Roster Building Analysis

The rules committee has been thinking generally about the cost and value of free agents in comparison to other ways of acquiring players.  I put together some data around this to try to quantify the three methods: franchising, draft, free agency.  The metric I used to measure value was the number of games played.  This is admittedly a flawed metric, but I think it has some value and was easier to collect then others I could think of.

I looked at every player from 2006 to the present (4 years since we expanded last). Heres what I found out:

	players	games	g/
franchising	429	2951	6.9
draft	442	2406	5.4
fa	277	1017	3.7

So roughly the same number of players come onto a roster every year from franchising as from drafting but franchised players have contributed just shy of 7 games per compared to 5.5 for drafted players and just under 4 per year from each free agent acquired.  However, these numbers shifted a bit after 2007.

Starting in 2008, we shortened the draft by 2 rounds.  In 2006 and 2007 the draft accounted for the most players on a roster: 42% compared to 34 and 37% for franchising and 21 and 24% for free agency.  After 2007, the draft accounted for around 36% compared to 39% for franchising and 25% for free agency.

This change also impacted games played as one would expect.  Prior to 2008, free agents accounted for 13% of games played compared to 39% for drafted players and 48% for franchised players.  In 2008 and 2009 free agents accounted for 19% of all games played.  In 2008 drafted players accounted for 33% and franchised for 48% while in 2009 drafted players accounted for 39% and franchised for 42%.

One final view of this.  In the chart below, I restated every player's number of games played as a percentage of their possible games and taking into account when a free agent was acquired during the season.  So a drafted player who plays 13 games played 100% of their possible games.  A free acquired before the last week of the regular season and who played 1 game that years also played in 100% of their possible games.  So the numbers below so the average free agent, drafted player, and franchised player for that year.

So, in some ways free agents were most successful last year averaging a start in just under 60% of their available games while franchised players have been performing less over these last 4 years and drafted players have continued to perform better just barely passing franchised last year.

I'm not sure I know what this all means, but I figured I'd share.