Kante
PREMIER
Below are the 2018-19 vs 2017-18 defensive improvements or declines by team this year (Sept thru Dec) vs last year.
As an fyi, last year San Diego and Los Angeles were one group. This year the two areas are separate groups. To normalize the comparison, I only took 2017-18 data from SD vs SD matches and LA vs LA matches as the baseline for the comparison to 2018-19.
Am using an algorithm called Goals Allowed % (GAP) to make the comparison between teams. This metric enables a mostly accurate look at how teams are likely to perform in future matches. Have been using to create match result predictions for 06, 05 and 04 matches and the algo has been getting pretty accurate.
If anyone has questions or wants more info on these types of metrics, here’s a post which explains in more detail.
http://www.socalsoccer.com/threads/how-to-objectively-best-measure-development.15976/
Here’s teams sorted by most to least improved defensively in 2018-19 vs 2017-18. Sample size is about ten matches for most teams.
Here's a look at which of these teams added new coaches for 2018-19. This info is from game cards and direct messages and may be incorrect. If you have a correction, please add to comment or send over via direct message.
Here’s a look at the correlation between teams adding new players and defensive improvement/decline.
As an fyi, the measure being used to quantify adding new player is number of games each new player (new = added to the roster in 2018-19) has been rostered on the 2018-19 game cards as a % of total players rostered x total number games played.
So, if a player was added at the beginning of the season and there are a total of 18 players on the roster, that new player would take up about 6% of the total game roster spots.
However, if that team has played 10 games, and the new player was added only for the 10th game, that new player would take up about .5% of the total game roster spots.
(Apologies for the word problem!)
Excluding both Albion and LAUFA (both teams had to backfill a # of key defensive players they lost to the SD Surf and LA Galaxy respectively), adding new players drove defensive about 70% of the defensive improvement y2y. Basically, for every new player added at the beginning of the season, it is expected that most teams improved defensively by about -20% goals allowed vs how they did in 2017-18.
Of note, Arsenal improved their defense significantly more than would have been expected, and Pateadores and Santa Barbara SC did not improve as much would have been expected for the # of players they added. One caveat is that the Pats lost a key defender to the Strikers in 2018-19.
Also of note, FC Golden State added four new players to their roster for their last game before the break, and it would be reasonable to expect that they will improve even more defensively in 2019.
As an fyi, last year San Diego and Los Angeles were one group. This year the two areas are separate groups. To normalize the comparison, I only took 2017-18 data from SD vs SD matches and LA vs LA matches as the baseline for the comparison to 2018-19.
Am using an algorithm called Goals Allowed % (GAP) to make the comparison between teams. This metric enables a mostly accurate look at how teams are likely to perform in future matches. Have been using to create match result predictions for 06, 05 and 04 matches and the algo has been getting pretty accurate.
If anyone has questions or wants more info on these types of metrics, here’s a post which explains in more detail.
http://www.socalsoccer.com/threads/how-to-objectively-best-measure-development.15976/
Here’s teams sorted by most to least improved defensively in 2018-19 vs 2017-18. Sample size is about ten matches for most teams.
Here's a look at which of these teams added new coaches for 2018-19. This info is from game cards and direct messages and may be incorrect. If you have a correction, please add to comment or send over via direct message.
Here’s a look at the correlation between teams adding new players and defensive improvement/decline.
As an fyi, the measure being used to quantify adding new player is number of games each new player (new = added to the roster in 2018-19) has been rostered on the 2018-19 game cards as a % of total players rostered x total number games played.
So, if a player was added at the beginning of the season and there are a total of 18 players on the roster, that new player would take up about 6% of the total game roster spots.
However, if that team has played 10 games, and the new player was added only for the 10th game, that new player would take up about .5% of the total game roster spots.
(Apologies for the word problem!)
Excluding both Albion and LAUFA (both teams had to backfill a # of key defensive players they lost to the SD Surf and LA Galaxy respectively), adding new players drove defensive about 70% of the defensive improvement y2y. Basically, for every new player added at the beginning of the season, it is expected that most teams improved defensively by about -20% goals allowed vs how they did in 2017-18.
Of note, Arsenal improved their defense significantly more than would have been expected, and Pateadores and Santa Barbara SC did not improve as much would have been expected for the # of players they added. One caveat is that the Pats lost a key defender to the Strikers in 2018-19.
Also of note, FC Golden State added four new players to their roster for their last game before the break, and it would be reasonable to expect that they will improve even more defensively in 2019.
Last edited: