Wednesday, March 26, 2014

NFL Draft Trade Machine 2014

Now that free agency is winding down it is time for teams to shift their focus to the draft. This handy trade machine will allow you (and them) to play GM by constructing trades and seeing winners and losers according to the classic draft value chart and a number of updated versions, including my own. 

For a look at the analysis that underlies my revised draft value chart, check out this post here.

For a look at evidence for team skill in drafting year over year, within a single draft and after making a trade up, check out the respective links.

If you've forgotten which teams have which picks here's a link to that.

Otherwise, enjoy the 2014 Sports + Numbers Draft Trade Machine:

If you enjoy this, have a look at my top posts from last year and my recent four-part analysis of the impact of changing teams on performance and value.

Wednesday, March 5, 2014

Cap goes up, football less interesting

The NFL makes a lot of money, every year (see here for some details from the only team that publishes financials) and lately the cap hasn't been going up at all. Given that background I am certainly happy for the players with this year's $10 million rise in the salary cap for each team.

As far as my recent analysis is concerned, however, this means the already-limited practice of free agency is getting even more limited. I talked in part 4 of my series on free agents about considering why players are available. One of the best options for a good value was to seek out players who were cut for cap reasons. With an extra $10 million even capped-out teams such as Carolina, Dallas, Washington and Pittsburgh are suddenly not capped-out.

Bill Barnwell hits on all these points in a good roundup of winners and losers from the change.

The bottom line is that free agency this year will be an even worse place to get value than it usually is. Teams who would previously have made decisions for non-football reasons are now left to hold onto basically all of the players they want to hold. Teams who might have scooped up those players, such as Cleveland, Oakland and Jacksonville, will be left fighting for scraps. A boost from having no money to $10 million is far more impactful than a boost from $55 million to $65.

You could look at this and craft a narrative that the teams with tons of cap space are getting screwed, but given the steadily increasing revenue for the league and promises of a percentage of that to players in the CBA it was bound to filter through the extremely complex formula eventually. Teams that bet on it happening this year are big winners. All teams, more able to retain their best players, will be that much more dependent on the draft for now. That and finding longshots in the CFL.

Sunday, February 9, 2014

The value of free agents in the NFL - Part 4

For those of you who haven't read Part 1, Part 2 or Part 3, I'll start with some facts:

Players who change teams perform worse 

This one is very straightforward. AV per player drops by a greater amount for both starters and non-starters when a player changes teams.
Figure 15 - Coefficients of Prior AV for Prior Starters

We can assume that teams are better at scouting and valuing their own players than players on other teams, so that’s part of the reason why players who change teams decrease in value. The other one as shown in this figure is the interaction between players in football. Offensive roles, especially on the line, require close cooperation with other players so changing teams results in a less effective player.

Players who change teams get paid less

This one is less intuitive. When I think of free agents who have changed teams, I am usually thinking of the too-big-to-turn-down offer from another team (to be fair I’m usually thinking of the Indians losing Manny Ramirez to Boston). What the data show us for the NFL during the 2003-09 period, however, is that players changing teams are actually earning less than comparable players who stay on the same team.

There are a couple things that could be affecting this. First of all, not all players can change teams in all years. If a player has a significant signing bonus spread over many years and a team in a tight cap situation, they may decide to keep him rather than cut him and take the cap hit. This would show up in our data as a player who maintained (plus or minus a bit) his salary from year to year even though the decision was not driven by the age and performance metrics we are looking at.

Second, and more interestingly, I think this may be an example of sticky wages. Employees tend to reject falling nominal wages, leading businesses to prefer layoffs over reducing salaries. Since behavioral factors make it tough to come to agreement on reduced wages players whose current teams would be willing to keep them at a reduced salary may be shifted into the “changed teams” pool. With players destined for a salary cut in the group the results are skewed that way even if other players take the Manny Ramirez route and land a big pay raise.

Teams with more homegrown players perform better

Way back in part 1 we saw that teams with more homegrown players, particularly with more homegrown starters, outperform those with fewer. The effect is small but persistent. In the 2010-2012 period there are still very few examples of teams with few homegrown starters performing well. The change is that more teams appear to be embracing draft-driven rebuilds (as opposed to free agent-driven) so there are more examples of teams with many homegrown starters performing poorly.

What about value?

We can’t say anything definitively about the value given the regression results in part 3. Instead we’ll look to Brian Burke’s article on the topic of draft picks, ‘Bricklayers or Gladiators,’ for inspiration. The conclusion of the article – that performance concentrated in one person is more valuable than equivalent performance from multiple people – applies directly to the results we have here. With players changing teams delivering value indistinguishable from players staying with their teams, the dominant factor in preferring those who stay with their teams is their performance level.

