MLS Scariness: 2008 Week 5

Okay, time for another look at MLS scariness. I’ve settled into the formula I hope to use for the rest of the season (I’m not going to go back and recalculate scariness for the previous weeks, but I am scaling the scores down to match the current range). The formula is as follows:

pts (game) + gd (game) + opp ppg (last 5 games) + 1 (away team)

I figure a weighted average over the last five weeks (or as many
weeks as the teams have played) to come up with the current
scariness.

Enough about math though, here’s the current table:

Team	Scariness	Ranking	Change
NY	4.9	1 (3)	+2
CLB	4.14	2 (2)	–
DAL	3.76	3 (4)	+1
KC	3.75	4 (1)	-3
TOR	3.72	5 (8)	+3
NE	3.49	6 (7)	+1
CHI	3.29	7 (5)	-2
DC	2.94	8 (12)	+4
LA	2.93	9 (14)	+5
COL	2.22	10 (6)	-4
HOU	1.98	11 (9)	-2
SAN	1.75	12 (11)	-1
SLC	0.93	13 (13)	–
CHV	0.9	14 (10)	-4

LA and DC moved way up the charts based on lopsided wins. Toronto also moved up, but I think that’s based more on a three game winning streak (man, BMO is looking like a fortress and a great place to catch a game¹.)

Chivas and Colorado were the big sinkers. Chivas isn’t playing very well in general and got clubbed by LA, while the Rapids are settling into what I think i a more realistic ranking after some surprising early wins.

My only other thought is that being an RSL fan kinda stinks … oh well, I guess I can always get my kicks watching the ‘Fire Kreis’ folks get wound up.

¹ I’ve gotta say, I really envy a young friend who just got called on an LDS mission to Toronto—he was on BYU’s reserve list last year and is going to love being in a soccer hungry area.

USL1 Scariness: 2008 Week 3

Okay, for a first post on Monday morning, I thought I’d put up my scariness rankings for the USL1 (now that they’ve got three weeks of games under their belts. I’ve also settled into a formula for each game:

pts (game) + gd (game) + opp ppg (last 5 games) + 1 (away team)

I figure a weighted average over the last five weeks (or as many weeks as the teams have played) to come up with the current scariness.

Team	Scariness	Rank
POR	4.27	1
MIN	3.9	2
CAR	3.05	3
ATL	2.88	4
MON	2.25	5
CHA	2.2	6
SEA	1.7	7
PUE	1.51	8
MIA	1.18	9
VAN	1.15	10
ROC	0	11

At this point, Portland is looking like the cream of the USL, unbeaten and untied in 3 games with a +4 goal differential makes them look awfully good. It will be interesting to see how they do in US Open Cup play.

Down near the bottom of the table, Miami and Vancouver are separated by only .03 points, and are ahead of only Rochester (who has yet to play a game).

Don’t look for a USL2 scariness post from me, but I am planning on doing one for the PDL Northwestern conference.

2008 Week Four Scariness

Last weeks scariness rankings were impacted by an error I made in caclulations … This week I’ve tried to ensure no such wonkiness has occured. This week’s edition also contains a new feature—I’ve added a sparkline-style graphic showing points earned over the last several games (I’ll add a similar graphic showing scariness fluctuations as the season progresses).

Kansas City has really pushed to the front of the pack this week, with Columbus looking like a strong second. The third through sixth spots aren’t as firm.

LA has cemented a place at the bottom of the table, but I think the 7-13 spots will continue to see a lot of flux as the teams build a bigger body of work to be evaluated against.

Team	Ranking	Scariness
KC	1	13.96
CLB	2	11.25
NY	3	10.33
DAL	4	9.78
CHI	5	8.73
COL	6	8.25
NE	7	7.92
TOR	8	7.88
HOU	9	5.4
SAN	10	5.25
CHV	11	5.18
DC	12	4.73
SLC	13	4.6
LA	14	3.8

2008 Scariness: Week 3

With three weeks under our belts, the scariness rankings are still looking a little odd, but better than last week. Scariness is calculated from results versus the opponents recent points, and then weighted so that recent results count more than those in the past. Here’s where things stand today:

Team	Ranking (LW)	Scariness	Pts
KC	1 (1)	6.36	7
NE	2 (7)	6.24	5
COL	3 (3T)	6.12	6
CHI	4 (2)	5.18	7
CLB	5 (12)	4.92	6
DAL	6 (6)	3.74	5
SLC	7 (8)	3.68	4
CHV	8 (5)	3.66	4
NY	9 (3T)	3.36	3
HOU	10 (9T)	2.68	2
TOR	11 (13T)	2.4	3
DC	12 (9T)	2.16	3
LA	13 (9T)	0.54	3
SAN	14 (13T)	0	0

Even with the tie, the Wizards are hanging on to the top spot, one of just three teams to hold position along with FC Dallas who look stuck on sixth and the Quakes who don’t look to leave the basement anytime soon.

