In Part I of the league discussion we did a review of when players are being used. And in Part II we discussed how players are signed, whether via the draft, amateur free agent (aFA) or others. Now we will discern where the amateur free agents are coming from.
Please remember that for this data analysis, the aFA identified do not necessarily have to come up through that organization. If that player is traded to another organization, he is still qualified as an aFA because that is how they were originally signed. From Part II, the amateur free agents are approximately are usually 20-25% of the all the players. This analysis will attempt to identify the markets where those amateur free agents are coming from. Going into this, my best guess would be that they mostly come from the Caribbean and upper South America locales, especially since this is where major league teams tend to setup their baseball academies.
And as previously, as the Tribe players themselves are only about 3.3% of the overall total, so looking at 25% of that total is open to even more fluctuation. I will continue to provide the median values for each category for completeness sake, but the graphs themselves are mostly meaningless, and the Tribe graphs are not included.
One last note about the amateur free agent, I have decided to split them up into five geographical areas: the Caribbean, Central/South America, North America, the Far East, and the Rest of World (ROW). The Caribbean includes Aruba, Cuba, Curacao, the Dominican Republic and Puerto Rico. Up until 1989, Puerto Rican players were allowed to sign as free agents. And usually about 80% of all Caribbean amateur free agents come from the Dominican Republic.
Central/South America includes Columbia, Nicaragua, Panama and Venezuela. Approximately 85% of all these amateur free agents come from Venezuela. North America includes Canada, Mexico and the USA. The Far East includes Japan, South Korea and Taiwan. This does not include established players that went through the posting process. And last but not least, the ROW includes Australia and the Netherlands.
So let’s start with all amateur free agents who appeared in a game each season.
Fig 1.1 - All Players (Core + Role + Backup + Coffee) and (Hitters + Pitchers): 
The Caribbean saturates this total, hovering right around 50% over the cycle. But there is a definite rise of the C/S American players, most notably in Venezuela. There has also been a slight rise of ROW players. Median values over the 12 year cycle: Caribbean 49.66%, C/S America 27.47%; N America: 16.21%; Far East: 2.67%; ROW: 2.13%.
The Tribe players leaned more heavily on C/S American players and a bit more on Far East players as well. Median values over the 12 year cycle: Caribbean 42.27%, C/S America 37.98%; N America: 11.86%; Far East: 7.69%; ROW: 0.00%.
Now let’s see if splitting up the hitters and pitchers changes anything.
Fig 1.2 - All Hitters (Core + Role + Backup + Coffee): 
Interestingly enough, the C/S American players have steadily risen and just passed the Caribbean as the majority while the other three are very consistent. Median values over the 12 year cycle: Caribbean 50.01%, C/S America 36.03%; N America: 11.47%; Far East: 1.46%; ROW: 1.43%.
The Tribe hitters have varied from N America to C/S America to the Caribbean with more of a lean to C/S America. Median values over the 12 year cycle: Caribbean 34.29%, C/S America 50.00%; N America: 6.25%; Far East: 6.25%; ROW: 0.00%.
Fig 1.3 - All Pitchers (Core + Role + Backup + Coffee): 
There is a clear separation here in favor for the Caribbean. Interestingly enough, the N American players out produced the C/S America players early in the decade, but have been passed up. Median values over the 12 year cycle: Caribbean 50.49%, C/S America 19.28%; N America: 20.92%; Far East: 3.86%; ROW: 3.46%.
The Tribe pitcher graph is all over the place, with a single ROW pitcher in 2000. Median values over the 12 year cycle: Caribbean 45.00%, C/S America 26.79%; N America: 16.67%; Far East: 0.00%; ROW: 0.00%.
As before, we’ll split up the groups into Core, Role, Backup and Coffee players. Core players got 500 or more plate appearances, made 25 or more starts or pitched 60 innings or more in relief. Role players got 300-499 plate appearances, made 15-24 starts or pitched 30-59.2 innings in relief. Backup players got 100-299 plate appearances, made 5-14 starts or pitched 15-29.2 innings in relief. And Coffee players got less than 100 plate appearances, made less than 5 starts or pitched less than 15 innings in relief. For combination pitchers (starts and relief appearances) I subtracted 6 innings pitched per start from the total innings and if the resultant number bumped them another grade, I moved them up.
Fig 2.1 - All Core Players (Hitters + Pitchers): 
The Core players are dominated from the Caribbean and C/South America with about 85% of the total. But C/S America is slowly gaining over the cycle. Median values over the 12 year cycle: Caribbean 55.15%, C/S America 30.74%; N America: 11.27%; Far East: 1.40%; ROW: 1.22%.
The Tribe Core players that are amateur free agents have very small totals, varying from one to maybe 4 per year, so these numbers look a bit scattered. Median values over the 12 year cycle: Caribbean 63.34%, C/S America 33.33%; N America: 0.00%; Far East: 0.00%; ROW: 0.00%.
Fig 2.2 - All Core Hitters: 
The Core hitters track the overall total fairly closely and even had one season with C/S America passing the Caribbean in 2010. Additionally, over 95% of the core hitters come from the Caribbean and C/S America. Median values over the 12 year cycle: Caribbean 60.40%, C/S America 37.75%; N America: 2.86%; Far East: 0.00%; ROW: 0.00%.
As above, splitting the few core players into hitters and pitchers is slicing the data awfully thin. There was one year with 5 core hitters (2007) and only two with 4 (2011 and 2001). Median values over the 12 year cycle: Caribbean 63.34%, C/S America 33.33%; N America: 0.00%; Far East: 0.00%; ROW: 0.00%.
Fig 2.3 - All Core Pitchers: 
The Core pitchers lean even more heavily on the Caribbean while the Far East has a bit of jump (which is expected). Median values over the 12 year cycle: Caribbean 51.86%, C/S America 25.00%; N America: 17.84%; Far East: 2.47%; ROW: 2.15%.
The Tribe Core pitchers are even fewer than the hitters. In fact in 2006 and 2009 the Tribe did not have a single core pitcher. Another oddity is that while 83.34% of all pitchers were Caribbean, only 25% of them were from the Dominican Republic. Median values over the 12 year cycle: Caribbean 83.34%, C/S America 0.00%; N America: 0.00%; Far East: 0.00%; ROW: 0.00%.
Fig 3.1 - All Role Players (Hitters + Pitchers): 
The Role players also have the C/S America increase in the recent part of the decade. The Role players are using the N America area much less and there looks like two bumps of Far East players around 2004 and 2008. Median values over the 12 year cycle: Caribbean 50.43%, C/S America 26.16%; N America: 18.03%; Far East: 1.69%; ROW: 1.84%.
The Tribe Role players again stick to mainly the Caribbean and C/S America, with C/S America more dominant. Median values over the 12 year cycle: Caribbean 33.33%, C/S America 50.00%; N America: 0.00%; Far East: 0.00%; ROW: 0.00%.
Fig 3.2 - All Role Hitters: 
The Role hitters again stick mainly with the Caribbean and C/S America. The N America group has been in steady decline. . Median values over the 12 year cycle: Caribbean 49.25%, C/S America 41.31%; N America: 10.44%; Far East: 0.00%; ROW: 0.00%.
The Tribe Role hitters are few and far between. There were no amateur role hitters from 2003 to 2006. Median values over the 12 year cycle: Caribbean 0.00%, C/S America 50.00%; N America: 0.00%; Far East: 0.00%; ROW: 0.00%.
Fig 3.3 - All Role Pitchers: 
The Role pitchers show much more variance than any of the previous graphs, with the highest reliance on the Caribbean so far. Median values over the 12 year cycle: Caribbean 56.83%, C/S America 18.83%; N America: 20.59%; Far East: 2.64%; ROW: 2.94%.
The Tribe Role pitchers had two years of blanks (2005 and 2007) but almost exclusively Caribbean overall. Median values over the 12 year cycle: Caribbean 41.67%, C/S America 0.00%; N America: 0.00%; Far East: 0.00%; ROW: 0.00%.
Fig 4.1 - All Backup Players (Hitters + Pitchers): 
The Backup continue to show the recent trend of more C/S American players with a steady decline of N American players. The Far East ad it’s spike around 2007. Median values over the 12 year cycle: Caribbean 47.33%, C/S America 27.97%; N America: 18.14%; Far East: 2.40%; ROW: 1.67%.
