MLS Points Per Match Table

by Dave Clark on May 8, 2011 8:00 AM PDT in Sounders: MLS Regular Season

Major League Soccer can do some odd things with the schedules. Sometimes a team or two get adjusted due to a CONCACAF Champions League Final. Other times you will see a team or two start with extra road games. Seven teams have already played nine games (this was week 8), and one has already played 10. Yes, that's the Seattle Sounders with two double weeks and so nearly a third through the season, even though this is the quarter pole on average.

While some would use games in hand as a balance, this often leads to the assumption that a poor performing team has the chance to collect points as if they didn't already suck. Take Sporting Kansas City for example. They have yet to perform well, should we really assume that the 4 games in hand that they have on Seattle will all be wins? Probably not, though they do have the advantage that they will have extra home games late in the year, and by a lot.

Here we tend to like Points Per Match. Jeremiah is throwing together a sidebar widget so you can look at that at any time (and you really should check out the sidebars we have a ton of data there). But right now, at this quarter pole for the league (again on average) is a chart of the league by PPM. Also included is "form" by last five matches and which teams would make the playoffs by PPM right now. Blue being the Western Conference and Red for the East. The Wild Cards are in Grey.

Club	PPM	PTS	GP	W	L	T	GF	GA	GD	Last 5
Real Salt Lake	2.50	15	6	5	1	0	9	2	7	12
New York Red Bulls	2.00	14	7	4	1	2	10	2	8	10
Philadelphia Union	1.86	13	7	4	2	1	5	3	2	7
Los Angeles Galaxy	1.67	15	9	4	2	3	11	9	2	8
Columbus Crew	1.63	13	8	3	1	4	8	6	2	9
Portland Timbers	1.63	13	8	4	3	1	11	13	-2	12
Colorado Rapids	1.56	14	9	4	3	2	11	9	2	5
FC Dallas	1.38	11	8	3	3	2	10	10	0	10
Houston Dynamo	1.33	12	9	3	3	3	13	10	3	7
Seattle Sounders FC	1.30	13	10	3	3	4	12	10	2	8
DC United	1.22	11	9	3	4	2	12	17	-5	7
Chivas USA	1.13	9	8	2	3	3	8	7	1	8
New England Revolution	1.11	10	9	2	3	4	8	12	-4	4
Toronto FC	1.11	10	9	2	3	4	9	14	-5	5
Chicago Fire	0.88	7	8	1	3	4	10	13	-3	3
Vancouver Whitecaps	0.78	7	9	1	4	4	11	14	-3	2
San Jose Earthquakes	0.71	5	7	1	4	2	6	10	-4	2
Sporting Kansas City	0.67	4	6	1	4	1	10	13	-3	1

One should note that only six teams have a goal differential greater than +/- 3, and more of those are poor than good. There is a lot of parity in the league as always, but two teams have risen above the rest. Some of the bad goal differential teams have offset their blowouts with a few narrow wins, most notably the Portland Timbers who would have a protected Playoff berth by PPM with a negative goal differential.

Tweet Comment 8 comments | 0 recs |

Do you like this story?

Comments

well were in 10th.. thats a playoff spot!

by Sandra_R on May 8, 2011 8:17 AM PDT reply actions

I choose to not recognize this table since the timbers are above us on it.

by Robert on May 8, 2011 8:53 AM PDT reply actions

I think it should be taped to the halls of Quest

As fuel for this weekend’s game.

by Dizzo on May 8, 2011 8:57 AM PDT up reply actions 1 recs

If we win this weekend

We’ll be ahead by a hair

by Tohoya on May 8, 2011 9:55 AM PDT up reply actions

Hmmm...

Take Sporting Kansas City for example. They have yet to perform well, should we really assume that the 4 games in hand that they have on Seattle will all be wins?

Certainly I wouldn’t presume any future games to be wins for any teams in the league, but at this point in the season, it’s not unreasonable to expect SKC to be league average in their games-in-hand, especially considering that they’ve had all 6 games on the road, and last year home teams in MLS had 1.69 PPM and road teams had 1.07 PPM. With 6 road games, we’d expect a league average team to have 6-7 points, and SKC has had 4 points. They’re not even close to being statistically different from average.

Since home field advantage matters so much in MLS, I’d suggest that it would be best to borrow from golf and use points above/below par. Golf’s system, when par is chosen close to average for the field, allows viewers to compare players that are on different parts of the course, and golfers who have played sections of the course that are of differing difficulty. All that’s needed to effectively compare is the golfer’s score above/below par and the number of holes he has played, the number of holes he has played giving us an idea of how much upside/downside is left for the golfer.

Pts over Par - Team ( GP )
6.72 - RSL ( 6 )
3.96 - NYRB ( 8 )
3.03 - PU ( 7 )
2.82 - LAG ( 10 )
1.96 - PT ( 8 )
1.89 - CR ( 9 )
1.34 - CC ( 8 )
-0.04 - FCD ( 8 )
-0.18 - SFC ( 10 )
-0.73 - HD ( 9 )
-1.73 - DCU ( 9 )
-2.04 - CUSA ( 8 )
-2.42 - SKC ( 6 )
-2.73 - NER ( 9 )
-3.35 - TFC ( 9 )
-4.04 - CF ( 8 )
-4.97 - SJE ( 7 )
-5.73 - VW ( 9 )

The rankings don’t change a whole lot, except for SKC, which has had a perverse schedule so far. I guess I think a big advantage to this over PPM is that it’s easier to figure how the outcome of the next game will change your score above par than it is to figure out how the outcome of the next game changes your PPM.

by ubelmann on May 8, 2011 12:04 PM PDT reply actions

I like this a lot

Though I don’t think it should replace ppm. Put something like an above/below par stat alongside it, with an option to sort by that, points, or ppm. And when determining par, I think it might be best to get a few more years of data.

I wonder if it would be possible to create a computer ranking/prediction system with this ranking system and a strength of schedule based on, say, SBNation’s power rankings?

by Tohoya on May 8, 2011 5:05 PM PDT up reply actions