The Revolution and the Crew rode wins (over Kansas City and Chivas, respectively) to big improvements. The New York Red Bulls were the big losers, dropping six spots on their loss.

One nice thing about this system, is that I don’t have to defend any teams position since I didn’t put them there. I guess I could spend time trying to improve the math behind the system, but that’s a topic for another day.

Week Two in 2008

Well, I managed not to post my predictions before the weekend—I guess that’s what happens when you try to juggle to many things. I had two results correct from the 5 games I called. Here’s what I thought would happen, alongside what really did.

Game	Projected Winner	Projected Score	Actual Result
COL @ KC	KC	1-2	2-3
TOR @ DC	DC	0-1	1-4
CLB @ NY	—	1-1	0-2
RSL @ CHV	—	1-1	1-3
DAL @ HOU	HOU	0-1	3-3

I’d called one other game, the Wednesday night game of New England @ Kansas City. I think Kansas City will take this one 1-2, New England has too many players out due to injury or suspension to win this one on the road against a good looking Kansas City.

My scariness rankings are starting to make sense. They’ll really clear up over the next couple of weeks though, at this point there’s juts not enough data. Here’s my table at the momement:

Ranking	Team	Points
1	KC	4.08
2	CHI	2.48
3	COL	1.92
3	NY	1.92
5	CHV	1.56
6	DAL	1.16
7	NE	0.96
8	RSL	0.8
9	DC	0.54
9	LA	0.54
9	HOU	0.54
12	CLB	0.48
13	SAN	0
13	TOR	0

Scariness is a function of a team’s PPG and their opponents PPG in a weighted moving average. Last year, I was fancier than this, but I’m not sure it really matters.

Week 23 Power Rankings

Weeks of non-soccer obligations and outright neglect have basically killed off my scariness rankings—at least for now, hopefully I can get them in shape in time for the playoffs. In their place, let me offer up my take on power rankings. The league splits pretty readily into 3 tiers, with some hints of movement between them. Here’s how I see it:

The Good

Up at the top of the stack are the good teams. These five clubs are pretty much assured of playoff spots and four of them should be in next years Super Liga (though there may be a surprise here).

1. DC United (48 pts, 24 games, 2 ppg)—These guys are red hot at the right time. Emilio is looking unstoppable and should handily break 20 goals for the season. Several other DCU players are playing well enough to look like MVP candidates on a lot of the league. Even better, the team as a whole is just clicking. Upcoming game: RSL @ DCU
2. Chivas USA (40 pts, 22 games, 1.81 ppg)—With a draw against DCU, a solid win over the Red Bulls, and a very soft schedule over the remaining eight games on their schedule the Goats are looking very good too. I expect to make a play for the Supporters Shield this year. Shoot, even their Sueno player is looking pretty good. Upcoming game: LAG @ Chivas USA
3. New England Revolution (42 pts, 24 games, 1.75 ppg)—Even dropping their last game to DC United, the Revs are looking pretty good. Their remaining schedule looks pretty soft too, though they have fewer games remaining than Chivas. Upcoming game: FC Dallas @ New England
4. Houston Dynamo (42 points, 25 games, 1.68 ppg)—They toughed out a win over RSL, and face RSL and the Galaxy (twice) over the rest of the season. They will be challenged by Chivas and FC Dallas games though. I don’t see them repeating as MLS cup winners. Upcoming game Dynamo @ LA Galaxy
5. FC Dallas (39 pts, 23 games, 1.69 ppg)—They beat up a sinking TFC and Denilson is looking pretty good, but it doesn’t look like enough. I think they’ll settle for four points from the next five games, and probably only eight more the rest of the season. That may be enough to dump them into the second tier of teams. Upcoming game: FC Dallas @ New England

The Fair

The next group of teams are battling it out for the playoffs (and a strong finish by on or two of them might even claim a spot in the Super Liga).

1. Chicago Fire (29 points, 23 games, 1.26 ppg)—Nine points in their last five games and a 7-4 aggrate score in those games makes this team look a lot better than their overall standing. Wanchope and Blanco have really been a boost to this team, and could push them into the top tier with continued solid play. Upcoming game: Red Bulls @ Fire
2. New York Red Bulls (36 points, 24 games, 1.5 ppg)—Injuries and recent results make this look like a team on the wane. It’s hard to count them out though with Jozy and JPA up front. Having RSL, TFC, and the Galaxy in the mix of their upcoming games should help them stack up some points though.
3. Kansas City Wizards (33 points, 25 games, 1.32 ppg)—Another team that’s dropping rapidly. A scary schedule that includes Chivas, DC United, the Red Bulls, and FC Dallas doesn’t bode well either, but I think they’ll be the eighth and final playoff team (will anyone come to watch their home games though?). Upcoming game: Columbus Crew @ KC Wizards
4. Columbus Crew (28 pts, 24 games, 1.16 ppg)—Three winnable games to finish out September then a tough October schedule don’t help the Crews chances of getting into the playoffs. I expect them to be on the outside looking in. Upcoming game: Columbus Crew @ KC Wizards.
5. Colorado Rapids (28 pts, 24 games, 1.16 ppg)—Two games against Chivas USA, one against the Revolution, and two against a resurgent RSL, and one against lowly TFC might see the Rapids only take six more points the rest of the year (they might not even keep the Rocky Mountain Cup this year). If that happens, Clavijo might not survive the off-season. The Rapids are on the verge of falling down to the bottom tier of teams. Upcoming game: Chivas USA @ Colorado

The Bad

If only the MLS and USL could work out relegation and promotion, there might be a reason to watch these guys finish out their season. These teams are just bad right now, though there may be some hope on the horizon—well at least the 2008 season.

1. Real Salt Lake (18 pts, 22 games, .81 ppg)—Their horrid start doomed RSL this year, but a coaching change and a roster overhaul are starting to pay dividends. RSL will play the spoiler this year, and should compete for a playoff spot ext year. Upcoming game: RSL @ DCU
2. Los Angeles Galaxy (17 pts, 20 games, .85 ppg)—Too many injuries and a terrible schedule are too much for the Galaxy to overcome. The good news is they won’t finish last in the league, but they won’t come close to being the super-team people expected. Upcoming game: Houston Dynamo @ LA Galaxy
3. Toronto FC (20 pts, 23 games, .86 ppg)—Except for a brief, shining moment, these guys have looked every bit the expansion club. With an injury list that might have them starting recruits from the PDL, things don’t look to improve anytime soon. At this point, just scoring again might seem like a victory. Upcoming game: RSL @ TFC

Week Twelve Scariness Rankings

Some really big changes this week, and some huge variance from the traditional power rankings. A big loss or a big win (especially over a team that had been doing well) really shows up here. Two tables for your viewing pleasure—first, the Adjusted Scariness, and second, the original Scariness. This time I’ll let you draw your own conclusions (though I’d sure like to hear them).

Rated by Adjusted Scariness

Team	Adj Scariness	Ranking (last week)
Houston	4.33	1 (4)
New England	2.49	2 (8)
DC United	2.13	3 (2)
Columbus	2.07	4 (10)
Salt Lake	1.62	5 (11)
Dallas	1.56	6 (9)
Toronto	1.49	7 (3)
New York	1.48	8 (6)
Chivas	1.31	9 (1)
Kansas City	0.91	10 (5)
Los Angeles	0.64	11 (7)
Chicago	0.36	12 (12)
Colorado	- 0.12	13 (13)

Rated by Scariness

Team	Scariness	Adj Scariness (diff)
Houston	3.00	4.33 (1.33)
Dallas	1.93	1.56 (- 0.37)
Chivas	1.91	1.31 (- 0.60)
DC United	1.82	2.13 (0.29)
Columbus	1.8	2.07 (0.27)
New England	1.64	2.49 (0.85)
Toronto	1.31	1.49 (0.18)
New York	1.22	1.48 (0.26)
Salt Lake	1.13	1.62 (0.39)
Kansas City	1	0.91 (- 0.09)
Los Angeles	0.89	0.64 (- 0.25)
Chicago	0.76	0.36 (- 0.40)
Colorado	0.33	- 0.12 (- 0.45)

Week 10 Scariness Ratings

I did fairly well on last week's pick, and I considered my adjusted scariness numbers¹ in making them, so I’m feeling better about them.

