Friday, March 13, 2009

AdamJT13 projects the compensatory picks...

For the eighth consecutive year and ninth overall, I’ve attempted to project all of the compensatory draft picks that the NFL will award. In my past seven projections, I’ve averaged 23.9 out of 32 exactly correct (going to the correct team in the correct round) and have been off by only one round on an average of 4.1 more. Last year, I got 25 correct and was off by one round on four more. With this year’s projections, I’m hoping to get a combined score of at least 30, although it’s possible that more than the usual number of them could be off by one round because so many projected compensatory picks fell near the cutoff points between rounds.

As the NFL explains, compensatory picks are awarded to teams that lose more or better compensatory free agents than they acquire. The number of picks a team can receive equals the net loss of compensatory free agents, up to a maximum of four. Compensatory free agents are determined by a secret formula based on salary, playing time and postseason honors. Not every free agent lost or signed is covered by the formula.

Although the formula has never been revealed, by studying the compensatory picks that have been awarded since they began in 1994, I’ve determined that the primary factor in the value of the picks awarded is the average annual value of the contract the player signed with his new team, with an adjustment for playing time and a smaller adjustment for postseason honors. It should be noted that the contract values used in the equation seemingly do not include things such as workout bonuses, incentives and conditional bonuses. (Also, keep in mind that the contract figures reported in the media often are incorrect.) And the playing time used in the equation seemingly is the percentage of offensive or defensive snaps played.

A simple method of determining for which qualifying free agents a team will be compensated is this – for every player acquired, cancel out a lost player of similar value. For example, consider a team that loses one qualifying player whose value would bring a third-round comp pick and another qualifying player whose value would bring a sixth-round comp pick but signs a qualifying player whose value would be in the range of a third-round pick. That team would receive a sixth-round comp pick because the signed player would cancel out the loss of the higher-valued player. If the signed player’s value was equal to a fourth-round pick or lower, however, the team would receive a third-round comp pick, because the signed player would cancel out the loss of the lower-valued player.

It is possible for a team to get a compensatory pick even if it doesn’t suffer a net loss of qualifying free agents. That type of comp pick comes at the end of the seventh round, after the normal comp picks and before the non-compensatory picks that are added if fewer than 32 comp picks are awarded. There have been 13 of these “net value” type of comp picks awarded, and in each case, the combined value of the free agents lost was significantly higher than the combined value of the free agents added. In all 13 cases, those teams lost the same number of qualifying free agents as they signed. No team has been awarded a comp pick after signing more qualifying free agents than it lost, no matter how significant the difference in combined value. This year, I’m projecting that Detroit and Arizona will receive a net-value comp picks. Detroit lost three qualifying players (Damien Woody, Boss Bailey and T.J. Duckett) and signed three qualifying players (Brian Kelly, Michael Gaines and Chuck Darby). Arizona lost three (Calvin Pace, Bryant Johnson and Keydrick Vincent) and signed three (Travis LaBoy, Clark Haggans and Bryan Robinson). The combined values of the players Detroit and Arizona lost each exceeded 50 percent more than the combined values of the players they signed. That would be the smallest difference in value of any net-value comp in the past six years, but I’m projecting that it will be enough for both teams.

For the second consecutive year, I’ve used a mathematical formula to weight the three factors that determine a player’s value in the comp equation (his contract, his playing time and his postseason awards). Using this formula, I’ve been able to reconstruct almost precisely the order of the comp picks that were awarded in 2006, 2007 and 2008. In two of those years, the only difference between the reconstructed order and the actual order was that a very small difference in values had the order of two consecutive picks switched. I don’t know if I have the factors weighted correctly, but given that my projected order last year (not the reconstructed order after the actual comps were awarded) matched the exact order of the comp picks in many cases – including one instance of 11 straight, out of the projected picks I had correct – I think I’m probably pretty close.

As always, please note that my comp pick formula is merely an attempt to project the results of the actual (secret) formula. I don’t pretend to know the actual formula. But I think previous results indicate that the formula I use is a pretty good simulation.

In order to qualify for the comp equation, a player must have been a true Unrestricted Free Agent whose contract had expired or was voided after the previous season (i.e., he cannot have been released by his old team); he must sign during the UFA signing period (which ended July 22 last year); if he signs after June 1, he must have been tendered a June 1 qualifying offer by his old team; his compensatory value must be above a specific minimum amount; and he cannot have been permanently released by his new team before a certain point in the season (which seems to be after Week 10) or, possibly, before getting a certain amount of playing time, unless he was claimed off waivers by another team.

