Centre of the Box

Return of the Blog

noreply@blogger.com (Bill Fish) — Fri, 09 Sep 2011 17:51:00 +0000

The blog has been on hiatus for the past month. Life got hectic for a while. We now have a baby on the way, work's been keeping me quite busy lately and we have an upcoming vacation to Ireland so posts will be somewhat sparse. I have also been thinking about the content of the blog and plan on shifting my focus towards player stats and less on team stats and modeling. I plan on covering the Premier League and La Liga but will need a few more matches before I can publish any meaningful data. So please stay tuned!

MLS Week 21 Recap

noreply@blogger.com (Bill Fish) — Wed, 10 Aug 2011 19:27:00 +0000

Had a so-so week with my predictions as you can see below:

Predicted Score				Actual Score				Points
Colorado	1	Columbus	1	Colorado	2	Columbus	0	-10
D.C. United	2	Toronto FC	1	D.C. United	3	Toronto FC	3	-10
New England	1	Chivas USA	2	New England	2	Chivas USA	3	15
Philadelphia	1	Houston	0	Philadelphia	1	Houston	1	-10
Kansas City	1	Seattle	1	Kansas City	1	Seattle	2	-10
Real Salt Lake	2	New York	1	Real Salt Lake	3	New York	0	10
Los Angeles	1	FC Dallas	0	Los Angeles	3	FC Dallas	1	10
San Jose	2	Portland	1	San Jose	1	Portland	1	-10
Vancouver	1	Chicago	1	Vancouver	4	Chicago	2	-10
						Total		-25

My accuracy was 3 for 9 on the outcomes and I did not predict any scores correctly. My outcome accuracy is now 39.6% for the season. Let's see how the new SRS ratings look:

Team	MOV	SOS	SRS	Change
Real Salt Lake	0.667	-0.008	0.659	-0.011
Los Angeles	0.600	-0.043	0.557	-0.096
Seattle	0.333	-0.016	0.318	0.034
FC Dallas	0.292	0.015	0.307	-0.067
Philadelphia	0.318	-0.016	0.302	-0.034
Colorado	0.160	0.028	0.188	0.077
New York	0.167	0.020	0.186	-0.107
Chivas USA	0.174	-0.018	0.156	0.010
Columbus	0.087	0.037	0.124	-0.081
Houston	0.000	0.024	0.024	0.005
Kansas City	0.087	-0.065	0.022	0.102
D.C. United	-0.190	-0.010	-0.201	-0.057
San Jose	-0.217	0.009	-0.209	-0.024
Portland	-0.318	0.027	-0.291	0.183
Chicago	-0.318	0.020	-0.298	-0.079
New England	-0.478	0.041	-0.438	-0.017
Vancouver	-0.478	-0.039	-0.517	0.092
Toronto FC	-0.880	-0.011	-0.891	0.029

Both Real Salt Lake and Los Angeles won one and lost one last week. However, LA's loss to Portland hurt their SRS rating more than RSL's loss to Columbus due to Portland being the weaker team. Conversely, Portland's victory over LA allowed them to register the biggest gain in the ratings.

My Thoughts on the MLS League Structure

noreply@blogger.com (Bill Fish) — Wed, 10 Aug 2011 03:31:00 +0000

A tough weekend for the Eastern Conference with none of their teams winning. There is now a definite imbalance between the East and West. Columbus leads the East with only 34 points. This point total would be good for only 6th place in the Western conference. I'm now questioning why the MLS is split into conferences. The geographic alignment of "East" and "West" should encourage rivalries between conference foes, but since every team plays every other team in the league twice, teams actually play more games against the other conference (18) than against their own conference (16). Based on the way that teams are seeded in the current playoff structure the MLS Cup will likely feature a matchup of the best team in the league (LA) against the 5th or 6th-best team (whoever comes out of the East). This is not the marquee matchup that the league needs to generate buzz for the Cup final.

If the league goes to a single table structure, which every other major soccer league in the world plays, they can seed the playoff teams 1 - 10 strictly on record and increase the likelihood that the two best teams in the league will meet in the MLS Cup final. It will make the regular season more meaningful because teams will be seeded based on their performance rather than conference alignment. For example, Colorado currently has the fourth best record in the league. In a single-table league they would be seeded fourth. In the current structure, although they have the 4th-best record, the three teams ahead of them are in their conference. So Colorado has to step aside while the league allocates the next three spots to the top teams from the East, which all have a worse record than Colorado. So the 4th-best team is relegated to a wild-card slot and has to play an extra-match while the weaker teams from the other conference get a bye. This is certainly not fair.

With the Montreal joining the league next season, the MLS has a golden opportunity to change their league structure. Going to a single table will, in my opinion, be better for the league. Regular season games will be more meaningful because teams will know exactly what they need to do to finish in the top 10. The playoffs will become more meaningful because we will have matchups based on performance rather than conference alignment. This will lead to better matchups in the MLS Cup final which should generate more buzz for this championship.

MLS Week 21 Weekend Predictions

noreply@blogger.com (Bill Fish) — Fri, 05 Aug 2011 16:59:00 +0000

We have a full slate of games this weekend. Here are my predictions:

Predicted Score
Colorado	1	Columbus	1
D.C. United	2	Toronto FC	1
New England	1	Chivas USA	2
Philadelphia	1	Houston	0
Kansas City	1	Seattle	1
Real Salt Lake	2	New York	1
Los Angeles	1	FC Dallas	0
San Jose	2	Portland	1
Vancouver	1	Chicago	1

MLS Week 21 Mid-week Predictions

noreply@blogger.com (Bill Fish) — Wed, 03 Aug 2011 18:38:00 +0000

Even though last weekend's predictions bombed I'm staying in the game. For July I finished with an accuracy rate of 42.2% which raised my season average to 40.8%. Let's see how we do in August. We'll be kicking off the month with ~~four~~ three matches tonight. Here are my predictions:

Predicted Score
Kansas City	1	Real Salt Lake	1
Chicago	1	Philadelphia	1
Portland	0	Los Angeles	1

MLS Season Projection Week 20

noreply@blogger.com (Bill Fish) — Wed, 03 Aug 2011 17:59:00 +0000

After a weekend full of upsets we have quite a shake-up in the season projection. Here are the projected standings based on results through week 20:

East
Pos	Team	P	W	D	L	GF	GA	GD	Pts
1	Philadelphia	34	13	16	5	42	29	13	55
2	Columbus	34	12	14	8	37	29	8	50
3	New York	34	11	16	7	54	42	12	49
4	D.C. United	34	10	16	8	46	49	-3	46
5	Kansas City	34	8	19	7	45	41	4	43
6	Houston	34	9	13	12	39	40	-1	40
7	Chicago	34	3	20	11	35	44	-9	29
8	New England	34	4	15	15	30	45	-15	27
9	Toronto FC	34	3	13	18	30	60	-30	22

West
Pos	Team	P	W	D	L	GF	GA	GD	Pts
1	Los Angeles	34	18	14	2	47	25	22	68
2	Real Salt Lake	34	16	14	4	45	25	20	62
3	FC Dallas	34	16	12	6	43	30	13	60
4	Seattle	34	14	15	5	47	37	10	57
5	Colorado	34	10	17	7	46	42	4	47
6	Chivas USA	34	9	16	9	42	38	4	43
7	San Jose	34	6	15	13	38	47	-9	33
8	Portland	34	9	6	19	39	58	-19	33
9	Vancouver	34	3	13	18	32	52	-20	22

Los Angeles moves ahead of RSL for the Western Conference title and best record in the league. In the East, Columbus is now projected to finish 2nd with D.C. United moving up to 4th and Kansas City dropping out of the top three into 5th. I'm now projecting United to make the playoffs as a wild card along with KC. The MLS Cup final will be Los Angeles vs Philadelphia. Since it will be played at the Home Depot Center, it will effectively be a home game for the Galaxy so I project them to win the Cup.

MLS Week 20 Recap

noreply@blogger.com (Bill Fish) — Tue, 02 Aug 2011 21:57:00 +0000

Good News and bad news for me over the weekend. The good news is that I had a perfect weekend with my MLS predictions. The bad news is that I was perfect the wrong way going 0 for 8. This is actually pretty hard to do, but that is small consolation. Here is a recap of the results:

Predicted Score				Actual Score				Points
Philadelphia	2	Colorado	1	Philadelphia	1	Colorado	2	-10
Vancouver	1	Los Angeles	1	Vancouver	0	Los Angeles	4	-10
Kansas City	2	New England	1	Kansas City	1	New England	1	-10
Houston	1	Seattle	1	Houston	3	Seattle	1	-10
Real Salt Lake	2	Columbus	0	Real Salt Lake	0	Columbus	2	-10
San Jose	1	D.C. United	1	San Jose	0	D.C. United	2	-10
Portland	3	Toronto FC	2	Portland	2	Toronto FC	2	-10
FC Dallas	1	Chivas USA	1	FC Dallas	1	Chivas USA	0	-10

It was an unusually bad weekend for home teams with three teams getting shut out and only two teams, Dallas and Houston winning. Perhaps that Texas heat gave them an advantage. Let's see how these results impacted the SRS ratings:

Team	MOV	SOS	SRS	Change
Real Salt Lake	0.684	-0.014	0.670	-0.148
Los Angeles	0.696	-0.042	0.653	0.131
FC Dallas	0.391	-0.017	0.374	0.045
Philadelphia	0.350	-0.014	0.336	-0.079
New York	0.304	-0.011	0.293	0.009
Seattle	0.304	-0.021	0.284	-0.094
Columbus	0.182	0.024	0.205	0.121
Chivas USA	0.136	0.010	0.146	-0.044
Colorado	0.083	0.027	0.111	0.070
Houston	0.000	0.019	0.019	0.121
Kansas City	0.048	-0.127	-0.080	-0.033
D.C. United	-0.200	0.056	-0.144	0.130
San Jose	-0.227	0.043	-0.185	-0.122
Chicago	-0.250	0.031	-0.219	-0.002
New England	-0.455	0.033	-0.421	0.019
Portland	-0.500	0.026	-0.474	-0.024
Vancouver	-0.591	-0.018	-0.609	-0.146
Toronto FC	-0.917	-0.003	-0.920	0.020

Real Salt Lake holds on to the top spot despite their upset loss to Columbus. LA's big win over Vancouver puts them right behind the Claret-and-Cobalt. They actually have the highest average margin of victory (MOV), but RSL's better strength of schedule (SOS) keeps them ahead of the Galaxy overall.

MLS Week 20 Predictions

noreply@blogger.com (Bill Fish) — Fri, 29 Jul 2011 15:34:00 +0000

The MLS returns to action this weekend after the All-Star break. Hopefully this weekend's action will allow the league's best to get over their 4-0 thrashing by Manchester United. We have a total of eight matches this weekend. New York will be playing in the Emirates Cup tournament in London which features Thierry Henry's return to the Emirates Stadium. Chicago is idle Here are my predictions:

Predicted Score
Philadelphia	2	Colorado	1
Vancouver	1	Los Angeles	1
Kansas City	2	New England	1
Houston	1	Seattle	1
Real Salt Lake	2	Columbus	0
San Jose	1	D.C. United	1
Portland	3	Toronto FC	2
FC Dallas	1	Chivas USA	1

Marginal Goal Analysis - Philadelphia Union

noreply@blogger.com (Bill Fish) — Mon, 25 Jul 2011 21:01:00 +0000

Since I'm projecting them as an MLS Cup finalist, I thought it was time to do a marginal goal analysis for the Union. Here are the results for their MLS matches through 7/17:

Player	Minutes	GS	GA	GS90	GA90	GD90	MGS90	MGA90	MGD90
Keon Daniel	756	13	5	1.548	0.595	0.952	0.286	-0.249	0.535
Danny Califf	1620	24	15	1.333	0.833	0.500	0.072	-0.011	0.083
Carlos Ruiz	1144	14	8	1.101	0.629	0.472	-0.160	-0.215	0.055
Carlos Valdes	1620	24	16	1.333	0.889	0.444	0.072	0.044	0.027
Sheanon Williams	1693	24	16	1.276	0.851	0.425	0.014	0.006	0.008
Sebastien Le Toux	1710	24	16	1.263	0.842	0.421	0.002	-0.002	0.004
Brian Carroll	1401	21	15	1.349	0.964	0.385	0.088	0.119	-0.032
Danny Mwanga	968	16	12	1.488	1.116	0.372	0.226	0.271	-0.045
Kyle Nakazawa	988	14	11	1.275	1.002	0.273	0.014	0.157	-0.144
Justin Mapp	1047	16	13	1.375	1.117	0.258	0.114	0.273	-0.159
Jordan Harvey	1241	13	13	0.943	0.943	0.000	-0.319	0.098	-0.417
	17051	239	160	1.262	0.845	0.417	0.000	0.000	0.000

I'm only listing regular or semi-regular players. The totals at the bottom are for the entire team.

This shows that midfielder Keon Daniel is their most effective player, by far. His marginal goal differential is almost half a goal better than his closest teammate. Daniel's stats are somewhat inflated by his low playing time, but when he is on the pitch, Philadelphia is most effective at both scoring and preventing goals. I suspect that if he gets more playing time the team's performance will improve.

This table also shows why Philadelphia was right to trade Jordan Harvey. When Harvey was on the pitch, the team was least effective at scoring goals and was not very effective at preventing them. Harvey's stats include the Union's 6-goal outburst against Toronto where he played the entire match. Take that away and his stats are even worse. Harvey was traded to Vancouver for allocation money.

MLS Week 19 SRS Ratings and Season Projection

noreply@blogger.com (Bill Fish) — Mon, 25 Jul 2011 15:22:00 +0000

Real Salt Lake's big win over San Jose had a big impact on the SRS table and the season projection. Here are the SRS ratings:

Team	MOV	SOS	SRS	Change
Real Salt Lake	0.833	-0.015	0.818	0.208
Los Angeles	0.545	-0.023	0.522	0.033
Philadelphia	0.421	-0.006	0.415	-0.005
Seattle	0.409	-0.031	0.378	-0.010
FC Dallas	0.364	-0.035	0.329	0.005
New York	0.304	-0.021	0.284	-0.177
Chivas USA	0.190	-0.001	0.190	0.131
Columbus	0.095	-0.011	0.084	-0.024
Colorado	0.043	-0.002	0.041	0.144
Kansas City	0.050	-0.097	-0.047	0.069
San Jose	-0.143	0.080	-0.063	-0.202
Houston	-0.095	-0.006	-0.102	-0.141
Chicago	-0.250	0.033	-0.217	0.014
D.C. United	-0.316	0.042	-0.274	-0.094
New England	-0.476	0.036	-0.440	0.107
Portland	-0.526	0.077	-0.450	0.012
Vancouver	-0.429	-0.035	-0.463	0.042
Toronto FC	-0.957	0.017	-0.940	-0.045

The SRS ratings are heavily influenced by margin of victory so RSL's 4-goal margin caused the biggest swings for them and San Jose. Let's see how it influenced the season projections:

East
Pos	Team	P	W	D	L	GF	GA	GD	Pts
1	Philadelphia	34	15	14	5	43	28	15	59
2	New York	34	11	16	7	54	42	12	49
3	Kansas City	34	11	16	7	46	40	6	49
4	Columbus	34	12	11	11	34	32	2	47
5	D.C. United	34	7	17	10	44	53	-9	38
6	Houston	34	7	15	12	36	41	-5	36
7	Chicago	34	3	22	9	35	44	-9	31
8	New England	34	4	13	17	30	46	-16	25
9	Toronto FC	34	3	12	19	28	61	-33	21

West
Pos	Team	P	W	D	L	GF	GA	GD	Pts
1	Real Salt Lake	34	17	14	3	47	22	25	65
2	Los Angeles	34	16	16	2	43	26	17	64
3	FC Dallas	34	15	13	6	43	32	11	58
4	Seattle	34	14	16	4	47	34	13	58
5	Colorado	34	9	17	8	44	43	1	44
6	Chivas USA	34	9	17	8	45	39	6	44
7	Portland	34	11	5	18	40	59	-19	38
8	San Jose	34	6	16	12	41	44	-3	34
9	Vancouver	34	3	16	15	34	48	-14	25

So I'm now projecting RSL to finish ahead of Los Angeles in the West. The Western Conference final will be between those two teams. With the better record RSL gets the home field advantage. This is key because these two teams have the highest home SRS Ratings (RSL @ 1.266, LA @ 0.839) so whichever team can secure home field will be in the driver's seat for reaching the MLS Cup. I'm still projecting Philadelphia to secure the Eastern Conference berth in the Cup final. My new prediction for MLS Cup champion is RSL.

MLS Week 19 Recap

noreply@blogger.com (Bill Fish) — Mon, 25 Jul 2011 14:16:00 +0000

With apologies to Prince, the MLS on Saturday decided to party like it was 1998. There were 22 goals scored in six matches, a rate of 3.67 goals/match. The league has not scored at that rate since '98 when they averaged 3.57 goals per match. Also, no 0-0 draws on Saturday. Nice to see the offenses coming back especially when there is a lot of international attention on the league due to the World Football Challenge.

I broke even on my Saturday predictions:

Predicted Score				Actual Score				Points
New York	2	FC Dallas	1	New York	2	FC Dallas	2	-10
Columbus	2	Portland	0	Columbus	1	Portland	0	10
Kansas City	1	Toronto FC	0	Kansas City	4	Toronto FC	2	10
Colorado	2	New England	1	Colorado	2	New England	2	-10
Real Salt Lake	2	San Jose	1	Real Salt Lake	4	San Jose	0	10
Chivas USA	1	Houston	1	Chivas USA	3	Houston	0	-10
						Total		0

On outcomes I am now at 50% for July and 43% for the season. For scores my accuracy is at 10.5% for July and 10.0% for the year.

MLS Week 19 Weekend Predictions

noreply@blogger.com (Bill Fish) — Fri, 22 Jul 2011 16:52:00 +0000

Here are the results of my mid-week predictions:

Predicted Score				Actual Score				Points
D.C. United	2	New England	2	D.C. United	0	New England	1	-10
Toronto FC	1	FC Dallas	2	Toronto FC	0	FC Dallas	1	15
Colorado	1	New York	1	Colorado	4	New York	1	-10
San Jose	2	Vancouver	1	San Jose	2	Vancouver	2	-10
Los Angeles	1	Columbus	0	Los Angeles	1	Columbus	0	30
						Total		15

Overall I had 2 of 5 outcomes and 1 of 5 scores predicted correctly. Here are my predictions for the six weekend matches:

Predicted Score
New York	2	FC Dallas	1
Columbus	2	Portland	0
Kansas City	1	Toronto FC	0
Colorado	2	New England	1
Real Salt Lake	2	San Jose	1
Chivas USA	1	Houston	1

For RSL this will be their first match in two weeks due to the postponement of last week's match in Vancouver. Although they have played the fewest matches in league play, they have also played six CONCACAF Champions League and two US Open Cup matches this season. So the extra week off should do them good and prepare them for a busy second-half of the season.

MLS Scoring Drought - Another Theory

noreply@blogger.com (Bill Fish) — Thu, 21 Jul 2011 21:56:00 +0000

I have written about this before and Steve Davis of SI.com recently proposed other theories as to why the MLS is scoring at a record-low level. Here is another - the league has gotten bigger than the talent pool. This table shows goals per match by season relative to the number of teams in the league:

Season	Total Goals	Matches	Goals per Match	Teams
1996	539	160	3.37	10
1997	522	160	3.26	10
1998	685	192	3.57	12
1999	549	192	2.86	12
2000	612	192	3.19	12
2001	518	158	3.28	12
2002	421	140	3.01	10
2003	433	150	2.89	10
2004	392	150	2.61	10
2005	551	192	2.87	12
2006	503	192	2.62	12
2007	519	195	2.66	13
2008	590	210	2.81	14
2009	569	225	2.53	15
2010	590	240	2.46	16
2011	437	182	2.40	18

The league has expanded seven times since it's formation. The first four expansions (1998, 2005, 2007 and 2008) each resulted in an increase in goals per match. However, the last three have resulted in decreased goals per match. I think this is because the talent pool has become diluted. The league is now 50% larger than it was five years ago and, let's face it, they are not drawing those additional players from the Premier League or La Liga or Serie A.

Here are the average goals scored by size of the league:

Teams	Goals per Match
10	3.04
12	3.06
13	2.66
14	2.81
15	2.53
16	2.46
18	2.40

So if the MLS wants to see more offense they should consider shrinking the league. I think a 12-team league would hit the sweet spot. Unfortunately the league will be adding a 19th team next season in Montreal. I expect to see even more 0-0 draws next season :(

MLS Week 19 Mid-week Predictions

noreply@blogger.com (Bill Fish) — Wed, 20 Jul 2011 20:27:00 +0000

We actually have 5 matches tonight. Here are my predictions:

Predicted Score
D.C. United	2	New England	2
Toronto FC	1	FC Dallas	2
Colorado	1	New York	1
San Jose	2	Vancouver	1
Los Angeles	1	Columbus	0

MLS SRS Ratings and Season Projection

noreply@blogger.com (Bill Fish) — Wed, 20 Jul 2011 20:15:00 +0000

Here are the updated ratings through Week 18:

Team	MOV	SOS	SRS	Change
Real Salt Lake	0.647	-0.037	0.610	-0.011
Los Angeles	0.524	-0.035	0.489	-0.001
New York	0.476	-0.015	0.461	-0.021
Philadelphia	0.421	-0.001	0.420	0.118
Seattle	0.409	-0.021	0.388	0.031
FC Dallas	0.350	-0.026	0.324	-0.028
San Jose	0.053	0.086	0.139	0.016
Columbus	0.105	0.003	0.108	-0.016
Chivas USA	0.050	0.009	0.059	0.025
Houston	0.050	-0.011	0.039	-0.013
Colorado	-0.095	-0.008	-0.103	-0.021
Kansas City	-0.053	-0.064	-0.116	0.003
D.C. United	-0.278	0.098	-0.180	0.034
Chicago	-0.250	0.019	-0.231	-0.082
Portland	-0.500	0.038	-0.462	0.077
Vancouver	-0.450	-0.055	-0.505	-0.004
New England	-0.579	0.032	-0.547	-0.120
Toronto FC	-0.905	0.010	-0.895	0.004

The results of the New England/Philadelphia match moved both teams two spots in the rankings - Philadelphia up two to 4th and the Revs down two to 17th. The Union's victory puts them on top in the Eastern Conference, two points clear of New York.

Here are the projected standings for the season:

East
Pos	Team	P	W	D	L	GF	GA	GD	Pts
1	Philadelphia	34	17	13	4	44	26	18	64
2	New York	34	12	17	5	54	37	17	53
3	Columbus	34	12	11	11	35	32	3	47
4	Kansas City	34	9	18	7	39	36	3	45
5	D.C. United	34	7	18	9	47	53	-6	39
6	Houston	34	6	19	9	36	37	-1	37
7	Chicago	34	3	22	9	35	43	-8	31
8	New England	34	3	13	18	28	48	-20	22
9	Toronto FC	34	3	12	19	27	59	-32	21

West
Pos	Team	P	W	D	L	GF	GA	GD	Pts
1	Los Angeles	34	16	16	2	44	26	18	64
2	Real Salt Lake	34	16	15	3	44	24	20	63
3	Seattle	34	14	16	4	47	35	12	58
4	FC Dallas	34	14	12	8	41	33	8	54
5	Chivas USA	34	8	18	8	41	39	2	42
6	San Jose	34	8	17	9	42	39	3	41
7	Colorado	34	8	17	9	39	42	-3	41
8	Portland	34	9	7	18	40	60	-20	34
9	Vancouver	34	3	15	16	32	48	-16	24

I'm projecting both Los Angeles and Philadelphia to finish at 64 points, but LA already owns the tiebreaker with a win and a draw against the Union in their two matches earlier this season. Still projecting those two to meet for the MLS Cup which I think the Galaxy will win.

MLS week 18 Recap

noreply@blogger.com (Bill Fish) — Wed, 20 Jul 2011 18:10:00 +0000

An interesting weekend but unfortunately not due to the quality of play. Vancouver found themselves in a dilemma. They had scheduled a friendly with Manchester City on Monday and had laid a temporary grass pitch over their turf field for that match. Their Saturday match with RSL would also be played on the grass pitch. However, heavy rains in Vancouver created a waterlogged pitch with several areas of standing water so the Whitecaps had to postpone their match with RSL. In other words, Vancouver rescheduled a league match so that their field would be ready for a friendly match two days later.

I understand that the MLS schedules these matches because the European teams are a great draw, but these matches don't count for anything. For the Euro teams they are just pre-season tune-ups and they are not included in the MLS standings. When the league disrupts their schedule to accommodate one of these friendlies they send the message that they're just fodder for the top international leagues.

Let's see how my predictions turned out:

Predicted Score				Actual Score				Points
Seattle	1	Colorado	1	Seattle	4	Colorado	3	-10
Columbus	2	San Jose	1	Columbus	0	San Jose	0	-10
Chicago	2	Portland	0	Chicago	0	Portland	1	-10
Houston	2	Kansas City	2	Houston	1	Kansas City	1	20
FC Dallas	2	D.C. United	1	FC Dallas	0	D.C. United	0	-10
Chivas USA	1	New York	1	Chivas USA	0	New York	0	20
New England	1	Philadelphia	1	New England	0	Philadelphia	3	-10
						Total		-10

Got 2 of 7 outcomes correct, but no scores. YTD my accuracy is now 42.6% for outcomes and 10.1% for scores.

MLS Week 18 Predictions

noreply@blogger.com (Bill Fish) — Thu, 14 Jul 2011 14:39:00 +0000

Once again there are only eight matches this weekend with Los Angeles and Toronto FC sitting idle. Here are my predictions:

Predicted Score
Vancouver	1	Real Salt Lake	1
Seattle	1	Colorado	1
Columbus	2	San Jose	1
Chicago	2	Portland	0
Houston	2	Kansas City	2
FC Dallas	2	D.C. United	1
Chivas USA	1	New York	1
New England	1	Philadelphia	1

I think the best matchup will be Chivas vs New York. In their previous matchup, Chivas shocked the Red Bulls at home thanks to a Justin Braun hat trick. Braun has scored only one other goal the rest of the season so don't expect a repeat performance. New York is the highest scoring team in the league overall (1.7 goals/match) and on the road (1.6 goals/match) but Chivas has been stingy at home allowing 0.9 goals/match. On offense, Chivas has struggled to score at home with 1.1 goals per match which is actually lower than their road scoring average of 1.4 goals/match. RBNY's road defense has been their Achilles heel, allowing 1.5 goals/match.

The outcome of this match could have big playoff implications. New York and Philadelphia are neck-and-neck for the Eastern Conference title. Anything less than a point for the Red Bulls will hurt their chances of winning the conference. I'm currently projecting Chivas for the final playoff spot. Anything less than a point puts their playoff hopes in limbo.

Marginal Goal Analysis - Los Angeles Galaxy

noreply@blogger.com (Bill Fish) — Wed, 13 Jul 2011 20:37:00 +0000

Since my latest projection has Los Angeles winning the MLS Cup, I thought that it was time to do a marginal goal analysis for the Galaxy. Here are the results for players with significant playing time through their 21 MLS matches:

Player	Primary Position	Minutes	GS	GA	GS90	GA90	GD90	MGS90	MGA90	MGD90
Todd Dunivant	D	1890	27	16	1.286	0.762	0.524	-0.006	-0.008	0.003
Sean Franklin	D	1664	24	15	1.298	0.811	0.487	0.007	0.041	-0.035
Juninho	M	1638	22	13	1.209	0.714	0.495	-0.083	-0.056	-0.027
Omar Gonzalez	D	1609	24	11	1.342	0.615	0.727	0.051	-0.155	0.206
A.J. DeLaGarza	D	1547	22	14	1.280	0.814	0.465	-0.011	0.044	-0.056
David Beckham	M	1338	20	11	1.345	0.740	0.605	0.054	-0.030	0.084
Juan Pablo Angel	F	1312	18	13	1.235	0.892	0.343	-0.057	0.122	-0.178
Mike Magee	M	1205	20	8	1.494	0.598	0.896	0.202	-0.172	0.375
Chris Birchall	M	1203	16	12	1.197	0.898	0.299	-0.094	0.128	-0.222
Chad Barrett	F	1182	16	11	1.218	0.838	0.381	-0.073	0.068	-0.141
Landon Donovan	F	1154	18	7	1.404	0.546	0.858	0.112	-0.224	0.337
Miguel López	M	896	14	4	1.406	0.402	1.004	0.115	-0.368	0.483

Miguel Lòpez has the highest MGD90 and MGA90 scores of any player I've rated so far. When he's on the field, the Galaxy shuts down their opponents. He also has been an effective contributor to their offense. Sticking out like a sore thumb is Chris Birchall's stat line. When he is on the field, the Galaxy are at their least effective both at scoring goals and preventing goals. I think that giving more midfield playing time to Lòpez and more bench time to Birchall will make the Galaxy a stronger team. Birchall may be getting more bench time anyway after this vicious tackle.

Of the three teams I have rated (Real Salt Lake and New York the others), only LA has three players (Lòpez, Mike Magee and Landon Donovan) with MGD90's above 0.300. I'm especially pleased to see the high rating for Donovan who is generally considered the best player in the MLS. To me this is validation that marginal goals are an effective way to measure a player's contribution to his team.