The Tribe Backup players again had much more variance, but never had a zero value like the role players. Median values over the 12 year cycle: Caribbean 36.67%, C/S America 36.67%; N America: 18.34%; Far East: 0.00%; ROW: 0.00%.
Fig 4.2 - All Backup Hitters: 
The Backup hitters have a high reliance of Caribbean players, but the C/S American players have made high strides, mainly to the N American players’ expense. Median values over the 12 year cycle: Caribbean 49.29%, C/S America 33.36%; N America: 15.13%; Far East: 0.00%; ROW: 0.00%.
The Tribe Backup hitters continued to jump from the Caribbean to C/S America on an almost yearly basis. Median values over the 12 year cycle: Caribbean 45.00%, C/S America 36.67%; N America: 0.00%; Far East: 0.00%; ROW: 0.00%.
Fig 4.3 - All Backup Pitchers: 
While the Backup pitchers continue the trend overall, here we see a higher reliance on the N American and Far East players, especially in 2007 and 2008. Median values over the 12 year cycle: Caribbean 49.24%, C/S America 20.53%; N America: 18.59%; Far East: 5.73%; ROW: 3.13%.
The Tribe Backup pitchers are few and far between overall with zeroes in 2000, 2005 and 2007. In fact, not a single grouping had a positive median score. Caribbean 0.00%, C/S America 0.00%; N America: 0.00%; Far East: 0.00%; ROW: 0.00%.
Fig 5.1 - All Coffee Players (Hitters + Pitchers): 
The Coffee players actually are one of the most stable of all the graphs. The N American players are very steady at around 20% and in 2011 we see the first spike of ROW, up to 8.99%. Median values over the 12 year cycle: Caribbean 48.59%, C/S America 27.89%; N America: 20.00%; Far East: 2.13%; ROW: 4.23%.
The Tribe Coffee players again have a huge variance, but not a single ROW player. Median values over the 12 year cycle: Caribbean 33.33%, C/S America 33.33%; N America: 20.84%; Far East: 0.00%; ROW: 0.00%.
Fig 5.2 - All Coffee Hitters: 
The Coffee hitters are almost exclusively the main two throughout this discussion. But it is interesting to see the N American spike from 2005 to 2008. Median values over the 12 year cycle: Caribbean 43.55%, C/S America 32.92%; N America: 14.48%; Far East: 2.64%; ROW: 3.39%.
The Tribe Coffee hitters were of a very low number, with zeroes in 2003 and 2006. Median values over the 12 year cycle: Caribbean 0.00%, C/S America 16.67%; N America: 0.00%; Far East: 0.00%; ROW: 0.00%.
Fig 5.3 - All Coffee Pitchers: 
The Coffee pitchers have the highest data for N American players, consistently even above C/S America. The Far East and ROW groups also have much higher totals than most of the graphs above. Median values over the 12 year cycle: Caribbean 45.47%, C/S America 16.03%; N America: 25.55%; Far East: 4.06%; ROW: 5.21%.
The Tribe Coffee pitchers were used pretty minimally, with zeroes in 2001, 2008, 2010 and 2011. Median values over the 12 year cycle: Caribbean 16.67%, C/S America 0.00%; N America: 0.00%; Far East: 0.00%; ROW: 0.00%.
Summary
Without looking at the data, I would have thought there would be much more fluctuation as the amateur free agents only make up about 25% of the league total. But, these numbers were more stable than I originally thought.
The dominance of the Caribbean, especially the Dominican Republic, had not wavered all that much in the past decade plus. There has been a slight decline in the past five years. But the increase in players from Central and South America, especially Venezuela, have been more at the expense of the North American amateur free agents.
Again, The Tribe numbers, while interesting, do not accurately reflect how the team is truly made up. This would need to be done in a separate computation (which could occur in a later post if there is enough demand).
Coming in Part IV, we finally attack the draft. My assumption is this might be the most anticipated piece to date.
I also have data corresponding to the team’s that originally signed the players. But with 30 teams, the graphs get a bit messy, so if there is enough demand, I could split Part V into perhaps six smaller (divisional) bite-size chunks. Or I could have six graphs per grouping and split the post into five separate smaller chunks.