With the MLS Championship finally at a close, we can turn our attention to preparing for the new year. There are a lot of things that we already know, and a lot of things that we still need to wait to find out. Sadly, that means there are a lot of things we need to wait for in order to lay out our preparations in full.
What we don’t need to wait on is the new starting points for existing leagues, and laying out the CONCACAF Champions League.
CONCACAF Champions League
I had mentioned previously that I would be handling the Champions League starting points differently. I was very dissatisfied with having all the non-Canada/US teams with the same starting point (1442.980 in 2018), especially given the large variance in quality between top Mexican clubs, and the CONCACAF League Champions.
As a result, I came up with a way to “rebalance” the starting points based upon the starting average of the Canada/US qualifiers. As CONCACAF has its own points scale, based upon the most recent 5-years of data, that provides a good basis for weighting.
The average elo rating of the 5 Canada/US teams was 1469.519. The average CONCACAF Champions League “points” for those same 5 teams was 69. To calculate a team’s starting elo for this tournament, I use the following formula:
x = 800 + (C/Y*(Z-800))
C = CONCACAF Champions League “Points”
Y = Canada/US Team Average CONCACAF Champions League Points
Z = Canada/US Team Average Elo Rating
As a result, we can see a top club like Monterray, who holds the MEX3 slot and has 116 CONCACAF points, will end up with a starting elo of 1922.315, whereas Herediano, which enters from the CONCACAF League with 5 CONCACAF points, will start with a meager 848.376 points. A win against a Mexican juggernaut will be worth more points than one against a Caribbean minnow, just as a loss to a Caribbean minnow will hurt more than losing to a Mexican juggernaut.
This format is going to be a lot more accurate than applying a flat starting rate, and ramps up the CONCACAF League (CL) representative as teams in that league become more accustomed to a higher level of competition than they may be used to seeing. How does it ramp up? Currently, the CL rep has 5 points, but they gain a minimum of 4 points per year up to 5-years of competition. Meaning if Herediano were to lose all their matches this year, they would still increase their CONCACAF Points to 9, thereby giving the 2020 CL representative a higher comparative elo rating when they enter the competition.
Starting Points (Existing Leagues)
Although there are quite a few new leagues this upcoming year, the exact format and level of those leagues remains to be seen (more details in next section).
What we can talk about is the starting points for existing leagues. Every year I get the opportunity to take another look at how the leagues have shaped up. We’ve seen large modifications in the past (when USL increased to 1225 from 1200), and we’ve seen a couple that I’ve held back because of insignificant difference.
First we need to look at how each league ended the year, and adjust to “rebalance” MLS at the 1400-point cap:
| League | 2018 League | 2018 New Team | 2018 End | 2018 Adjusted |
| PCSL | 800.000 | 800.000 | 800.000 | 800.000 |
| L1O | 899.630 | 900.000 | 894.609 | 895.670 |
| PLSQ | 900.000 | 900.000 | 903.202 | 904.263 |
| NPSL | 1000.000 | 1000.000 | 1000.028 | 1001.089 |
| USL-2 | 994.936 | 1000.000 | 994.404 | 995.465 |
| USL-C | 1226.282 | 1225.000 | 1225.818 | 1226.877 |
| MLS | 1398.900 | 1400.000 | 1398.939 | 1400.000 |
As we can see, League1 Ontario (L1O) under performed compared to an over-performing PLSQ. Additionally, USL-2 (formerly PDL), maintained its competitive level, which was below the previous benchmark they were granted. Other leagues were within 2.5 points of their league benchmarks.
| League | 2019 New Team |
| PCSL | 800.000 |
| L1O | 895.000 |
| PLSQ | 905.000 |
| USL-2 | 995.000 |
| NPSL | 1000.000 |
| USL-C | 1225.000 |
| MLS | 1400.000 |
As a result, I’ll be altering 3 benchmarks this year. L1O will have their “new teams” benchmark altered to 895.000, PLSQ will grow to 905.000, and USL-2 will have theirs reduced to 995.000.
Once departing teams are removed and new teams are added, the league averages shift, but the starting point is based on the end of the previous year’s elo averages. Not all team movements have been announced for lower division leagues, so I won’t get into the new league averages until after schedules are launched in the new year. What’s important is knowing where new teams will start on the table.
An interesting thing to consider here is how the leagues now compare to one another. With the change in starting point, League1 Ontario now stands slightly below PLSQ is quality. This makes a lot of sense when you consider that the CPL will be raiding the league’s top players for their inaugural 2019 season. That said, I can see these two leagues waffling back and forth depending on the format and outcome of the Voyageur’s Cup. Again, this makes a lot of sense considering the importance of the competition.
Just as interesting, the change in position of USL-2 in relation to USSF’s 4th-division NPSL. Based on the data, the NPSL has managed to remain level in US Open Cup, while USL-2 sides have suffered slightly. This will probably change in the upcoming year as some of the top-NPSL sides are splitting off to compete in NPSL Pro/Founder’s Cup, but we await full details on that competition (including Open Cup entry information).
