Every figure on the site, in one place

The ten samples are defined across the site; this page reduces each of them to the one calculation it rests on, so a reader can re-derive every number rather than take it. Nothing here is observed from a real operator, player or match.

Desk spec

samples: 10
figures: all
method: re-derivable
origin: invented

the projectionWhat a player is expected to do: an expected-minutes figure multiplied by a per-90 rate. The sample player is expected to play 78.0 minutes at 1.40 shots on target per 90, so the model expects 1.21 of them rather than the 1.40 a full match would imply.

the participation definitionWhat the operator's rules mean by "to play". Across 400 sampled single-player markets the same bets stand 308 times under "named in the starting XI" and 369 times under "takes the field", so the definition decides whether the bet is settled at all.

the confirmationThe source that counts the stat after the match and the audit behind it. 14 of 1,200 sampled markets were corrected after first publication, 5 of them more than an hour later, and 3 after the payout had already been made.

Direct answer

Every figure on this site derives from ten invented samples: a projection (78.0 minutes at 1.40 per 90, output 1.21, fair 2.93), four participation definitions (308 to 369 of 400 stand), a confirmation rate (14 of 1,200 corrected), conditional scoring rates (21.25% overall), two ceilings (800.00 and 12,000.00) and a month of 8,000 bets retaining 7.92% of settled stakes.

Sample A and sample J - the projection and its band

sample A - the core calculation expected output = 1.40 x 78.0 / 90 = 1.21 P(2 or more) with a mean of 1.21 = 1 - e^-1.21 x (1 + 1.21) = 34.10% fair price = 1 / 0.3410 = 2.93; offered = 2.70 sample J: 66.0 minutes -> 3.63; 90.0 minutes -> 2.45 band = 3.63 - 2.45 = 1.18 = 40.3% of 2.93.

Sample B - the definitions, and the void

sample B - stand and void across 400 markets named in the starting XI: 308 stand, 92 void (23.0%) takes the field: 369 stand, 31 void (7.75%) plays 60 minutes: 331 stand, 69 void (17.25%) completes the match: 297 stand, 103 void (25.75%) scaled to 8,000 bets: 620 to 2,060 voids, a swing of 1,440 bets.

Sample C, D, H and I - confirmation, correlation, lag and tie

samples C, D, H and I sample C: 14 of 1,200 corrected = 1.17%; 3 after payout = 0.25%; 6 a woodwork shot sample D: 0.30 x 0.60 + 0.10 x 0.25 + 0.05 x 0.15 = 21.25% -> fair 4.71 sample H: median settlement lag 22 minutes; 68.0% inside 30; p90 3 h 05 m sample I: 6 of 1,200 counting disputes = 0.5%; 9 of 36 pages name a provider

Sample E and sample F - the ceilings and the record

sample E and sample F sample E: player maximum 800.00; team maximum 5,000.00 = 16.0% 800.00 at 2.70 = 2,160.00; at 3.63 = 2,904.00 payout ceiling 12,000.00 caps a 2.70 price at a stake of 4,444.44 sample F: 36 rule pages -> 22 settle, 14 define, 9 name, 12 void, 6 limit, 8 tie, 1 all six the one page stating all six = 2.8%.

Sample G - one month in money

sample G - the month 8,000 bets at a mean stake of 9.40 = 75,200.00 turnover settled 7,380 (92.25%); void 620 (7.75%) = 5,828.00 returned cleared 7,380 x 0.3410 = 2,517; paid 2,517 x 9.40 x 2.70 = 63,881.46 retained 69,372.00 - 63,881.46 = 5,490.54 = 7.92% of settled stakes a year = 5,490.54 x 12 = 65,886.48.

Every figure above is illustrative and derives from the ten invented samples defined across this site. No real operator, player, team or match is described, and none of the numbers is observed from a live market. The calculations are shown so that any reader can re-derive them, which is the whole purpose of putting them in one place.