If the players changing teams definitively provided better value we would have our work cut out for us figuring out whether the lower performance is worth it in light of the higher value. Luckily for me I don’t have to look into that. 

One last stat to consider on the Bricklayer/Gladiator topic (click the link above and read the article if you haven’t already) – among prior starters who change teams performance drops from 5.76 to 4.53 or 21%. Among prior starters who don’t change teams performance drops only 14% from 7.40 to 6.41. Starting from a lower base those who change teams drop more in relative and absolute terms.

The impact of the CBA

To think about how the agreement between the NFL and the players union affects player personnel decisions, consider the following 2 by 2[1]. The y-axis indicates whether a player’s current team wants to keep them at the market rate. The x-axis indicates whether any other team in the league would want to sign the player at the market rate.

All of this assumes the market rate for a player, which varies depending on the interest from other teams, the CBA minimum salary and, for some players, the franchise tag.

Starting in the top right, the CBA provides teams with significant leverage over players because most contracts contain significant non-guaranteed money. If a team wants to keep a player they can continue under the contract or renegotiate with the remaining years of the original contract as leverage for good players and the threat of release as leverage for lesser players. Most starters on their rookie contract fall in this quadrant as other teams would gladly take them at the price of their current contract.

Teams also enjoy ongoing leverage over elite players via the franchise tag. If a team wants to they can hold a player indefinitely via the tag – two years is the practical max – and the fact that players know this is more power for the team in contract negotiations. Players who are wanted by their current team and other teams are more likely to end up with their original team even if another team might conceivably give them more money. The other team can’t make that offer because the current team can invoke the tag and pull the player off the market (I know this is out of sequence with the tag coming before the free agent period opens, but stick with me). The big players – Brees, Brady, and Manning, in the future perhaps Colin Kaepernick, JJ Watt or Richard Sherman – never hit the market and sign deals that are below what they would get on the open market.

The players who get to free agency just below this elite level, those who aren’t worth the franchise tag but still perform on the field, frequently end up in the bottom right. This type of player is too expensive for the current team – either in the team’s independent judgment of their value or in the context of their broader salary cap situation. These players are also most susceptible to the effects of the Winner’s Curse since they have multiple teams bidding on them. In the 2013 free agent market players like Mike Wallace, Paul Kruger and Wes Welker typify this group.

In addition to these there are the players whose team does not want them because of other known reasons – someone who needs a “change of scenery” would fit in this category, as would veterans chasing a ring leaving a team that is obviously rebuilding or players displaced by a switch in scheme.

Finally, there are players whose team does not want them because of non-publicly-known information. Teams might believe a player has an elevated risk of injury. They might know about behavior issues that have not been leaked to the media or caused any league discipline. They might simply have more information about the player’s ability – reports from the coaching and scouting staff based on practice, off-season workouts and other sources.

The left side as a whole does not have a lot of players in it. As for the top left, other than a smattering of franchise favorites on elevated veteran’s minimum salaries who wouldn’t be picked up by other teams, it is likely to be filled with players on their rookie contracts who have relatively low cap hits and provide depth on the roster. Other teams don’t have any reason to believe they are better than their end-of-roster players – though cast-offs from successful franchises might generate more interest.

The part of the CBA that really affects this quadrant is the minimum salary. If a player in this quadrant could be replaced by a rookie undrafted free agent, that’s probably the end of their time with their team. When they hit the market, however, they end up in Canada or the Arena League rather than another team because once their existing team doesn’t want them they move top left to bottom left.

If you’re a GM, how do you deal with all this?

This is going to sound simplistic: acquire players most likely to provide value for you.

Players on their rookie contract are hit or miss, but the misses are cheap and the hits more than cover them. I (and many others) have devoted considerable time to demonstrating that there is no persistent skill demonstrated in the draft. Teams who had a good draft the prior year are no more likely to have one this year. Teams who make a good pick are no more likely to have their subsequent pick be a good one. Selections made after trading up, when teams are so confident that they pay a premium to obtain the pick, are no better than those made with an assigned pick.

Given the low probability of any individual pick becoming a home run teams need to acquire as many picks as they can and supplement those with undrafted free agents and promising players from other leagues. The only constraint on this process should be the ability of coaches and scouting staff to evaluate the players while keeping up with all of their other duties.

Even teams in relatively good shape in terms of roster depth should be preoccupied with acquiring players to develop into starters so they will maintain roster flexibility as existing starters become too expensive over time. This is a classic problem for Super Bowl winners as there are too many players looking for a raise – and plenty of teams are willing to provide one – to fit under one salary cap.

If you need to plug holes in free agency, behave like a value investor. Seek out places where value is most likely by identifying where there is favorable reason for the player to be available. Players jettisoned in a salary cap-related purge over those let go from a team with tons of cap space. Players released due to off field concerns over those available cheaper than you thought (e.g., Davone Bess) for no obvious reason. Players who haven’t had the chance to prove themselves – stuck behind stars if possible – over those who have started and failed to impress.

None of this is revolutionary stuff. Teams face long odds going through free agency because a player’s original team has far more information about them than anyone else. All other teams considering the player are left with whatever the player has done in games, whatever information is still relevant from the draft scouting process and whatever they glean from coaches or players who have second- or third-hand information.

As demonstrated way back in the first post of this series, teams with more homegrown starters tend to outperform those with fewer. Under a rookie wage scale that favors teams even more than before don’t expect that to change anytime soon.

[1] As a consultant I am obligated to use 2 by 2s in any situation in which a concept could be clearly explained, but would be made slightly more opaque and mysterious through use of a stylized graph, implying analytical rigor and insight not present

Wednesday, January 29, 2014

The value of free agents in the NFL - Part 3

Check out Part 1 and Part 2 of this ongoing and overly long series on the value of free agents in the NFL

It doesn’t matter if your players perform slightly worse if you can get a good deal on them. We’ve been looking at performance but value is the real target here.

Before we get all the way to value we’ll start with a look at cost.

The average player who changes teams is not cashing in on a big payday in the way we might think. While players like Paul Kruger (in 2013 to the Browns) or Matt Cassel (in 2009 to the Chiefs) come to mind when looking at free agency, NFL contracts are basically a series of one year options held by the team. 

For good players, the ability to hold out can give them some leverage for a raise as long as they continue to be good. If a player is merely average they can only hope that the team would have to take a big cap hit to cut them, most won’t see a dollar more from the team if they are let go[1].

Figure 10 - Change in SC% (Prior Starters)

Figure 12 - Changed Team
Figure 11 - Same Team

In the season before changing teams, the average salary cap hit for a player who will change is 2.20% (1.24% for non-starters, 2.87% for starters). Once they change their salary cap hit is 1.56% (0.89% for prior non-starters, 2.03% for prior starters). Most of these players are taking a pay cut to stay in the league.

Tuesday, January 14, 2014

The value of free agents in the NFL - Part 2

The previous post in this series (check out Part 1 here if you missed it earlier) looked at whether teams do well based on how many homegrown players they have or start. There’s some small ability to explain performance but it doesn’t get to the level of detail needed to inform specific allocation decisions. Looking at how well individual players perform when changing teams, however, is a promising way to get enough information for just that kind of decision.

Player Performance

There are a couple trackable and quantifiable factors we might expect to predict a player’s performance in the next season. First and foremost is how they are performing this year. We will be using Approximate Value as created and calculated by the good folks at here as a proxy for performance.

As I qualify any analysis based on this data set, AV is necessarily and admittedly approximate (it’s in the name if you look closely). By using it in a data set as large as this we should avoid having random one-offs impact the results significantly. A great AV typically results from a great player on a good to great unit (offensive or defensive). A good AV means a player played a significant role on a good unit or a huge role on a middling one. A middling AV might be a decent player on a bad team or a rarely-used player on a good team. A low AV is a player who didn’t play or a player who played on a very bad unit.

Average AV by Performance
Pro Bowl (All)
PB (not All-Pro)
Starter (All)
Starter (no AP/PB)

Next up is whether a player changed teams. It’s the point of the analysis and it would be a bit disappointing if we didn’t include it here. Those players who change teams should be expected to suffer a decline in performance.

We could conceivably see differences in the ability to integrate into a new team based on the position played with less complex positions potentially requiring less adjustment and yielding better performance for team-changers. To attempt to address this, we’ll split data set into four sets with one for each combination of offense/defense and skilled/unskilled and run the regression for each[1].

The age of a player and their number of years in the league might also be informative so we’ll add them to the regression[2]. They will interact somewhat but we’ll let the model determine which one reflects the positive bias in the data (see below) and which one reflects the inevitable decline of a player as they age. In place of a constant, these two terms will interact to create a base for the average player/years played combination. For example in this data set the average player in their first season is 23.33 years old, 24.32 in their second, 25.29 in their third and so on.