Adjusted Scariness

Team	Adj Scariness	Rating (Last Week)
DC United	3.11	1 (3)
Chivas	2.85	2 (4)
Kansas City	2.74	3 (1)
Houston	2.73	4 (6)
New York	1.86	5 (2)
Dallas	1.78	6 (12)
New England	1.61	7 (7)
Salt Lake	1.46	8 (9)
Toronto	1.18	9 (5)
Colorado	1.01	10 (7)
Chicago	0.62	11 (11)
Los Angeles	0.39	12 (10)
Columbus	- 0.25	13 (13)

One thing that I’m getting used to is that there’s a lot of movement on these charts each week. I’m still a bit puzzled by some of the numbers (why in the world is Chivas rated number two??) but overall, I think I’m coming to peace with things.

¹ The Adjusted Scariness probably deserves a bit of explanation:

• For each game, I add up the team’s points from a game, the opponents PPG average for the last five games, and add a bonus point for playing on the road, then divide by two to create a value for the game.
• I add the goal differential to the value for each game, to create a modified value.
• Finally, I take a weighted average of the last five modified values (using weightings of .5, .7, .9, 1.1, and 1.3).

Week 9 Collective Rankings and Scariness

Jeff Bull has released his Week 9 Collective Rankings, so it’s time for me to take a look at how the my new and improved scariness ratings compare.

Team	Scariness Rating	Collective Ranking (and difference from Scariness)
Kansas City	1	1
New York	2	2
DC United	3	4 (+1)
Chivas	4	8 (+4)
Toronto	5	7 (+2)
Houston	6	9 (+3)
Colorado	7	5 (-2)
New England	7	3 (-4)
Salt Lake	9	13 (+4)
Los Angeles	10	11 (+1)
Chicago	11	10 (-1)
Dallas	12	6 (-6)
Columbus	13	12 (-1)

A teams scariness rating averages 2.23 points off of their position in the collective rating, which seems like a pretty big skew. A lot of this has to do with the fact that I’m not really rating a teams power—just how they’ve done against recent strength of schedule (and a bit of fudging for playing on the road). Because I weight recent results more heavily, there’s going to be a lot more fluctuation. For example, here are the scariness scores and ratings updated for last night’s game.

Team	Scariness	Scariness Rating (and previous)
New York	3.14	1(2)
Kansas City	2.74	2(1)
DC United	2.35	3
Chivas	2.18	4
Houston	1.65	5(6)
New England	1.61	6(7)
Colorado	1.61	6(7)
Salt Lake	1.46	8(9)
Toronto	1.18	9(5)
Los Angeles	1.16	10
Chicago	1.00	11
Dallas	0.80	12
Columbus	-0.12	13

New York piled on an extra .52 points (and rose a slot in the ratings), while Toronto lost .97 points (and dropped a whopping 4 slots). Is that fair? I think some TFC fans might tell me it’s not, but I think it is.

These ratings still don’t reflect personel changes. Is Kansas City without Eddie Johnson just as scary as they were in beating the Red Bulls last weekend? Not even close, but I’m not really sure how to adjust the numbers to reflect for players who are injured, sitting out a game due to penalties, or playing for their national team.

Scariness For Jun 5 (and a little twist)

Well, I got an 0-fer on my last set of picks, which is so depressing I won’t even bother to go over them (though, RSL did pull out a tie that I’m pretty happy about). Instead, I thought I’d spend some extra time on my scariness ratings. This time I present the older style along with a new (improved?) version.

Traditional Scariness

Team	Scariness	Rating
DC United	2.2	1 (+1)
Kansas City	2.06	2 (+2)
Toronto	1.96	3 (+4)
Colorado	1.92	4 (-3)
New England	1.86	5 (0)
Dallas	1.62	6 (0)
New York	1.42	7 (-4)
Houston	1.2	8 (+4)
Chivas	1.14	9 (-1)
Chicago	0.94	10 (+3)
Los Angeles	0.92	11 (-1)
Salt Lake	0.8	12 (-1)
Columbus	0.4	13 (-4)

Adjusted Scariness

Team	Adj Scariness	Rating	Diff from Scariness Rating
Kansas City	2.74	1	+1
New York	2.62	2	+5
DC United	2.35	3	-2
Chivas	2.18	4	+5
Toronto	2.15	5	-2
Houston	1.65	6	+2
Colorado	1.61	7	-3
New England	1.61	7	-2
Salt Lake	1.46	9	+3
Los Angeles	1.16	10	+1
Chicago	1	11	+1
Dallas	0.8	12	-6
Columbus	-0.12	13	0

The Adjusted Scariness probably deserves a bit of explanation:

• For each game, I add up the team’s points from a game, the opponents PPG average for the last five games, and add a bonus point for playing on the road, then divide by two to create a value for the game.
• I add the goal differential to the value for each game, to create a modified value.
• Finally, I take a weighted average of the last five modified values (using weightings of .5, .7, .9, 1.1, and 1.3).

