BetFlux launches soon, and this is your chance to get in on the ground floor — the founders and engineers building it are right here in the room with you. Bring what you're working on, ask the hard questions, and help shape what we build next. There's nothing else like this.
Spent all morning on this — my NHL backtest keeps grabbing two closing lines per game and it's wrecking my CLV numbers. Am I querying it wrong?
Not you — postponed games get a second close once they're rescheduled. Filter to the latest close per game and they collapse to one. Dropping a query that does it in #data-and-sources now.
That was it. Hours of head-scratching, fixed in five minutes — no other data vendor would've just told me that.
That's the whole reason we're in here. Stuck on anything else, tag me.
what this isn't
Want a result? Share the process that got you there.
why join
Most betting Discords are picks and screenshots. This one is the team behind BetFlux working alongside you — on how you model, how you source data, and how you actually beat the close.
Share how you build and pressure-test how others do — feature design, model selection, what's actually predictive versus what just backtests well.
Sourcing, pipelines, storage, normalization. The unglamorous plumbing that turns a clever idea into a model you can actually run every day.
CLV, line movement, market timing. A great model still has to beat the close — the room talks about how to get filled at the number that matters.
Ask hard questions and get straight answers from the founders and engineers — not a support queue or a community manager. Hands-on help you won't get from any other betting product.
If you want lock-of-the-week, this isn't it. Want to share a result? Share the methodology that got you there, or don't share it at all.
Hit a rough edge? Want a feature? The people who can change the product are in the room — and they ship. Tell us what you need and help shape what we build next.
how the room works
Post how you model, where you source data, and how you beat the close — and get honest, specific critique back. From peers who do the work, and from the founders and engineers building the data underneath it. Bring a result and we'll dig into the process that got you there.
Rebuilt my puck-line model from scratch. Features: rest days, travel, goalie xGA, pace. Backtested two seasons against BetFlux closing lines — +3.1% CLV over 412 bets. Where's it overfit?
Goalie xGA over how many starts? Small samples will lie to you — I'd regress to league mean under ~15 starts or you're just modeling noise.
Show CLV by month, not aggregate. Variance hides in the average. If it's all from one hot March you don't have an edge yet — you have a story.
Travel as raw miles or a back-to-back flag? When I tried both, the B2B flag carried almost all the signal. Raw miles was mostly collinear noise.
inside the server
welcome
What BetFlux is, and what this room is for — in one read.
start-here
Three rules and the pinned resources to get oriented fast.
introduce-yourself
Your sport, what you're building, and what you're stuck on.
ask-the-devs
Stuck on something? The founders and engineers answer here — usually same day.
general-modeling
Methodology, approaches, and honest model critique.
data-and-sources
What data people use, where it comes from, and how it's wired up.
betflux-feedback
Bugs, feature requests, and things you wish the tool did.
announcements
Product updates, new features, and the changelog.
Cookies
A few cookies keep you signed in, and the rest are analytics and marketing that show us what's working. You can accept, reject, or customize, and you can change your mind anytime from the footer.
Drop your email to download normalized, graded sample datasets for every sport right now.
No spam. Unsubscribe anytime.
Building a model? Our Discord is where you talk shop with the founders and other sharp bettors. No picks, no touts — it's free.
Join our Discord