Standings & projected finish
Results
Power rankings
Who beat who
Upcoming games
How accurate are the predictions?
How the predictions work
Team ratings
Each team gets an attack and a defense rating, fit to every completed game with a Poisson goals model, the standard approach for predicting soccer scores. Expected goals for the home side = league average × home edge × home attack × away defense weakness, and likewise for the away side. Home teams get more goals, the average edge across four past ECNL seasons. The goals model on its own expects more draws than youth games produce, so draws are scaled to the rate seen in past seasons. Each team starts the season from last season's rating, at three-quarters strength: the club's team for the same birth year, one age group younger (for U13 and U18/19, also the club's team at this age). This season's results take over as games are played. Replaying four past ECNL seasons, it lifts correct picks over the first five weekends from 56% to 62% (boys) and 58% to 67% (girls).
From ratings to odds
From the two expected-goal numbers the model works out the chance of every scoreline (0–0, 1–0, 2–1, …), which gives the win, draw and loss chances and the most likely score. A small Dixon-Coles adjustment makes 0–0 and 1–1 a bit more likely (and 1–0 and 0–1 a bit less), because plain Poisson models under-predict draws; its strength is estimated from the results so far (currently ). The pick is the more likely winner, or "too close to call" when the two are within 5 points; the margin is the expected goal difference, rounded. The projected finish plays out every remaining game 10,000 times using those chances. Each run also redraws every team's rating within how sure the model is about it, which is wide after a few games and narrow late in the season, so early-season chances don't swing as hard. Checked on two past seasons, this made the playoff and title chances more accurate, most of all early on.
The standings
Tables follow ECNL's rules: 3 points for a win, 1 for a tie, ranked by points per game, then head-to-head (two-team ties only), total goal difference and total goals scored. After every update all tables are checked against ECNL's official standings. On October 2, 2026 every record and position matched for all 1,812 teams.
National rankings
Conferences don't play each other in the regular season, but they meet at ECNL RL showcases (Arizona, Florida, Virginia, Texas, California, North Carolina) and the RL playoffs and finals. Those games set each conference's strength, starting from last season's and updated as this season's showcases are played. A team's national rating is its rating inside its conference plus its conference's strength. Checked on past seasons, this predicted games between conferences better than treating every conference as equal (correct picks about 50% → 55%). Replaying 2022–23 to 2025–26, the higher-rated team won 69% of the decisive games between conferences, against 64% for the team with more points per game, the order ECNL seeds its playoffs by. The rating counts goal margins in full: counting them only up to 3 goals predicted those games slightly worse. A grey rank is provisional: fewer than 6 games, so last season's rating still counts for a lot.
Postseason chances
Each simulated season ends with ECNL's official 2026–27 qualification rules for that conference and age group: how many teams reach the Champions League (ECNL National Playoffs) and, for the girls, the North American Cup and Showcase Cups. That includes the top teams in each group, wildcards and play-ins. A wildcard spot goes to whichever of two conferences' teams has more points per game, so the two conferences' seasons are simulated side by side. League Cup winners can't be predicted from this data, so a spot that the League Cup winner can take is treated as going to the team in the table. A team headed for a play-in game (a single game against another conference's team for a Champions League spot) counts under Postseason, not in the Champions League chance, since that game decides the spot.
How it was tested
Every setting was chosen by replaying past seasons one weekend at a time: each weekend is predicted only from the games played before it, then compared with what happened. That's 113,264 games in all: 56,734 ECNL games (boys and girls, 2022–23 to 2025–26), 32,306 ECNL RL games (2025–26) and 24,224 MLS NEXT games (Homegrown 2023–24 to this season so far, and Academy this season). The model's most likely result happened in about two games out of three (66%) in all three leagues, against 45–47% for always picking the home team and 33% for a blind guess. A change was kept only if it made the predictions better in every season tested; that's how the draw rate, the home edge and the starting ratings were set (October 2026). “How accurate are the predictions?” above runs the same check on this season's games.
What it leaves out
Not included: roster changes, injuries, travel, weather, out-of-conference events, and the fact that youth results are noisy. Treat these numbers as a guide, not a guarantee.