05 DA National Rankings

[Kante]

have been asked several times about applying the algo to create national rankings. so, made the algo earns its money and had it run more than 8,500 simulated matches to produce the Top 25 ranking below. Also included below are the southwest group team national rankings, even if they're not in the Top 25.

The inputs todrive the algo rankings include DA Cup matches, regular season matches and GA Cup matches. Leveraging the GA Cup results, also created and applied a group strength/rpi-esque discount i.e. not all groups are equal (looking at you, southeast and midwest...) and so a win/goal scored is one group is not the same as a win/goal scored in another group. Northwest was the strongest group offensively for goals scored, and Southwest was the strongest group defensively for goals allowed.

The ranking below is not based on a team's complete record, but, instead, where that team is right now and their projected results if a match were held today. There are several highly ranked teams that started slow (to be fair, some of this may have been better competition via DA Cup) but then figured some things out and started to roll. Examples of this include FC Dallas and Real Salt Lake.

If folks are interested in more info/background, just send a dm for more detail on methodology.
_________

Without further adieu, here's the top 25 (fyi, TDS Top 25 November Ranking is in (parens))

1. LAFC (3)
note: putting aside ties to SD Surf and FCG, and a GA cup loss to FCD, LAFC is setting the standard. Algo says that, in a rematch against FCD, LAFC wins by .48 goals, and that LAFC beats both SD Surf and FCG by three goals in a more wide open rematches. LAFC has the 3rd best defense in the nation, and the second best offense, and have frequently put that quality on display against an array of highly ranked opponents. LAFC has added some offensive firepower this season and only had one player called up to the 05/u15 ynt, so these boys may have something to prove this season.

2. FC Dallas (4)
note: FC Dallas started a little slow, with an early season 0-3 loss to InterMiami and later tied Vancouver 1-1 at the GA Cup. However, FCD is also the only team in the country to beat LAFC. And FC Dallas also had seven players called up to the u15/05 ynt to go to Poland. And their defense is the 2nd best in the country while their offense is fourth best. So why isn't FC Dallas ranked 1st? Fair question. One word. Strengthofschedule. FCD just hasn't played nearly as many high quality opponents with as solid of results as LAFC. Send your letters and comments to scheduling@ussda.com.

3. Seattle Sounders (1)
note: so close to greatness and yet so far. The Sounders tied Real Salt Lake (but played a man down due to injury for some of the 2nd half), tied Colorado Rapids 1-1 after, per game report, the Rapids got a goal in the 1st minute and lost to the Sacramento Republic 1-2 (three Sounders were playing w/ the ynt that weekend). Meanwhile their wins haven't quite been as dominating as FCD and LAFC. Hopefully, these three top teams will make it thru the qualifying round of the GA Cup and then get a chance to see who's got what.

4. Real Salt Lake (9)
note: for most of the RSL team, this is their first year playing together and it showed at the beginning of the season with ties against Portland Timber and the Colorado Rapids, and losses to Vancouver and TFA. However, things seem to be coming together a bit, and RSL now has the #1 defense in the country with five shutouts in the their last eight matches. Offensively, RSL are still finding their legs, and their offense is only at #20 in the country, so more work/creativity needed there.

5. BW Gottschee (n/a)
note: this Queens, NY is the #1 non-MLS DA team in the country, and ranked ahead of northeast regional peers like NYCFC, New York Red Bulls and the Philadelphia Union. The caveat is that BW Gottschee has yet to play any of those teams, so this ranking may be a little bit of mirage. Will see.

6. Dallas Texans (n/a)
7. New York Red Bulls (5)
8. San Jose Earthquakes (2)
9. Barca Academy (TDS: (n/a)
10. Philadelphia Union (8)
11. Chicago Fire (12)
12. Toronto FC (n/a)
13. Sacramento Republic (n/a)
14. United Fubol Academy (n/a)
15. Montreal Impact (n/a)
16. San Antonio (21)
17. Real SoCal (n/a)
18. Baltimore Armour (11)
19. Cedar Stars - Bergen (n/a)
20. Queen City Mutiny (but #1 for best club name in the country...) (n/a)
21. Chargers (n/a)
22. Colorado Rapids (15)
23. Sporting KC (16)
24. Houston Dynamo (14)
25. NYCFC (10)

Other Southwest Group Team National Rankings (again, there's 93 total 05/u15 DA teams):