MLS Season Projection

noreply@blogger.com (Bill Fish) — Wed, 13 Jul 2011 13:52:00 +0000

We are now halfway through the 2011 MLS season. Here is my latest projection of the final standings based on results through week 17:

East
Pos	Team	P	W	D	L	GF	GA	GD	Pts
1	Philadelphia	34	15	12	7	40	27	13	57
2	New York	34	13	17	4	56	38	18	56
3	Columbus	34	14	10	10	39	33	6	52
4	Kansas City	34	9	18	7	40	37	3	45
5	D.C. United	34	7	17	10	49	57	-8	38
6	Houston	34	6	19	9	38	38	0	37
7	Chicago	34	5	22	7	39	42	-3	37
8	New England	34	3	15	16	29	45	-16	24
9	Toronto FC	34	3	12	19	27	59	-32	21

West
Pos	Team	P	W	D	L	GF	GA	GD	Pts
1	Los Angeles	34	16	16	2	43	26	17	64
2	Real Salt Lake	34	16	15	3	45	23	22	63
3	Seattle	34	14	16	4	44	31	13	58
4	FC Dallas	34	14	12	8	43	34	9	54
5	Colorado	34	7	20	7	36	39	-3	41
6	Chivas USA	34	7	18	9	43	42	1	39
7	San Jose	34	7	17	10	44	42	2	38
8	Portland	34	8	7	19	39	65	-26	31
9	Vancouver	34	3	15	16	32	48	-16	24

Now projecting Los Angeles and Philadelphia as the conference champs. RSL dropped two points because they have the toughest remaining schedule. It includes two matches against New York, Los Angeles on the road and Philadelphia at home. New York's loss to D.C. United over the weekend cost them some points, allowing Philadelphia to move ahead of them. For the MLS Cup I'm now projecting Los Angeles to defeat Philadelphia.

MLS week 17 Recap

noreply@blogger.com (Bill Fish) — Tue, 12 Jul 2011 18:55:00 +0000

I did extremely well with my predictions this weekend:

Predicted Score				Actual Score				Points
Kansas City	1	Chivas USA	1	Kansas City	1	Chivas USA	1	30
New York	2	D.C. United	1	New York	0	D.C. United	1	-10
Houston	2	Toronto FC	1	Houston	2	Toronto FC	0	10
Real Salt Lake	2	FC Dallas	1	Real Salt Lake	2	FC Dallas	0	10
Colorado	1	Vancouver	0	Colorado	2	Vancouver	1	15
Los Angeles	2	Chicago	1	Los Angeles	2	Chicago	1	30
San Jose	1	Philadelphia	1	San Jose	0	Philadelphia	0	20
Portland	1	Seattle	1	Portland	2	Seattle	3	-10

That's 6 out of 8 matches where my model predicted the correct outcome. Also predicted the correct final score in 2 matches. For the month of July, my accuracy rate has been 60% on outcomes and 15% on scores. Season-to-date those rates are 43% and 11% respectively.

Updated SRS ratings after the jump:

Team	MOV	SOS	SRS	Change
Real Salt Lake	0.647	-0.026	0.621	0.112
Los Angeles	0.524	-0.034	0.490	0.011
New York	0.500	-0.018	0.482	0.130
Seattle	0.381	-0.024	0.357	-0.012
FC Dallas	0.368	-0.016	0.352	-0.101
Philadelphia	0.278	0.024	0.302	-0.014
Columbus	0.111	0.013	0.124	0.039
San Jose	0.056	0.068	0.123	-0.116
Houston	0.053	-0.001	0.052	0.076
Chivas USA	0.053	-0.019	0.034	0.118
Colorado	-0.050	-0.032	-0.082	0.057
Kansas City	-0.056	-0.064	-0.119	-0.004
Chicago	-0.211	0.062	-0.149	-0.020
D.C. United	-0.294	0.080	-0.214	0.120
New England	-0.444	0.017	-0.427	0.006
Vancouver	-0.450	-0.051	-0.501	-0.075
Portland	-0.588	0.050	-0.539	0.002
Toronto FC	-0.905	0.005	-0.899	-0.254

Real Salt Lake continues their run at the top. Their win over FC Dallas improved their rating quite a bit. But, New York's 5-0 win had a much greater impact on their rating, so much that they are still most improved despite a 1-0 loss to D.C. United and their old friend DeRo. Toronto dug a huge hole for themselves with two losses by a combined score of 7-0. They're basically buried at the bottom of the table unless these guys can turn things around for them.

MLS Week 17 Predictions

noreply@blogger.com (Bill Fish) — Fri, 08 Jul 2011 17:41:00 +0000

I was one for four on my midweek predictions, getting the NY-Toronto outcome right, but certainly not the magnitude! Let's see how we do for the weekend. There are only eight games with Columbus and New England sitting idle:

Predicted Score
Kansas City	1	Chivas USA	1
New York	2	D.C. United	1
Houston	2	Toronto FC	1
Real Salt Lake	2	FC Dallas	1
Colorado	1	Vancouver	0
Los Angeles	2	Chicago	1
San Jose	1	Philadelphia	1
Portland	1	Seattle	1

Marginal Goal Analysis - New York Red Bulls

noreply@blogger.com (Bill Fish) — Thu, 07 Jul 2011 22:01:00 +0000

Following up on my analysis of Real Salt Lake, I decided to do a similar analysis for New York. MLS released their Castrol Index today and Thierry Henry continues to hold the top spot. I want to measure marginal goals for Henry and his teammates to see if I can get a similar result. Here is the table for the Red Bulls. This data includes last night's 5-0 stomping of Toronto FC:

Player	Minutes	GS	GA	GS90	GA90	GD90	MGS90	MGA90	MGD90
Roy Miller	1709	34	23	1.791	1.211	0.579	0.006	0.001	0.006
Joel Lindpere	1705	33	23	1.742	1.214	0.528	-0.042	0.004	-0.046
Jan Gunnar Solli	1629	33	22	1.823	1.215	0.608	0.039	0.005	0.034
Teemu Tainio	1545	32	22	1.864	1.282	0.583	0.080	0.071	0.009
Thierry Henry	1247	25	14	1.804	1.010	0.794	0.020	-0.200	0.220
Tim Ream	1170	25	13	1.923	1.000	0.923	0.139	-0.211	0.349
Dwayne De Rosario	1113	23	19	1.860	1.536	0.323	0.076	0.326	-0.250
Dane Richards	1092	23	14	1.896	1.154	0.742	0.111	-0.057	0.168
Mehdi Ballouchy	999	19	16	1.712	1.441	0.270	-0.072	0.231	-0.303
Luke Rodgers	949	19	12	1.802	1.138	0.664	0.018	-0.072	0.090
Carlos Mendes	880	19	15	1.943	1.534	0.409	0.159	0.324	-0.165
Rafael Marquez	810	15	8	1.667	0.889	0.778	-0.118	-0.322	0.204

17100

339

230

1.784

1.211

0.574

0.000

I've limited the list to show those players that have played roughly half of the total minutes over RBNY's 19 MLS matches. New York's MVP according to my analysis is defender Tim Ream. His MGD90 rating of 0.349 goals is equivalent to 11.9 goals in a 34-match season. Ream has played in 13 matches and his team has allowed exactly one goal per match with him in the lineup. In the 6 matches he has missed (due to playing for the US Men's team) RBNY allowed 1.67 goals per match. Ream is also contributing to New York's league-leading offense with the second-best MGS90 score.