The most difficult part about projecting the comp picks is determining all of the cutoff points – the minimum value needed to qualify and the value ranges for the comp picks in each round of the draft. The comp picks awarded in previous years suggest that the cutoff points increase each year by a small percentage – approximately the same percentage by which the leaguewide salary cap increases. From 2007 to 2008, the cap went up 6.96 percent, so I used a 7 percent increase when estimating the cutoff points for this year’s comp picks.

Last year, the lowest-paid player who is known to have qualified for the NFL’s comp equation was Michael Myers, who signed for $825,000 per season and saw significant playing time. The highest-paid player who is known to have not qualified was Mike Doss, who signed for $900,000 per season by saw very little playing time except on special teams. The non-qualifying player with the highest value using the compensatory formula was Chris Liwienski, who signed for $740,000 per season and played almost 90 percent of his team’s offensive snaps. This year, only one player was “on the bubble” for qualifying – Tony Richardson, who left the Vikings and signed with the Jets for $860,000 per season. However, regardless of whether Richardson qualifies, the Vikings and the Jets each signed more qualifying players than they lost, which means Richardson doesn’t affect the comp picks at all.

I’m fairly confident that the players I consider a little “above the bubble” this year (Terry Cousin, Keydrick Vincent and Danny Clark) will qualify for the equation, and that the players I consider slightly “below the bubble” (Alex Stepanovich and Aaron Glenn), will not qualify. The lowest-valued player “above the bubble,” Danny Clark, has a value in the formula that is more than 20 percent higher than that of last year’s lowest-valued qualifying player. And the highest-valued player “below the bubble,” Alex Stepanovich, has a value that is less than that of the lowest-valued qualifying player last year (Michael Myers) and less than 1 percent more than the highest-valued non-qualifying player last year (Chris Liwienski). If I’m wrong about any of those players, it will represent by far the largest or smallest increase in the minimum value needed to qualify that the NFL has used since comp picks were first awarded.

There were two unusual cases this year, one involving Keary Colbert and the other involving Marques Douglas, and they each might or might not count in the comp picks equation.

Colbert was a UFA for Carolina who signed with Denver, was traded to Seattle on Sept. 17, then was cut by the Seahawks on Nov. 12. The only clue about how the NFL handles a player like this in the compensatory formula is the case of Qadry Ismail in 1998. He was a free agent for Minnesota in 1997, signed with Green Bay, then was traded to Miami during the preseason. He did not count in the equation for Minnesota or Green Bay. It’s not known whether he counted for Miami, because the Dolphins didn’t receive any comp picks in 1998, so the NFL never revealed which players counted as lost or signed for Miami. In Colbert’s case, it doesn’t matter whether he counts for Carolina and/or Denver, because neither team will receive a comp pick either way. They each signed more qualifying players than they lost, regardless of whether Colbert counts. It does matter for Seattle, though. If Colbert counts as a player added, the Seahawks would receive three comp picks. If he does not, the Seahawks would receive four. Because Colbert was a member of the Seahawks for only eight weeks, I am projecting that he will not count in the equation.

Douglas was a UFA for San Francisco who signed with Tampa Bay, then was traded to Baltimore on Aug. 27 and played in every game for the Ravens. Whether he counts for Tampa Bay and/or Baltimore is irrelevant, because they each signed more qualifying players than they lost. However, it does matter if he counts for San Francisco. If he counts as a player lost by the 49ers, they will receive two comp picks. If not, they will get one comp pick. I am projecting that Douglas will count as a player lost by the 49ers.

Last year, regardless of playing time or postseason honors, the third-round comp players had signed for at least $6.25 million per season, the fourth-round comp players had signed for $4.67 million to $5.225 million, the only fifth-round comp player had signed for $4.5 million, the sixth-round comp players had signed for $2.25 million to $3.75 million, and the seventh-round comp players had signed for $2.5 million or less per season. Note that there are huge gaps between some rounds, and that there is an overlap between the sixth and seventh rounds because of the adjustments for playing time. You’ll find the contract values for each round of this year’s projected picks in the list a few paragraphs below this one.

As I alluded to earlier, the NFL adds non-compensatory picks if fewer than 32 comp picks are awarded. The non-compensatory picks are given, in order, to the teams that would be drafting if there were an eighth round, until the maximum of 32 has been reached. If there are 28 true comps, for example, the NFL would give additional picks to the teams that would have the first four picks in the eighth round, if there were one. This year, I’m projecting that 30 true comps will be awarded, including Detroit’s and Arizona's comp picks for net-value losses, which I mentioned earlier. Therefore, I’m projecting that Detroit and Kansas City will receive non-compensatory picks to fill out the maximum number of picks. If the NFL’s equation results in more than three non-compensatory picks being added, the next six teams in line to receive one would be St. Louis, Cleveland, Seattle, Cincinnati, Jacksonville and Oakland, in that order.