Match Weights
There is no change in how match weights are determined. The math behind it, as originally laid out in 2017, allows for a lot of modifications on the fly if the competition structure changes.
There are some minor modifications from what was posted in the 2018 update, as the MLS altered their playoffs to have single-match knock-out round prior to a home-and-away quarter & semi-finals.
This alteration was taken into account while calculating the elo ratings during the playoffs, which demonstrates how simple it was to fix the weighting on the fly.
I anticipate that we will see the match weighting tweaked this upcoming year. I feel it is unlikely the Voyageur’s Cup will remain a home-and-away competition. Additionally, I am uncertain whether the MLS will continue with its latest playoff format. There are a lot of changes that could occur between now and the end of the season.
17 Dec 2018 Edit:
The MLS has announced a new playoff structure, moving entirely to a single-elimination tournament. As a result, all MLS Playoff matches will have a weight of 35, Conference Finals will have a weight of 40, and the MLS Final continues to be weighted 65.
Other Known Changes
A couple of other key things will be changing.
First, I will only be tracking the cumulative ratings. Performing these ratings is pretty time-intensive, despite a lot of the automation that I’ve built into my excel spreadsheets. Given that there are at least 3 new leagues (CPL, USL-1, NPSL Pro/Founder’s Cup), and the possibility of a 4th (NISA), I need to trim back on the time spent performing calculations.
Additionally, if NISA launches with the season format they’ve been talking about, they will be running through the winter, with some matches in 2019, and some in 2020. Depending on when the season would end, this would cause issues, especially if it runs into the Champions League season. As a result, it’s a better idea to bypass the whole issue by sticking with a single, rolling set of ratings.
Secondly, I’ll be drastically changing the layout of the elo rating article.
I’ll be starting the article by highlighting the top Canadian-teams by league. I’ll look at how the team managed over the past month, who they surpassed to take the spot, and who is threatening them from behind.
This will then be followed by looking at the biggest movers in and shakers. This will be focusing on the teams that moved up and down the most places in the ratings, as well as the teams that moved up and down the most elo rating points.
This will then be followed by any news on the Canadian Championship, including updated competition radials (if possible). As the exact format of the Voyageur’s Cup remains to be seen for 2019, this could get changed around quite a bit as we roll it out. This section could be switched for (or in addition to) L1O Cup, PLSQ Cup, or play-off coverage, with the appropriate radials, if it is appropriate at the time.
Kicking off before the Canadian Championship, we have the CONCACAF Champions League. Our patreon’s already have a copy of the first 2019 Champions League radial, and will receive a “predictive” version after the first match of each round is completed. For those who aren’t patreons, our monthly elo-articles will include a version updated for the most recently completed round. Expect the first one sometime in January 2019.
We’ll then have the full ratings. No longer will there be two sets of large vertical charts. Instead, we’ll be looking at a single chart, broken up in three sections, spanning the width of the screen. Patreons will have a peak at this chart (and the potential starting points for the new leagues) this upcoming Friday. Everyone else will need to wait for the 2nd part of this article.
Thirdly, teams that aren’t big movers or top of the ratings for their league, will have precious little written in these articles. Let’s face it, most people would have skimmed over those teams anyways.
To Be Determined
Sadly, there’s a fair bit of information that has yet to be disclosed about the new leagues.
- Canadian Premier League
- Season Format
- Play Off Existence/Format (Rumours say None)
- Salary Cap/Realistic Level of Play (Latest rumour $750k, above USL-C, below MLS)
- Canadian Championship
- CONCACAF League / CONCACAF Champions League qualification
- NPSL Pro/Founder’s Cup
- Season Format
- Play Off Existence/Format
- Realistic Level of Play
- USSF App
- Open Cup 2019 Qualified?
- NISA
- Teams
- Season Format
- Play Off Existence/Format
- Salary Details
- USSF Approval Status (Rumours Say Good-to-go)
With the information currently available, I can confidently put together starting ratings and match weights for USL-1. I can make a loose guess for CPL, but it would be based mostly on rumours at this point. Hopefully some more details will be revealed swiftly.
There isn’t much of a priority for the Founder’s Cup or NISA, since they don’t kick off until August (Founder’s), or sometime in the fall (NISA). That being said, we don’t know if Founder’s Cup teams will be qualified for the 2019 US Open Cup, which would make knowing that information a little more pressing. It is possible they may be granted some of the NPSL’s slots, for those teams that competed in 2018.
NISA sides should not have entry into the 2019 US Open Cup as they will not have played any league matches by that point. The NASL-NPSL qualification match that was put into play in 2018, should not apply to these clubs for the 2019 Open Cup. That being said, we’ve seen stranger things happen, and I wouldn’t be surprised if USSF opts to skip this lawsuit given its low priority and the number with which they’re already inundated.