Confidence range
25 to 100.
WB3 (76 to 95) is the same band the curated HOTSHEET ships from. On Terminal it is one segment of the full board, not the whole product.
// HOW THE TERMINAL WORKS
The full method. Nothing held back.
The Terminal scores every pick on the board from 25 to 100 and grades the result in public. This is how that number is built, how to read a pick, how the market read and risk pricing work, and how the Terminal compares to the alternatives.
// HOW THE MODEL WORKS
One score. One price. One read.
Every signal on the daily board carries a single confidence score from 25 to 100. That score is the model's read on the line and the Points cost to unlock it. Below is the scale, a sample card, and the four principles the whole Terminal runs on.
// THE CONFIDENCE SCORE
The full board. Priced 25 to 100.
Mainlines, props, totals, F5, full game, across every active sport. Nothing below 25 publishes. At 96 and above, the pick ships as a GEM.
Tier breakdown
- 25–49WB1
Lower-conviction signals. Lower unlock cost. Useful for board awareness or speculative low-bracket positions.
- 50–75WB2
Solid model conviction. The bulk of the daily board. Most subscribers allocate the majority of their Points here.
- 76–95WB3
The band the curated HOTSHEET ships from. On Terminal, every signal in WB3 is unlockable, not just the top three.
- 96–100GEMS
Premium signal band. The strongest model conviction on the board. Allocation per Trader plan.
// ANATOMY OF A PICK
Score is price. Conviction is sizing.
Every game card surfaces the picks the model is willing to score. A 96 GEM costs 96 Points; a 76 standard costs 76. Wager size scales with the bracket the trader picks at unlock time.
MLBTODAY · 7:05 PM ET
NYY@BOS
Yankees @ Red Sox
MONEYLINE
NYY −115
CONFIDENCE96
UNLOCK COST96 POINTS
RUN LINE
NYY −1.5
CONFIDENCE81
UNLOCK COST81 POINTS
TEAM TOTAL
NYY OVER 4.5
CONFIDENCE74
UNLOCK COST74 POINTS
MONEYLINE
NYY −110
CONFIDENCE61
UNLOCK COST61 POINTS
TOTAL
OVER 4.5
CONFIDENCE58
UNLOCK COST58 POINTS
This is one example. The model may surface fewer or different picks for any given game depending on confidence and market signal.
// CARD ANATOMY
Read the number, send the wager.
The Terminal card is the operator surface. Every score is the model's read on the line, and the unlock cost equals that score in Points. No upsell, no decision fatigue.
- Full game and F5 (first five) on a single card. Segment markets price independently of the mainline.
- 25 to 100 score from the model. GEMs at 96 and above read in green. Below 25 does not publish.
- Unlock cost equals confidence rating. The score on the card is the price on the wallet.
- Wager size scales with confidence using bracket multipliers, not flat sizing. The trader picks the bracket; the model picks the score.
// THE FRAMEWORK
Define risk. Measure edge. Execute. Detach.
Four operating principles. The model generates the read; the trader holds the discipline. Position sizing and detachment do the work the model cannot.
- 01
Define risk in brackets.
Position size scales with conviction. WagerBird uses bracket multipliers (1.75, 2.0, 2.25, and up) keyed to confidence band. Brackets preserve the difference between a 71 percent pick and a 66 percent pick that flat sizing would round away. The bracket is the discipline.
- 02
Measure edge.
Every signal carries a confidence score from 25 to 100. The score is the model's read on the line. Unlock cost equals the score in Points. The market sets the price; the model sets the conviction.
- 03
Execute.
Unlock the pick, place the wager at the confidence-mapped bracket, log the result. Discipline replaces feel. No chasing, no doubling, no off-card improvisation.
- 04
Detach.
The model does not care about the last pick. Neither does the trader. Outcomes feed back into the engine. Emotion does not.
// THE MARKET READ
How the model reads the market.
Three reads layered into every confidence score. None of these stand alone; together they price what the mainstream line misses.
- A
MARKET STRUCTURE
Where inefficiencies live.
Sportsbooks price mainstream lines tight and props loose. Liquidity, line shopping, and book bias create a structural gap on derivative markets. The Terminal surfaces signals where the price lags the model the most, not where the volume is loudest.
- B
BOOKMAKER PSYCHOLOGY
Trap vs value.
Books shade lines toward bettor psychology, not expected value. A line that looks easy can be the trap. A line that reads against the public can carry conviction. The Terminal evaluates each line situationally. The model weighs book posture, market depth, and prior cycles before scoring; the customer sees the score, not a formula to extract.
- C
SITUATIONAL CONTEXT
What the model layers on top.
Weather on totals. Day-game splits after night travel. Lineup deltas, bullpen state, rest advantage, referee tendency. The variables that move a fair line by a tick before any market data fires. Quantified, weighted, and re-scored each cycle.
// RISK-PRICING PHILOSOPHY
Brackets, not units.
Most picks services flatten everything into units. WagerBird prices conviction as a multiplier and lets the bracket carry the difference between a strong read and a marginal one.
- α
THE UNITS PROBLEM
Units flatten conviction.
A flat-unit model treats a 71 percent pick the same as a 66 percent pick at the same unit count. The bettor loses the ability to scale risk to confidence. The book takes the gap.
- β
THE BRACKETS ANSWER
Brackets scale with confidence.
WagerBird uses bracket multipliers (1.75, 2.0, 2.25, and up) keyed to the confidence band. A 96 GEM gets a different bracket than a 76 standard. The size matches the read.
- γ
WHY IT COMPOUNDS
Conviction priced, not averaged.
Over a long sample window, bracket-based sizing materially outperforms flat sizing. Edge does not arrive in flat doses; it arrives weighted. The framework prices weight directly.
// POINTS AND GEMS
The wallet. Points, Points-Back, and GEMs.
// HOW TERMINAL COMPARES
Different operating model. Different product.
Most of the picks economy is built for entertainment. Terminal is built for allocation. Four short reads on how Terminal differs from the alternatives a serious customer is already evaluating.
vs Picks services
Numerical confidence, not a hot take.
Picks services sell the personality of one handicapper. Terminal publishes a confidence score generated by a model and graded daily against the close.
See the comparison →vs AI / algo prediction sites
A working framework, not a black box.
Most algo sites show a percentage and a one-click bet. Terminal explains the framework, prices conviction in Points, and refuses to publish below the model floor.
See the comparison →vs Sportsbook in-app tools
Book tools work for the book.
Sportsbook same-game-parlay generators and boost feeds optimize for the house margin. Terminal exists to find the gaps the book did not close.
See the comparison →vs Discord groups
A research surface, not a chat room.
Discord groups operate on social proof. Terminal operates on graded outcomes and a logged ledger at /results. Different product, different operating philosophy.
See the comparison →
Full comparison lives at /why-wagerbird. Updated as markets shift.
// THE FOUNDATIONS
Built on the work, not the noise.
The Terminal model reads sports markets the way capital markets are read. Public research on book psychology and forecast confidence informs how every signal is scored.
WagerBird methodology
Confidence scoring, not handicapping.
WagerBird is an analytics product. Every Terminal signal is priced in Points by the model. The customer chooses when to act and at what size. WagerBird does not accept wagers and does not place wagers on a customer's behalf.
For the free trial product, see /hotsheet.
Academic foundations
Public research, working framework.
- Sportsbooks shade lines toward bettor psychology, not expected value. Edge sits in the gap.Levitt, 2004 · Why are gambling markets organised so differently from financial markets?
- Confidence calibration matters more than directional accuracy. The right score on the right pick beats the next pick.Feddersen, 2013 · Confidence as a signal in sports prediction markets.
- Bracket-based position sizing materially outperforms flat sizing across long sample windows.Goto and Yamada, 2023 · Position sizing in consumer-facing prediction markets.
These citations frame the operating philosophy. They are not the model. The model is proprietary and runs daily against the live board.
24 / 7 customer service desk
A real desk on the other side.
Subscribers reach a human at any hour. Account questions, billing, magic-link recovery, plan changes. WagerBird does not run a chatbot maze; tickets route to a person.
Account access, billing, magic-link recovery, plan changes. Tickets route to a person, not a queue.