Here are the projected picks for 2009, along with the compensatory player, their average contract value, their games played, their games started and other notes (I’ve also noted the nine picks that fall near a cutoff point and could end up in a different round) –


THIRD ROUND
New England (Asante Samuel, $9.3567 million per season, 15 GP/15 GS, Pro Bowl)
Pittsburgh (Alan Faneca, $7.8 million, 16/16, Pro Bowl)
Cincinnati (Justin Smith, $7 million, 16/16) – possibly a fourth-round pick

FOURTH ROUND
Chicago (Bernard Berrian, $6.9 million, 16/13) – possibly a third-round pick
N.Y. Giants (Gibril Wilson, $6.5008 million, 16/15) – possibly a third-round pick
San Diego (Michael Turner, $5.75 million, 16/16, Pro Bowl) – possibly a third-round pick
San Diego (Drayton Florence, $5.9333 million, 15/8)
Tennessee (Antwan Odom, $5.9 million, 12/8)
Indianapolis (Jake Scott, $4.8 million, 16/16) – possibly a fifth-round pick

FIFTH ROUND
San Francisco (Kwame Harris, $4.6667 million, 14/11)
New England (Donte Stallworth, $4.5393 million, 11/7)
Dallas (Jacques Reeves, $4 million, 16/16) – possibly a sixth-round pick
Tennessee (Travis LaBoy, $4.4 million, 13/12) – possibly a sixth-round pick

SIXTH ROUND
Tennessee (Randy Starks, $3.885 million, 16/4)
New England (Randall Gay, $3.3125 million, 14/13)
Dallas (Julius Jones, $2.9 million, 15/10)
Cincinnati (Landon Johnson, $2.733 million, 15/0) – possibly a seventh-round pick
San Francisco (Marques Douglas, $2.525 million, 16/0) – possibly a seventh-round pick

SEVENTH ROUND
Tennessee (Ben Hartsock, $2.25 million, 11/11)
Washington (Mark Brunell, $1.755 million, 2/0)
Seattle (Chuck Darby, $1.467 million, 15/15)
Seattle (Ellis Wyms, $1.4 million, 16/0)
Chicago (John Gilmore, $1.333 million, 16/10)
Seattle (Kevin Bentley, $1.3 million, 16/7)
Chicago (Brendon Ayanbadejo, $1.223 million, 16/0, Pro Bowl)
Cincinnati (Bryan Robinson, $1.2125 million, 16/15)
Seattle (D.J. Hackett, $1.2 million, 9/2)
Jacksonville (Terry Cousin, $1.115 million, 16/0)
Detroit (net-value comp pick, lost three for $10.721 million, 38/22; signed three for $7.134 million, 42/31)
Arizona (net-value comp pick, lost three for $9.45 million, 46/42; signed three for $6.9625 million, 40/27)
Detroit (non-compensatory pick)
Kansas City (non-compensatory pick)



As noted, the values of nine comp picks fell near the cutoff points between rounds, so it wouldn’t surprise me if the comp pick for Smith is in the fourth round, if the comp picks for Berrian, Wilson and/or Turner are in the third round, if the comp pick for Scott is in the fifth round, if the comp picks for Reeves and LaBoy in the sixth round or if the comp picks for Johnson and/or Douglas are in the seventh round. (Actually, if Douglas’ value falls in the seventh round, the 49ers’ comp pick would be for Maurice Hicks and would fall between Chicago’s pick for John Gilmore and Seattle's pick for Kevin Bentley.)

Of course, other projected picks could be off by one round (or more) if the NFL happened to change the formula or increase the cutoff points by significantly more or less than I projected.


Here are the qualifying players lost and signed (in order of value) for the 15 teams that I’m projecting will receive comp picks –

ARIZONA
Lost: Calvin Pace, Bryant Johnson, Keydrick Vincent
Signed: Travis LaBoy, Clark Haggans, Bryan Robinson


CHICAGO
Lost: Bernard Berrian, John Gilmore, Brendon Ayanbadejo
Signed: None

CINCINNATI
Lost: Justin Smith, Madieu Williams, Landon Johnson, Bryan Robinson
Signed: Antwan Odom

DALLAS
Lost: Jacques Reeves, Julius Jones
Signed: None

DETROIT
Lost: Damien Woody, Boss Bailey, T.J. Duckett
Signed: Brian Kelly, Michael Gaines, Chuck Darby

INDIANAPOLIS
Lost: Jake Scott
Signed: None

JACKSONVILLE
Lost: Bobby McCray, Ernest Wilford, Sammy Knight, Terry Cousin
Signed: Drayton Florence, Jerry Porter, Cleo Lemon

