Football tier · Reference card · PM Academy

Football PMs.
One sheet.

Everything you need open while you trade. Vig math, xG ranges, public-bias formula, hold/sell tree, the five-point pre-flight. Pin it; print it; come back to it.

Back to Module 13

Vig math

M08

Vig is the bookmaker’s margin baked into the odds. Implied probabilities sum to 100% + vig, not 100%.

implied = 1 / decimal_odds
break-even win rate = implied at the price you took

Common vig rates:

Book	Vig	Min hit rate
Sharp / Pinnacle	2.0–2.5%	51.2%
Mainstream	4.0–5.0%	52.4%
Limitless PM	spread + fee	~50.5%

xG quick read

M09

Per-shot quality:

0.0–0.10	Speculative
0.10–0.25	Standard chance
0.25–0.50	Strong chance
> 0.50	Big chance

Match totals. Cumulative xG by half-time predicts second-half goals. xG >1.5 with 0 goals at HT → favourites mean-revert. xG <0.5 at HT → thin chances; expect a draw or 1-0.

Public bias

M10

bias = market_implied − fair_implied
positive = market overpriced, fade the favourite

Typical magnitudes on Premier League weekends:

Big-six home	+2 to +4%
Mid-table fixture	±1%
Underdog away	−1 to −2%
Storyline match	+5%+

Caveat. Bias evaporates when the public is right (late team news). Always check why a price moved before fading.

Kelly sizing

M03

f = (b·p − q) / b
p = your prob, q = 1−p, b = decimal_odds − 1

Worked: Arsenal at $0.55, you say 62%.

b = (1/0.55) − 1 = 0.818
p = 0.62, q = 0.38
f = (0.818·0.62 − 0.38) / 0.818
= 16.5% of bankroll

Rules of thumb. Cap at 5% bankroll. Use half-Kelly in your first year. Skip the trade if Kelly returns negative.

Hold-vs-sell decision tree

M11

Price moved. Did your fair-value model move with it?

Model unchanged, price up. → You’ve been carried by variance. Sell. Don’t anchor on entry.
Model up, price up by less. → Edge widened. Hold or add.
Model up, price up by more. → Edge gone. Sell at market.
Model down, price down. → Thesis is breaking. Sell.
Model down, price up. → Counter-trade emerging. Sell hard, consider taking the other side.
You can’t articulate why your model moved. → It didn’t. Sell.

Five-point pre-flight

M11

Two+ baselines. Sharp line, xG aggregator, similar fixture close.
Specific tactical edge. Not a feeling.
Liquidity verified. At your price + one level deep.
Kelly-sized. Capped at 5% bankroll, halved in year one.
Exits set. TP at X¢, SL at Y¢, before entry.

Order types

M01

	Limit	Market
Price	controlled	fills now
Fill	maybe	guaranteed
Slippage	none	up to spread
Use when	edge thin	thesis time-sensitive

Default. Limit at the displayed price. Switch to market only when the in-game move is structural and seconds matter.

Mistakes that blow accounts

Cross-module

Single-source edge. Your model alone isn’t an edge, it’s a hypothesis.
Sizing up after a loss. Kelly assumes independence; chasing breaks it.
Holding because you’re already in. Sunk cost. The relevant question is forward EV.
Ignoring liquidity gaps. Mid-table fixtures with sub-$5k books punish size.
Public bias one-way trade. Sentiment fades only when sentiment is wrong.
Skipping fees. A 7% theoretical edge is a 3% realised edge after spread + fee.