This takes a little bit longer, but gives a clearer picture (I think). There are still some surprises there (how did Chivas and RSL climb so high?), but I think I’m getting closer. I’m still looking for feedback, so please let me know what you think. (Oh, this is getting hairy enough that I don’t expect to do it for any other leauges.)

May 31 Scariness Rankings

I’ve not seen Jeff Bull’s latest collective rankings, so I can’t compare them this time around. Instead, I’m providing the my scariness score, and the corresponding rating (along with the change versus last week’s rankings).

Team	Scariness	Rating (change)
Colorado	2.28	1 (4)
DC United	2.16	2 (3)
New York	2.04	3 (-)
Kansas City	1.96	4 (-)
New England	1.96	4 (-3)
Dallas	1.8	6 (-4)
Toronto	1.38	7 (-1)
Chivas	1.14	8 (1)
Columbus	0.92	9 (1)
Los Angeles	0.76	10 (-2)
Salt Lake	0.62	11 (1)
Houston	0.54	12 (-1)
Chicago	0.24	13 (-6)

This means that the average change in scariness rankings was just over two slots, with big changes occuring all over the board.

Some of these rankings look a bit off, as I mentioned in my post A Bit More on Scariness. I’m still trying to come up with a way to clean them up—I’m open to suggestions.

(Oh, by the way, the USL1 and the PDL Northwest Division still have teams without enough games played to fill out their tables—maybe next week.)

A Bit More on Scariness (Including Last Night's Game)

Here’s a look at how much of a change you might see on a game by game basis.

New York’s win raises their scariness rating by .1 point, which meant that they swap positions with Dallas. In this case, I think my system underrates the scariness of the Red Bulls with JPA but as he plays more games, and they get results, it gain accuracy.

Team	new rating	last week
New York	2.04	1.94
Dallas	1.98	1.98

Chicago’s loss hurt them a lot more. The dropped .54 points, which means they slipped below LA, Chivas USA, Columbus, and Houston. They’re now only .06 points scarier than Real Salt Lake—which makes me feel a bit better about my most recent picks

Team	new rating	last week
Chicago	0.48	1.02
Salt Lake	0.42	0.42

There are a several things that should still be worked into this to create a better scariness rating:

• Goal differential (but on a per game or a last 5 basis?).
• Opponents scariness (for teams that win or draw)
• Fatigue (when a team has played an extra game already this week (e.g., both RSL and the Fire will be playing their second game of the week on Sunday, so they should both be considered ‘less scary’ than normal)

Are there other things that ought to factor in here? Does the rating make sense so far?

MLS Scariness Ranking

In my post on USL1 power rankings, I used a simple ‘points per game’ (PPG) to rank teams instead of the even simpler ‘total points’ used on most standings reports. I mentioned that I wasn’t completely happy with it though, and that I might work on something using a weighted average. Well, I finally put it together, and have ranked the MLS and USL1, and PDL (Northwest Division) according to it.

I looked at the teams’ last five league results, and gave each result a weighted score. The earliest was points x 0.8, the next was points x 0.9, and so on, until the most recent game which was points x 1.2. I then divided these by the sum of 0.8, 0.9, 1.0, 1.1, and 1.2 (which is 5). The resulting value gives an approximation of how well the team has played recently, with the most emphasis on the most recent games. I call this a team’s ‘scariness rating’.

Here’s what the MLS looks like when compared to the Collective Power Ranking put together by Jeff Bull.

Team	Scariness	Collective Power Ranking
New England	2.64	1
Dallas	1.98	3
New York	1.94	2
Kansas City	1.86	4
Colorado	1.58	5
DC United	1.58	6
Toronto	1.26	11
Chicago	1.02	7
Los Angeles	1.00	8
Chivas	0.94	10
Columbus	0.76	12
Houston	0.60	9
Salt Lake	0.42	13

Because some teams in the USL1 and PDL (Northwest) haven’t played five games yet, I’m not going to post them yet. As the tables are fleshed out, I’ll add them to a ‘Scariness Ranking’ post each Wednesday. I’ll also try to remember to add the scariness rankings to teams where appropriate (like my predictions posts). If you’d like to grab them and use them yourself, please feel free—just point people here for an explanation of what they mean.