NEW YORK GIANTS
Lost: Gibril Wilson, Kawika Mitchell, Reggie Torbor
Signed: Sammy Knight, Danny Clark

NEW ENGLAND
Lost: Asante Samuel, Donte Stallworth, Randall Gay
Signed: None

PITTSBURGH
Lost: Alan Faneca, Clark Haggans
Signed: Mewelde Moore

SAN DIEGO
Lost: Michael Turner, Drayton Florence
Signed: None

SAN FRANCISCO
Lost: Justin Smiley, Kwame Harris, Marques Douglas, Maurice Hicks
Signed: Justin Smith, Bryant Johnson

SEATTLE
Lost: Josh Brown, Niko Koutouvides, Chuck Darby, Ellis Wyms, Kevin Bentley, D.J. Hackett
Signed: Julius Jones, T.J. Duckett

TENNESSEE
Lost: Jacob Bell, Antwan Odom, Travis LaBoy, Randy Starks, Ben Hartsock, Chris Brown
Signed: Jake Scott

WASHINGTON
Lost: Mark Brunell
Signed: None

Anyone else who was lost or signed by one of those teams last offseason is not projected to qualify for the equation, for one reason or another. Remember, players have to meet certain criteria in order to qualify for the equation (see the eighth paragraph of these projections for a summary of the criteria), so a lot of players will not count in the equation. Most of the time, it’s either because the player had been released by his previous team and was not a true UFA, or because the player didn’t sign for enough money to qualify.

If I’m wrong about the values of certain players or whether some players will or will not qualify for the equation, that would affect the comp picks. Here’s what would happen in certain instances –

If Jacob Bell’s value as a player lost for Tennessee falls into the range of a third-round comp pick, instead of a fourth-rounder as I projected, then the Titans would get a third-round comp pick for him instead of a fourth-round comp pick for Antwan Odom.

If Jake Scott’s compensatory value is in the fifth round, then the Colts would get a fifth-round pick for him, and the Titans would get a comp pick for Jacob Bell (either in the third or fourth round) instead of a fifth-round comp pick for Travis LaBoy.

If Keary Colbert does qualify as a player signed for Seattle, the Seahawks would not get a seventh-round comp pick for Chuck Darby but would get the other three seventh-round comp picks, and St. Louis would get a non-compensatory pick to fill out the draft order.

If Marques Douglas does not qualify as a player lost by San Francisco, the 49ers would not get a sixth-round comp pick for him. And as I mentioned earlier, if Douglas’ value falls in the range of a seventh-round comp pick, the 49ers’ comp pick would not be for him, but for Maurice Hicks, and it would fall between Chicago’s pick for John Gilmore and Seattle's pick for Kevin Bentley.

If Terry Cousin does not qualify as a player lost by Jacksonville, the Jaguars would not get a seventh-round comp pick for him, and St. Louis would get a non-compensatory pick to fill out the draft order.

If Keydrick Vincent does not qualify as a player lost by Arizona, the Cardinals would not get a third-round comp pick for Calvin Pace, and St. Louis would get a non-compensatory pick to fill out the draft order.

If Danny Clark does not qualify as a player signed by the Giants, they would get a sixth-round comp pick for Reggie Torbor, between New England’s pick for Randall Gay and Dallas’ pick for Julius Jones, and Kansas City would not get a non-compensatory pick at the end of the seventh round.

If Alex Stepanovich does qualify as a player lost by Cincinnati, the Bengals would receive a seventh-round comp pick for him, between Jacksonville’s comp pick for Terry Cousin and Detroit’s net-value comp pick, and Kansas City would not get a non-compensatory pick at the end of the seventh round.

If Aaron Glenn does qualify as a player lost by Jacksonville, the Jaguars would receive a seventh-round comp pick for him, after their comp pick for Terry Cousin (and after Cincinnati’s comp pick for Stepanovich, if the Bengals get one) and before Detroit’s net-value comp pick, and Kansas City would not get a non-compensatory pick at the end of the seventh round.

Under no circumstances will more than 32 picks be awarded, so if I have made numerous significant mistakes and there are more true comp picks than I have projected, one or more of the lowest-valued picks in my projection (starting with the non-compensatory picks and going backward) might not be awarded, if they’re not one of the 32 highest-valued comp picks. Only the 32 highest-valued comp picks are awarded. So, for example, if Stepanovich and Glenn both qualify as players lost, there would be no non-compensatory picks at the end of the seventh round.

The NFL typically awards the compensatory picks on the second day of the Annual Meeting, which would be March 23 this year (the meeting will be March 22-25 in Dana Point, Calif.). After the comp picks are announced, I’ll review what the NFL did and where my projections were incorrect (although I’ve already presented some other possibilities).