Kante
PREMIER
Apologies for the length of the post and congrats to folks who make it all the way through. Below are the 2018-19 vs 2017-18 offensive comparative improvements or declines by team ytd this year (Sept thru Dec) vs same period last year.
Am using a algorithm called Goal Scoring % (GSP) 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 offensively in 2018-19 vs 2017-18, for matches played Sept 1 thru Dec 9. Sample size is about ten matches for most teams.
Here’s which teams added what # of players for the new 2018-19 season.
When you look at both charts – offensive improvement and # of players added – there’s an interesting inverse relationship between the # of players being added to a team and to the degree to which they improved (or did not improve) offensively y2y, particularly if you remove LA Galaxy and LAFC as anomalies.
Would argue that LAG is an anomaly because of the degree to which they replaced their 2017-18 team by recruiting players from other academies (most teams don’t have this kind of recruiting muscle). LAFC is an anomaly because of the drop-off after their 12 month run of wins.
Here’s a chart comparing the # of players added to offensive improvement/decline without LA Galaxy and LAFC. It’s a little busy, but the short version is that there’s a decent correlation between adding a new player and a decline in offensive effectiveness the more players a team adds.
Running the stats, the correlation shows that the adding new players drives about 30% of the variation in y2y offensive effectiveness improvement/decline, and could be the driver behind an average -13% decline in offensive effectiveness for each new player added.
Put more simply, according to the numbers, maintaining and developing the bulk of a team's players over time looks like it is more be a more effective way to improve your offense than wholesale recruiting/adding in new players.
Intuitively, this makes sense, and there's some numbers to back this up.
Am using a algorithm called Goal Scoring % (GSP) 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 offensively in 2018-19 vs 2017-18, for matches played Sept 1 thru Dec 9. Sample size is about ten matches for most teams.
Here’s which teams added what # of players for the new 2018-19 season.
When you look at both charts – offensive improvement and # of players added – there’s an interesting inverse relationship between the # of players being added to a team and to the degree to which they improved (or did not improve) offensively y2y, particularly if you remove LA Galaxy and LAFC as anomalies.
Would argue that LAG is an anomaly because of the degree to which they replaced their 2017-18 team by recruiting players from other academies (most teams don’t have this kind of recruiting muscle). LAFC is an anomaly because of the drop-off after their 12 month run of wins.
Here’s a chart comparing the # of players added to offensive improvement/decline without LA Galaxy and LAFC. It’s a little busy, but the short version is that there’s a decent correlation between adding a new player and a decline in offensive effectiveness the more players a team adds.
Running the stats, the correlation shows that the adding new players drives about 30% of the variation in y2y offensive effectiveness improvement/decline, and could be the driver behind an average -13% decline in offensive effectiveness for each new player added.
Put more simply, according to the numbers, maintaining and developing the bulk of a team's players over time looks like it is more be a more effective way to improve your offense than wholesale recruiting/adding in new players.
Intuitively, this makes sense, and there's some numbers to back this up.