in the first place) are sunk costs.  They occurred in the past.  Rick Smith, the GM of the Texans, can only focus on the status (and future status) of his roster at each moment. During the draft, he clearly decided that the price, including the desperation markup, to acquire Watson was worth it, exclusive of earlier moves he made.

4.  Seahawks trade #26 to the Falcons for #31, #95, and #249

5.  Seahawks trade #31 to the 49ers for #34 and #111

The Seahawks traded out of the first round last night, starting with trading down to #31 and then trading that pick for #34.  I don’t know who they’re targeting with their first pick (I would guess an offensive linemen) but they clearly thought that the player (or players) would be available at #34.  In the process they picked up an extra third rounder, fourth rounder, and seventh rounder.

In terms of the first transaction, the Falcons overpaid by the value of the 177th pick according to the Jimmy Johnson chart or by the value of the 111th pick according to my chart.  The Falcons saw a player they liked, Takkarist McKinley, and knew that their new climate-controlled stadium in Atlanta was perfect for him.

The transaction between the Seahawks and 49ers is interesting to me from a mathematical perspective.  This is the first trade where it appears that my chart shows a lower desperation markup than the Jimmy Johnson chart.  My chart places a greater value on mid-round picks which means that #111 is worth more on my chart.  Because a lot of teams continue to use the Jimmy Johnson chart and this is supposed to be a deeper draft, I think we’ll see a lot more teams trade up during the middle rounds (and sacrifice other mid-round picks in the process – Note: there were 38 trades during the 2017 draft, breaking the previous record of 34 in 2008).  The consequences of trading up are lower than they think they are and the benefit is supposedly higher than usual.

6.  Packers trade #29 to the Browns for #33 and #108

This trade should be great news for Browns fans.  First, it shows that the Browns are willing to use their massive amount of draft capital rather than always trust the process.  Second, the Browns paid a really small desperation markup (equal to a seventh round pick).  This makes sense to me given their recent analytics bent.  They were not about to pay an irrational price to move up which they accomplished.  They also now have five years of control over David Njoku (who’s only 20!) rather than four had they selected him at #33.

Conclusion

In all six cases, the team that traded up “lost” the trade.  We won’t know for years if these teams actually lost these trades until their careers play out but for now, they paid a desperation markup to trade up and acquire a player.  Let’s hope for the fans of these teams (and for the job security of the GM) that it works out for these teams.  For the teams that traded down, you’ve acted rationally and acquired more draft capital.  Now it needs to show on the field.

Leave a Reply

Your email address will not be published.