Props Betting
The bets our models find, sent as we place them. Built on the statistical models we have developed across the props markets since 2020.
- Stake sizing and the odds we took
- Every bet logged to the public sheet
- Delivered through Telegram
We come from data science and machine learning, and have modelled the props markets since 2020. Subscribe to get our models bets, or use our apps to price your own combinations and track your accounts. Every bet we send is logged to a public sheet, wins and losses alike.
Prefer to watch first? Join the free Telegram channel.
Every bet is written to a public spreadsheet, with the stake, the odds taken and the bookmaker. The figures below read directly from that sheet, drawdowns included.
Split by book so you can see the edge is not carried by a single soft market.
| Bookmaker | Profit (u) | Staked (u) | ROI | Bets | Bets/mo | Avg odds |
|---|---|---|---|---|---|---|
| Loading… | ||||||
A subscription for the bets themselves, and two apps for the bettors who want to build and manage positions on their own.
The bets our models find, sent as we place them. Built on the statistical models we have developed across the props markets since 2020.
Estimated odds from every model in one place. Combine selections to capitalise on correlations the books price independently.
One append-only ledger for every bookmaker account you hold, with multi-currency balances and a real view of your position.
Decide what you are willing to stake over a season and set one unit to 1% of it. Every bet we publish is sized in units, so the staking plan is already done.
Bets arrive with the selection, the stake in units and the odds we took. Place them at whichever bookmaker you hold, since nothing ties a bet to a specific book.
Results are logged publicly as they settle. Judge the service the way we do, over thousands of bets and several seasons.
Props markets are priced with less attention than main lines. Books post thousands of player and team totals every week and cannot model each one as carefully as a match result.
We build statistical models across those markets, estimate a fair price for each, and bet only where the offered odds are far enough above our estimate to survive the margin.
It is unglamorous. Average odds sit near 1.8 and average stake near 1.3 units. The edge is small per bet and only shows up in volume, which is exactly why the full record matters more than any individual result.
BK Digital is a small, dedicated team with professional backgrounds in data science and machine learning, and a genuine passion for football. We have been building and refining sports betting models since 2020, taking a fully data-driven approach to finding value in the props markets.
Everything we publish is built for serious, disciplined long-term bettors: people who stake consistently, keep to a fixed unit, and give a service thousands of bets before judging it.
That is also why the full record sits on this page. We would rather you scrutinise nearly four years of logged bets before paying us anything, so read the results and make up your own mind.
A sample of our bets, no card required
Each bet, as it is found
Get premium
access
Price your own combinations
Every account in one ledger
Bets are written to a public spreadsheet at the moment they are sent, before the result is known. The numbers on this page are read live from that sheet, and the free analytics app lets you filter the same data yourself by league, bookmaker, market and date.
You should still treat any tipster's self-reported record with scepticism. The best we can do is keep the log complete, so the losing bets are as easy to find as the winning ones.
Enough that one unit (1% of it) is a stake you are comfortable losing, repeatedly. The historical record includes a peak-to-trough drawdown of roughly 48 units, so a bankroll that cannot absorb a swing of that size is too small for this style of betting.
No. The record is tracked across Bet365, Kambi, NordicBet and Betano so you can see where the bets have been available, and the Betbuilder app is bookmaker-agnostic. You will get the most value from books that post props early and price them loosely.
Volume follows the fixture calendar. Across the logged record the average is roughly 40 bets on an active day, heavily concentrated around weekend league rounds, with nothing at all during breaks.
Probably, eventually. Bookmakers restrict consistently profitable accounts, and props betting at scale attracts attention faster than most markets. Spreading turnover across several books, as the breakdown above reflects, is part of how the record has stayed viable.
Subscriptions are handled by MyMembers, and you can cancel from your account there. Check the current terms on the checkout page for the notice period that applies to your plan.
Every bet since September 2022 is public, including the bad months. If the numbers hold up to your scrutiny, the bets are one click away.