Every day, our AI scans hundreds of matches across Europe's top leagues and delivers a curated set of picks. If you are new to CalibrSports, the amount of information per pick might feel overwhelming at first. This guide breaks down every element so you can read our picks with confidence.
The Anatomy of a Pick
Each pick on our Hot Picks page or Telegram channel contains five key elements:
- Match: The two teams playing, e.g. Liverpool vs Arsenal.
- Market: What we are betting on. This could be 1X2 (match result), Over/Under 2.5 Goals, Both Teams to Score, Asian Handicap, Double Chance, Correct Score, or any of 12+ markets we cover.
- Pick: Our specific selection within that market, e.g. "Home", "Over 2.5", "BTTS Yes", or "Away +0.5".
- Odds: The best available price from bookmakers at the time of prediction.
- Stake %: How much of your bankroll we recommend risking on this bet. For example, 2.5% means £25 on a £1,000 bankroll.
Confidence Badges Explained
Every pick carries a colored confidence badge that tells you how it was generated and how much conviction is behind it. These are not just labels — they represent fundamentally different levels of analysis.
🟢 POWER PICK
Our highest conviction bets. A Power Pick means both the machine learning model and the AI Advisor independently agree this is a strong bet. The model found a statistical edge, and the AI Advisor confirmed it after reviewing news, injuries, form, and odds movements. When you see green, the system is speaking with one voice.
🟠 OPTIMIZED
The model identified value, but the AI Advisor made an adjustment. This could be a stake change (reducing from 3% to 1.5% because of a key injury) or a line change (suggesting Over 1.5 instead of Over 2.5 for a safer entry). Optimized bets are still backed by the model — the advisor just fine-tuned them.
🔵 SCOUT PICK
These are advisor-only bets that the model did not flag. The AI Advisor found a specific catalyst — breaking team news, significant odds movement, a form divergence the model has not priced in yet, or a calibration gap in our 30-day track record. Scout Picks always require a named reason. "The numbers look good" is never enough — there must be a concrete catalyst.
🟡 LONGSHOT
Low-conviction bets that the system has parked. These appear because we believe in full transparency — we would rather show you a marginal bet with a yellow warning than hide it. Longshots might have been vetoed by the advisor due to mass injuries, poor market track record, or match disruption risks. Proceed with extreme caution, or skip entirely.
Understanding Stake Percentages
The stake percentage is arguably the most important number on each pick. It is calculated using the Kelly Criterion, a mathematical formula that sizes bets proportionally to the detected edge. Here is what the ranges mean in practice:
- 3-5% stakes: Strong edge, high confidence. These are our best bets of the day. The model sees a significant gap between its probability estimate and the bookmaker's implied probability.
- 1-2% stakes: Moderate edge. Solid value but with more uncertainty. These are bread-and-butter bets that compound steadily over time.
- Below 1% stakes: Small edge. The value is there but marginal. These bets contribute to long-term profit but individually will not move the needle.
The beauty of Kelly sizing is that it self-regulates risk. When the model is unsure, stakes are tiny. When it sees a clear opportunity, stakes are larger. You never need to guess how much to bet — the math does it for you.
Reading the Probability Bar
Every match on CalibrSports shows an H / D / A probability bar — a horizontal bar split into three colored sections representing our model's estimated probability of Home win, Draw, and Away win.
The key insight: compare our probabilities to the implied probabilities from the bookmaker's odds. If our model says Home has a 65% chance but the odds imply only 50%, that 15 percentage point gap is where the value lives. The wider the gap, the stronger the edge.
For example:
- Our model: Home 65%, Draw 20%, Away 15%
- Bookmaker odds: Home 1.90 (implied 52.6%), Draw 3.50 (implied 28.6%), Away 4.50 (implied 22.2%)
- Edge on Home: 65% - 52.6% = +12.4% → this is a value bet
- Edge on Draw: 20% - 28.6% = -8.6% → no value, skip
Multiple Picks Per Match
You will often see more than one pick for the same match. This is by design. A single fixture might offer value on the 1X2 result, the Over/Under line, and an Asian Handicap simultaneously. Each market is analyzed independently, so a Home win pick and an Over 2.5 pick on the same match are not contradictory — they target different aspects of the game.
What to Expect Long-Term
Not every pick wins. No system in the world wins every bet, and anyone who claims otherwise is lying. Our target is approximately 60% win rate across hundreds of bets, with Kelly-sized stakes that compound the edge over time.
Some key expectations:
- Losing days happen. Even a 3-bet day where all three lose is normal variance, not a system failure.
- Losing weeks happen. Football has natural variance. A 5-day cold streak followed by a 10-day hot streak is typical.
- The edge is real but small per bet. Profitability comes from consistent, disciplined execution over hundreds of bets, not from any single pick.
Trust the process, manage your bankroll, and never stake more than you can afford to lose. To see how this plays out in practice, check our public track record where every bet and result is published transparently.