2018 Roster Discussion

Interesting graph here

DlETf1XW4AA-AE6.jpg:large
NYCFC has a fundamental problem with their roster creation & subsequent playing minutes distribution. In medical terms, the club is living a very unhealthy lifestyle with a family history of high blood pressure while refusing to forgo fatty/rich food.
 
Here is the source of the graph:


Later in the thread he indicates that USL means a player from a USL affiliate.

Seems like kind of a misleading approach he clearly uses % to hide the fact that he's created very narrow definitions for each category and if a player does not meet the definitions he's just excluding them so some teams have all of their min accounted for while others have large numbers of players missing. Ibeagha is the perfect example, he should either be USL because he came from SA our affiliate or free agent because that's how we signed him. So we can assume USL only means a player directly acquired from an affiliate while still under contract, FA only means MLS free agent, etc. This approach makes the graph completely misleading because he's clearly excluding at a minimum, Sweat, Ibeagha and TMac from our numbers because he didn't make a category that fits them.

If he wants to be even reasonably accurate he needs an "other" category and dump any min that don't meet his categories into there. As it stands he's essentially saying this is the % break down for 70% of NYCFCs min compared with percentages to 100% of some other teams min but I'm not going to show that in any way easier to pretend it's all apples to apples. Sorry for the rant but crap analytics that are misleading due to laziness annoy the hell out of me.
 
Seems like kind of a misleading approach he clearly uses % to hide the fact that he's created very narrow definitions for each category and if a player does not meet the definitions he's just excluding them so some teams have all of their min accounted for while others have large numbers of players missing. Ibeagha is the perfect example, he should either be USL because he came from SA our affiliate or free agent because that's how we signed him. So we can assume USL only means a player directly acquired from an affiliate while still under contract, FA only means MLS free agent, etc. This approach makes the graph completely misleading because he's clearly excluding at a minimum, Sweat, Ibeagha and TMac from our numbers because he didn't make a category that fits them.

If he wants to be even reasonably accurate he needs an "other" category and dump any min that don't meet his categories into there. As it stands he's essentially saying this is the % break down for 70% of NYCFCs min compared with percentages to 100% of some other teams min but I'm not going to show that in any way easier to pretend it's all apples to apples. Sorry for the rant but crap analytics that are misleading due to laziness annoy the hell out of me.
I wonder why Tmac doesn't show up in the "expansion draft" category
 
  • Like
Reactions: Kjbert and gbservis
I wonder why Tmac doesn't show up in the "expansion draft" category

It's a good question, I just read his thread someone called him on missing Ibeagha. The problem is for players like Ibeagha and TMac he either has them in the wrong category or he excluded them in which case the bar should not total 100% for NYCFC.

Edit: I thought for a second maybe it was only players acquired this season but in his examples he states that Michael Bradley is in the INT bucket because he was acquired from abroad as he was playing for Roma at the time. So clearly he's attempting to bucket all players if he's going back that far.
 
I wouldn’t worry about Haak. If he wants to play here, he will.
 
I was fully expecting Lewis to be number one based on stats with minimal minutes. He didn't make the list. Minimum 500 minutes. Lewis has less than 200.
I think the writer also allowed himself some subjectivity, so even if there were no minute minimum, I doubt he would rank someone #1 based on so little playing time.
 
I think the writer also allowed himself some subjectivity, so even if there were no minute minimum, I doubt he would rank someone #1 based on so little playing time.
How can an objective list include subjective reasoning? Doesn’t A equal A and B equal B and if A>B then it is what it is?
 
How can an objective list include subjective reasoning? Doesn’t A equal A and B equal B and if A>B then it is what it is?
"This is mostly an objective look at the discussion through the lens of data with a slightly subjective twist of how I see things from tactical perspective. A lot of this would match my own personal eye test, but a few certainly surprised me."

Seems like a reasonable approach to me.

ETA: I think he basically went with the numbers but allowed himself some discretion on the margins, probably meaning if it was close he might flip 2 players.