San Jose v Seattle: Local Player Showdown
What if San Jose and Seattle played each other using only players from their respective areas rather than players from all of the country and indeed the world? I thought I'd try to explore this using R and MLS Fantasy points (as determined prior to tonight's match between these two MLS teams).
I decided on a 20 mile radius from each team's home stadium as the limits of the "local area" from which the teams could draw their players. Seattle has 5 players with hometowns within 20 miles of CenturyLink Field, whereas San Jose has 4 players with hometowns within 20 miles of Avaya Stadium. The cumulative totals of MLS Fantasy points (through Week 5) for this 5 v 4 match-up produced the following result:
Despite having fewer players, San Jose came out on top, doubling Seattle's points total. However, these two teams are too small to play a match, and only one of them (San Jose) has a goalkeeper. I decided to expand the two cities' respective radii to the points at which each city could field at least 6 field players and 1 goalkeeper (7 being the minimum number of players required to play a match according to FIFA's Laws of the Game).
For Seattle to field at least 6 field players and a goalkeeper, the radius around CenturyLink Field has to increase from 20 miles to just over 23 miles. At that point, Seattle picks up three more field players, all of whom are from the same hometown (Federal Way). By contrast, the radius around Avaya Stadium has to increase from 20 miles to just over 38.5 miles for San Jose to pick up 6 field players and a goalkeeper. In fact, at that distance, San Jose picks up two goalkeepers, but since teams can field only one, I dropped the more distant goalkeeper from San Jose's team. This left the 8 players for Seattle pitted against the 7 players for San Jose, with the following outcome:
Again, despite having fewer players, San Jose crushes Seattle. Even if San Jose were limited to the same 23 mile radius that Seattle had, San Jose still comes out on top:
At this radius, San Jose picks up no additional players but still has more points than Seattle (although no goalkeeper).
For completeness, only one player's hometown is San Jose, CA (Steven Beitashour), while Seattle is the hometown of three players (Seyi Adekoya, Aaron Kovar, and Spencer Richey). Of these players, two actually play for their hometown MLS team (Seyi Adekoya and Aaron Kovar). For these four players, their cumulative MLS Fantasy scores (through Week 5) for their respective hometowns are:
So, despite being outnumbered 1 against 3, San Jose takes the win again, and by quite a margin.
This exercise involved my first use of the mapdist function from the ggmap package. mapdist takes a "from" and "to" argument and produces a variety of information by querying Google, including distance and time of travel. For this exercise, I was interested only in miles. The "from" and "to" arguments can even be place names. Here is an excerpt from the code that I used for this project
I had a bit of trouble making it work, but the "as.character" function seemed to solve it.
I decided on a 20 mile radius from each team's home stadium as the limits of the "local area" from which the teams could draw their players. Seattle has 5 players with hometowns within 20 miles of CenturyLink Field, whereas San Jose has 4 players with hometowns within 20 miles of Avaya Stadium. The cumulative totals of MLS Fantasy points (through Week 5) for this 5 v 4 match-up produced the following result:
San Jose Seattle 42 21
Despite having fewer players, San Jose came out on top, doubling Seattle's points total. However, these two teams are too small to play a match, and only one of them (San Jose) has a goalkeeper. I decided to expand the two cities' respective radii to the points at which each city could field at least 6 field players and 1 goalkeeper (7 being the minimum number of players required to play a match according to FIFA's Laws of the Game).
For Seattle to field at least 6 field players and a goalkeeper, the radius around CenturyLink Field has to increase from 20 miles to just over 23 miles. At that point, Seattle picks up three more field players, all of whom are from the same hometown (Federal Way). By contrast, the radius around Avaya Stadium has to increase from 20 miles to just over 38.5 miles for San Jose to pick up 6 field players and a goalkeeper. In fact, at that distance, San Jose picks up two goalkeepers, but since teams can field only one, I dropped the more distant goalkeeper from San Jose's team. This left the 8 players for Seattle pitted against the 7 players for San Jose, with the following outcome:
San Jose Seattle 108 29
Again, despite having fewer players, San Jose crushes Seattle. Even if San Jose were limited to the same 23 mile radius that Seattle had, San Jose still comes out on top:
San Jose Seattle 42 29
At this radius, San Jose picks up no additional players but still has more points than Seattle (although no goalkeeper).
For completeness, only one player's hometown is San Jose, CA (Steven Beitashour), while Seattle is the hometown of three players (Seyi Adekoya, Aaron Kovar, and Spencer Richey). Of these players, two actually play for their hometown MLS team (Seyi Adekoya and Aaron Kovar). For these four players, their cumulative MLS Fantasy scores (through Week 5) for their respective hometowns are:
San Jose Seattle 21 2
So, despite being outnumbered 1 against 3, San Jose takes the win again, and by quite a margin.
This exercise involved my first use of the mapdist function from the ggmap package. mapdist takes a "from" and "to" argument and produces a variety of information by querying Google, including distance and time of travel. For this exercise, I was interested only in miles. The "from" and "to" arguments can even be place names. Here is an excerpt from the code that I used for this project
> SEA_geo<-paste(SEA$Home.Town, SEA$Home.State) > SEA_u<-as.character(SEA_geo) > SEA_dist<-mapdist(SEA_geo, "CenturyLink Field")$miles
I had a bit of trouble making it work, but the "as.character" function seemed to solve it.
Comments
Post a Comment