Thierry Henry scores pretty highly with and MGD90 of 0.220. Although he leads the league in goals scored, his marginal contribution is pretty small on offense with an MGS90 of only 0.020. New York is the highest scoring team in the league and Henry is reaping the benefits of that. I suspect that if Henry played for another team he wouldn't be leading the league in goals. Not saying he is bad or overrated, just pointing out that his value should not be measured just by his goals. His MGA90 of -0.200 is just behind Ream's so this indicates that his presence on the field helps prevent opponents from scoring.

One thing I would like to point out is that this table shows why RBNY traded Dwayne De Rosario. Considered a surprising move at the time, but DeRo's MGD90 was at -0.250 and he had the worst MGA90 at 0.326. Compare his stats to Dane Richards. They're both primarily midfielders who played about the same minutes. When De Rosario was on the pitch the team gave up five more goals than when Richards was on the pitch. So this was a good move by the Red Bulls.

MLS Week 16 SRS Rankings and Season Projection

noreply@blogger.com (Bill Fish) — Thu, 07 Jul 2011 15:38:00 +0000

Another drawn out weekend in the MLS (get it?) with 6 of 9 matches ending all tied up and two of those were 0-0. The most entertaining match was Real Salt Lake's 3-3 draw at home against New England. This was a matchup of the best defense (RSL) against the worst offense (NE) and it was played at Rio Tinto where RSL is nearly unbeatable. Problem was that RSL dug themselves into a hole when Nat Borchers drew a red card in the 3rd minute. The Revolution converted on the subsequent penalty kick and added another to go up 2-0 on 10-man RSL.

RSL fought back to tie it in the 2nd half but New England took the lead back in the 63rd minute and seemed poised to pull off a huge upset. But a 2nd yellow on Ryan Cochrane evened up both sides at 10 men. This opened up some space for RSL's offense which tied up the game on this goal. Overall this match featured 6 goals, 2 penalty kicks, 6 yellow cards and 2 red cards. Referee Yader Reyes generated a lot of controversy for his handling of the match.

Here are the SRS ratings through Week 16:

Team	MOV	SOS	SRS
Real Salt Lake	0.563	-0.053	0.509
Los Angeles	0.500	-0.021	0.479
FC Dallas	0.500	-0.047	0.453
Seattle	0.350	0.019	0.369
New York	0.333	0.018	0.352
Philadelphia	0.294	0.022	0.316
San Jose	0.188	0.051	0.239
Columbus	0.059	0.026	0.085
Houston	-0.056	0.032	-0.024
Chivas USA	-0.059	-0.026	-0.084
Kansas City	-0.063	-0.052	-0.115
Chicago	-0.167	0.038	-0.129
Colorado	-0.111	-0.028	-0.139
D.C. United	-0.375	0.041	-0.334
Vancouver	-0.389	-0.037	-0.426
New England	-0.444	0.011	-0.433
Portland	-0.563	0.022	-0.541
Toronto FC	-0.632	-0.013	-0.645

RSL continues to top the rankings. They have played the easiest schedule in the league thus far. The Western Conference now has the top 4 spots and 5 of the top 7. Toronto remains at the bottom despite their victory over Vancouver. The Whitecaps were not strong enough to really improve Toronto's rating.

After the jump I will have my updated season projection:

Eastern Conference
Pos	Team	P	W	D	L	GF	GA	GD	Pts
1	New York	34	13	18	3	54	40	14	57
2	Philadelphia	34	14	14	6	42	29	13	56
3	Kansas City	34	12	15	7	41	34	7	51
4	Columbus	34	13	8	13	38	36	2	47
5	Chicago	34	6	21	7	40	42	-2	39
6	Houston	34	6	17	11	38	40	-2	35
7	D.C. United	34	5	15	14	50	62	-12	30
8	New England	34	3	15	16	29	45	-16	24
9	Toronto FC	34	3	13	18	29	54	-25	22

Western Conference
Pos	Team	P	W	D	L	GF	GA	GD	Pts
1	Real Salt Lake	34	17	14	3	44	25	19	65
2	Los Angeles	34	16	16	2	41	26	15	64
3	FC Dallas	34	14	14	6	45	33	12	56
4	Seattle	34	13	17	4	41	31	10	56
5	San Jose	34	10	16	8	49	42	7	46
6	Colorado	34	8	16	10	34	38	-4	40
7	Portland	34	9	7	18	37	62	-25	34
8	Chivas USA	34	5	18	11	40	45	-5	33
9	Vancouver	34	4	16	14	35	46	-11	28

Philadelphia had the only significant change from last week with their projected points dropping from 62 to 56 and dropping from first in the Eastern Conference to second. In fact I'm now projecting that they will lose in the conference finals to New York. The Union's projection was hurt by their draw against D.C. United, a match they were expected to win. That cost them two points and after reprojecting based on week 16 they are expected to lose an additional four points over the course of the season. They

So my projected MLS Cup final is RSL vs New York, but I'm still projecting RSL as the MLS Cup champions.

MLS Week 16 Recap

noreply@blogger.com (Bill Fish) — Wed, 06 Jul 2011 18:36:00 +0000

Had a pretty good week with my predictions as you can see below:

Predicted Score				Actual Score				Points
Toronto FC	1	Vancouver	1	Toronto FC	1	Vancouver	0	-10
D.C. United	1	Philadelphia	2	D.C. United	2	Philadelphia	2	-10
FC Dallas	1	Columbus	0	FC Dallas	2	Columbus	0	10
Chivas USA	1	Chicago	1	Chivas USA	1	Chicago	1	30
San Jose	1	New York	1	San Jose	2	New York	2	20
Portland	3	Kansas City	2	Portland	1	Kansas City	2	-10
Colorado	1	Houston	1	Colorado	0	Houston	0	20
Real Salt Lake	2	New England	0	Real Salt Lake	3	New England	3	-10
Los Angeles	1	Seattle	1	Los Angeles	0	Seattle	0	20
						Total		60

For the week I had correct predictions for 5 of 9 outcomes and 1 score. My accuracy for the year is up to 41.8% on outcomes and 10.0% on scores.

