Bet record

All leagues Premier League LaLiga Bundesliga

Every recorded line, what the model said about it, and how it settled. The point is to build a track record before staking anything on one — a model's own edge estimate is a claim about itself, and this is the only thing that can check it. Predictions come from the fitted model where a league has enough history for one, at σ = 8.02. Rows marked ^* came from the naive midpoint instead, because that league cannot fit a model yet — see below the table.

Lines recorded

Settled

Backed by the model
positive edge only

-1.92

Profit / loss (units)

-48.0%

Return on stake
on 4 bets — not a result

Should you be betting this yet?

Not yet.

only 4 settled bets. Below about 50 the result is noise whichever way it falls.

Bets the model actually backed, settled	4
Won	1 (25%)
Breakeven at these prices The average of 1/odds across the bets taken — the rate that merely breaks even	51%
Standard errors clear of breakeven Below 2 the result is within what luck produces	-1.0

The rule: bet when the model has beaten the market by more than luck explains — not when it claims an edge. A model's own edge estimate is a claim about itself; the only thing that can check it is a settled record. Two standard errors is the usual bar, and at these prices that means roughly 100–200 settled bets. That is not pessimism, it is what it costs to tell a real 5% edge from a coin that landed well.

You do not have to stake money to get there. Every recorded line is graded whether or not a bet was placed, so the record builds itself for free.

5 settled — far too few to conclude anything. Detecting a real 5% edge with any confidence takes something closer to 100–200 settled bets. Until then this page is a record, not a verdict, and a run of wins is as likely to be luck as skill.

Kickoff	Match	Line is about	Line	Bet	Odds	Model	P(side)	Edge	Stake	Actual	Result
13 Sep	RSO v ATM	Atlético de Madrid	52.6	no bet	-	52.6^*	50%	-4.3%	0/10	-	skipped
13 Sep	LEV v BAR	FC Barcelona	70.6	UNDER	1.84	60.1^*	89%	+35.0% ★	1/10	74.9	lost
12 Sep	RMA v RAY	Real Madrid	57.1	OVER	2.16	57.1^*	50%	+3.6% ★	1/10	55.9	lost
12 Sep	ATH v ELC	Elche CF	54.6	UNDER	2.08	54.8^*	49%	+1.1% ★	1/10	33.3	won
12 Sep	OSA v ESP	CA Osasuna	51.6	UNDER	1.84	50.2^*	56%	+2.1% ★	1/10	60.6	lost
11 Sep	SEV v VAL	Sevilla FC	52.1	no bet	-	51.3^*	54%	-0.6%	0/10	45.7	skipped

^* 6 rows did not come from a fitted model. Their league has too little history to fit one, so the figure is the naive midpoint — the side's own recent possession averaged against what its opponent concedes, on as few as two matches each. The midpoint pulls hard towards 50, so it understates a mismatch: a strong side away at a weak one reads lower than it should, and the edge against a bookmaker's line can come out the wrong way round entirely. Treat the Bet, Edge and Stake columns on these rows as unproven.

Adding lines

Edit odds.txt in GitHub's web editor, paste a row, commit. The hourly run imports it. Columns are date · club · club · whose-possession · line · over · under, whitespace-separated, with - for a price you do not have. Club order does not matter. Use the abbreviations below, or names without spaces.

ARS Arsenal AVL Aston Villa BOU Bournemouth BRE Brentford BHA Brighton CHE Chelsea COV Coventry CRY Crystal Palace EVE Everton FUL Fulham HUL Hull IPS Ipswich LEE Leeds LIV Liverpool MCI Man City MUN Man Utd NEW Newcastle NFO Forest SUN Sunderland TOT Spurs