July P Score: Reactive Games Lead to Fewer Goals / by Jared Young

This is our monthly P score update. Click here for last month's article. To read how P score is calculated, click here.

By Jared Young (@jaredeyoung)

Change was good for some teams in July and not for others. FC Dallas and Colorado Rapids had perfect months while playing more proactively than earlier in the season. Meanwhile, Orlando City SC and Seattle Sounders had difficult months and were more reactive than they had ever been.

Below is the latest P Score table. Orlando still maintains their top spot, but Montreal caught up by being the most proactive team in July. This month I added each team’s opponent's P Score. Interestingly but not surprisingly, teams play more reactively against the New York Red Bulls and Sporting Kansas City.

RankLastTeamP ScoreOpp P ScoreDifferenceYear PPGJuly P ScoreJuly PPG
11ORL7.1 6.2 1 1.144.80.3
23MON6.4 5.5 0.9 1.337.51.5
34NYRB6.1 3.5 2.6 1.536.72
42SEA5.9 5.6 0.2 1.454.30.8
57LA5.8 5.3 0.5 1.467.32
66CLB5.8 5.6 0.2 1.395.82
79POR5.7 5 0.7 1.456.51
85TOR5.5 5.8 -0.3 1.474.31.3
912NYCFC5.5 5.1 0.4 1.146.31.8
1010PHI5.4 6 -0.6 15.71
118DC5.4 5.3 0.1 1.6541
1214NE5.1 6.1 -1 1.2251
1311HOU5 4.9 0.1 1.293.81.8
1413CHI4.8 5.6 -0.8 0.9531
1517DAL4.7 5.5 -0.8 1.8163
1615VAN4.7 5.7 -1 1.643.81
1719COL4.4 4.7 -0.3 1.253
1816SKC4.4 3.9 0.5 1.742.72
1920RSL4 4.5 -0.5 1.324.51.8
2018SJ3.9 5.5 -1.6 1.2520

The fate of Dallas, Colorado, Orlando and Seattle got me wondering if there was more than a trend here. So I plotted the change in P Score against the change in points per game. Sure enough, the more reactive a team was relative to their play all season long, the more they lowered their points per game in July.

That’s a trend to keep watching. Should teams be pushing to be more proactive as the season progresses? Is that an actual thing?

This month’s P Score bonus analysis

This month I was curious if the styles of play impact total goal scoring. For example, if two teams are focused on shape and direct play and therefore playing very reactively, are less goals scored in those games? And conversely, if two teams are playing very proactively, are more goals scored?

To get to this I added the P Score of both teams and looked at the average goals scored in each grouping. I grouped total P Score 1 to 9, 10 to 11, and 12+ which was as close as I could get to an even number of games in each bucket. Sure enough, reactive games have over 10% fewer goals scored than proactive games.

The sample sizes get small but at the lower extreme the trend continues. There were only 10 games with a combined P Score of 5 or lower but only 1.4 goals were scored in those games. At the higher end of the spectrum the trend is basically flat. After combined 12 P Score the goals scored are roughly 2.7 per game.

It’s not a huge difference but the P Score does appear to be capturing a team’s style of play and how that impacts the flow and scoring in games. The impacts are not large but they are present and this season, finding any rhyme or reason for a team’s success is a huge challenge. That’s a topic that I am going to cover in depth for the August review of P Score. Until then here’s a random ambient chart plotting the percentage of long passes played in each game in MLS this season.