The MLS has four mid-week games, all being played tonight. Here are my predictions for those matches:

Predicted Score
New York	2	Toronto FC	1
Kansas City	1	Colorado	0
Vancouver	1	Columbus	0
Chivas USA	1	San Jose	1

Measuring a Player's Contribution to His Team

noreply@blogger.com (Bill Fish) — Fri, 01 Jul 2011 02:14:00 +0000

The goal I had when starting this blog was to analyze soccer statistically, "Bill James-style". My focus so far has been on team stats. I'm now going to take a crack at player stats. Unlike baseball or football, soccer is not a sport that lends itself well to statistical analysis. This is because of the fluid nature of the game. Baseball has discrete events, such as pitches or plate appearances, that serve as the foundation of sabrmetric research. Advanced football statistics are based on analysis of plays. Soccer does not have many discrete events. Although there are set pieces such as corner kicks, free kicks and throw-ins, these are a small part of the game. Most of the action consists of the passing and dribbling and heading that is constantly happening. Each touch of the ball can be considered a discrete event, but there are few meaningful statistics that can be calculated from this data at this point in time.

What I'm trying to do here is determine how valuable a player is to his team using a rather simplistic method - measuring the difference between goals scored and goals allowed for a team when a particular player is on the pitch. This is roughly equivalent to the Plus/Minus stat used in hockey. I'll be using Real Salt Lake as an example. Here are their stats for MLS matches through 6/26:

Player	Minutes	GS	GA	GS90	GA90	GD90	MGS90	MGA90	MGD90
Kyle Beckerman	1287	16	9	1.119	0.629	0.490	-0.071	0.041	-0.112
Ned Grabavoy	1244	18	8	1.302	0.579	0.723	0.112	-0.009	0.122
Nat Borchers	1170	15	7	1.154	0.538	0.615	-0.036	-0.050	0.014
Tony Beltran	1030	13	8	1.136	0.699	0.437	-0.054	0.111	-0.165
Chris Wingert	1013	13	8	1.155	0.711	0.444	-0.035	0.123	-0.157
Jamison Olave	958	12	6	1.127	0.564	0.564	-0.062	-0.025	-0.038
Will Johnson	780	8	4	0.923	0.462	0.462	-0.267	-0.127	-0.140
Fabian Espindola	760	12	6	1.421	0.711	0.711	0.231	0.122	0.109
Andy Williams	719	10	7	1.252	0.876	0.376	0.062	0.288	-0.226
Robbie Russell	625	8	4	1.152	0.576	0.576	-0.038	-0.012	-0.026
Collen Warner	615	9	2	1.317	0.293	1.024	0.127	-0.296	0.423
Alvaro Saborio	594	5	2	0.758	0.303	0.455	-0.432	-0.285	-0.147
Jean Alexandre	539	7	4	1.169	0.668	0.501	-0.021	0.080	-0.101
Chris Schuler	450	6	3	1.200	0.600	0.600	0.010	0.012	-0.002
Javier Morales	405	6	2	1.333	0.444	0.889	0.143	-0.144	0.287
Arturo Alvarez	396	3	1	0.682	0.227	0.455	-0.508	-0.361	-0.147
Luis Gil	323	5	5	1.393	1.393	0.000	0.203	0.805	-0.602
Paulo Araujo Jr.	194	7	0	3.247	0.000	3.247	2.058	-0.588	2.646
Nelson Gonzalez	174	3	1	1.552	0.517	1.034	0.362	-0.071	0.433
Artur Aghasyan	98	0	1	0.000	0.918	-0.918	-1.190	0.330	-1.520
Rauwshan McKenzie	90	2	0	2.000	0.000	2.000	0.810	-0.588	1.398
Total	13464	178	88	1.190	0.588	0.602	0.000	0.000	0.000

Minutes = minutes played by each player.

GS = goals scored by the team during the time that player was on the pitch.

GA = goals allowed by the team during the time the player was on the pitch.

GS90 = GF per 90 minutes.

GA90 = GA per 90 minutes.

GD90 = GF90 - GA 90.

MGS90 is Marginal Goals Scored per 90 minutes = Player's GS90 - Team's GS90.
MGA90 is Marginal Goals Allowed per 90 minutes = Player's GA90 - Team's GA90.
MGD90 is Marginal Goal Differential per 90 minutes = Player's GD90 - Team's GD90.

The key figures are the "Marginal" stats which measure the player's contribution over and above the team's performance. For example RSL averages 1.190 goals scored per 90 minutes. A player with a GS90 below 1.190 is not performing at the same level as his teammates so is actually hurting the team offensively. If the player's GS90 is greater than 1.190 he is performing above the level of his teammates and is helping the team. In the above table higher MGS90 scores are better while lower (negative) MGA90 scores are better. MGD90 captures a player's total performance (offense plus defense) and higher scores are better.

The marginal scores of the players at the top and bottom of the table are affected by low amounts of playing time so tend to be extremely high or low. Of those that have played over 1000 minutes, midfielder Ned Grabavoy has the highest MGD90 score of 0.122. When Grabavoy is on the field, RSL's goal differential is 0.122 better than when he isn't playing. Grabavoy's contributions can't be explained by goals and assists. In fact he has none of either. But as a midfielder he is contributing via tackles, passes, free kicks, headers, etc. The effect of all these touches is subtle and hard to measure directly so I'm using marginal stats to infer the impact of the player's performance. His MGD90 rate of 0.122 goals is worth just over 4 additional goals in a 34-game season.

The 1000-minute player with the worst MGD90 score is Tony Beltran at -0.165. His MGS90 is -0.054 which means he is hurting the team offensively. His MGA90 of 0.111 hurts the team defensively (remember that negative MGA scores are better). His MGD90 means that the team's goal differential is 5.6 goals worse over a full season when he is on the pitch.

So let me know what you think in the comments about my methods. I plan on adding other teams to this analysis. Let me know which teams you would like to see!