32. FC Golden State (#32 offense; #36 defense)
38. TFA (#25 offense; #64 defense) note: TFA has been tough to predict all year since they're playing a number of 06s up with frequency (good on them btw). For example, the last match of 2019, where TFA lost to the Strikers 3-4, TFA played six 06s significant minutes in that match. and yes, am aware TFA's combined offense/defense ranking is worse than LAG's. The algo makes the ranking by running simulated matches, and in those simulations, TFA has a very slight edge over LAG, hence the overall higher ranking.
44. LA Galaxy (#18 offense; #61 defense)
45. LAGSD (#52 offense; #14 defense)
51. Nomads (#30 offense; #68 defense)
53. Strikers (#28 offense; #78 defense) note: almost all their overall lower ranking is due to the Strikers defensive trend line. In their last four matches, the Strikers allowed an average of 3.25 goals per match, whereas they only 1.17 goals per match in their first six matches.
68. Pateadores (#21 offense; #86 defense)
71. LAUFA (#55 offense; #82 defense)
81. Albion (#90 offense; #59 defense)
93. Arsenal (#92 offense; #67 defense)

 

[texanincali]

I think your algo must be pretty good. Your list matches the eye test as well - in most cases.

I’ve seen most of these teams play, and the only change I’d have is FCD over LAFC. Mainly down to beating them twice in 6 months.

 

[Kante]

I think your algo must be pretty good. Your list matches the eye test as well - in most cases.

I’ve seen most of these teams play, and the only change I’d have is FCD over LAFC. Mainly down to beating them twice in 6 months.

thanks. Fair point.

The first loss would have been before (?) LAFC added SR from the Strikers, and he's been good 1.5 goals per 80 minutes played.

The 2nd 0-1 loss at the GA Cup is more difficult to reconcile.

have been trying to track down video of that match to take a look, but can't find any, so am defaulting to strength of schedule, which favors LAFC.

Any insight into why so many FCD players (7?) were called to the YNT u15 trip to Poland? That was the other data point that gave pause.

Also any insight into Dallas Texans and San Antonio? Do they look about right?

The weak spot in the algo is that it can only work with available data.

This, right now, effectively puts the foot on the scale for non-MLS teams since, in most cases, their schedule to date hasn't been as challenging as the MLS teams' schedules.

 

[Kante]

have received feedback, had a chance to mull and do some bug hunting/troubleshooting.

below are revised rankings, and include rankings for all 93 DA teams, overall, best offense and best defense.

​

Here's ranking user guide:

• The rankings are based on several iterations now of more than 8,500 simulated match results where every team played every other team. For reference, have also included the TDS rankings as of November 2019.
• The best offense measure "normalized GS%" means that it's the GS% - goals scored by a team in particular match as % of average goals allowed by particular team - trended over time and normalized/discounted by team group - i.e./e.g. a goal scored against a Southwest team (the best defensive group) is worth more than a goal scored against an Atlantic team (the worst defensive group).
• The best defense measure "normalized GA%" is similar and is the GA% - goals allowed by a team in a particular match as % of average goals scored by particular team - trend over time and normalized/discounted by team group.
  • The GD% is goal differential %, and is simply GS% minus GA%.
    • A positive value for GS% is good, and means that the team scores more goals than other teams typically allow. Conversely, a negative value for GA% is good and means that teams allow fewer goals than other team typically score. A positive value for GD% is also good. More simply, the text in red is bad, the text not in red is good.
  • The phrase "in next game" means that the % scores are if each team played today, how would they do today at this point in time right now.
• There are some instances where one team looks like they're better than another, higher ranked team. For example, Arsenal looks like they should be ranked higher than Breakers. In all of these instances, the algo ran the simulated match and the higher ranked team had the goal differential edge in the simulated match.
• The only instance where there was tinkering with the algo's output is the Nomads and the Strikers. The algo had the Nomads ranked ahead of the Strikers. The two teams played about a month ago, and the Strikers won 4-0, scoring all four goals int he first half. So, the Strikers are ranked ahead of the Nomads.
• Some of the rankings also may have changed from what was originally posted due to fixing bugs and double-checking the work/math.

As always, if there's questions, post below or send over a DM.

Here's best overall rankings for all 05/u15 DA teams:




Here's best offensive rankings for all 05/u15 DA teams:



Here's best defensive rankings for all 05/u15 DA teams: