Predictions

Ongoing probability research with virtual stakes and bankrolls. Compare model and bookmaker probabilities.

Public beta · Virtual simulation

We benchmark model probabilities against bookmakers and tune for simulated returns with virtual stakes and bankrolls. No real money is wagered or won. Simulated profits do not promise real returns.

Independence, method & limitations

An ongoing public-beta experiment comparing model probabilities with bookmaker-implied probabilities for football statistical outcomes. We tune and evaluate our models to improve simulated returns using virtual stakes and bankrolls. Recorded profit and loss are hypothetical; no real money is wagered or won here.

StatsGem is independent. We are not a bookmaker, gambling operator, betting broker or bookmaker affiliate. We accept no bets or gambling deposits, hold no player funds and offer no cash prizes or withdrawals. Subscriptions pay for research access. Bookmaker names and logos identify data sources; they do not imply a partnership or endorsement.

Simulated performance does not predict real-world profit. Models can be wrong, odds can change, and historical results can overstate an apparent advantage. Gambling can cause financial harm and addiction. Get support

PredictionsPredictions Player propsProps FantasyFantasy BenchmarksBenchmarks

2 Free0 VIP0 Free Active0 VIP Active

2 picks shown with Paddy Power prices. Recorded performance covers all suggestions for the selected tier and markets.

Board

Matches

2 matches

10:00LaziovMilanSerie A1 historical estimate

Corners · Lazio total

Open match →Historical estimate

under 3.5

Paddy Powerreference price

Reference odds2.20

Closing odds2.20

Reconstructed from saved model and bookmaker data. This was not a saved pre-match recommendation and is excluded from recorded performance.

Bookmaker prices and pick details

Bookmaker price options · Choose one. Different lines have different model chances.

Paddy Powerreference priceClosing estimateunder 3.5
Reference odds2.20
Closing odds2.20
Model chance55.6%EV+22.4%
BetMGM UKClosing estimateunder 3.5
Reference odds2.15
Closing odds2.15
Model chance55.6%EV+19.6%
Bet365Closing estimateunder 3.5
Reference odds2.10
Closing odds2.10
Model chance55.6%EV+16.8%

17:30DC UnitedvAtlanta UnitedMLS1 historical estimate

Corners · DC United total

Open match →Historical estimate

under 5.5

Bet365reference price

Reference odds2.10

Closing odds2.10

Reconstructed from saved model and bookmaker data. This was not a saved pre-match recommendation and is excluded from recorded performance.

Bookmaker prices and pick details

Bookmaker price options · Choose one. Different lines have different model chances.

Bet365reference priceClosing estimateunder 5.5
Reference odds2.10
Closing odds2.10
Model chance57.2%EV+20.1%
Paddy PowerClosing estimateunder 5.5
Reference odds2.10
Closing odds2.10
Model chance57.2%EV+20.1%

Selected day

Sat, 12 Sept 2026

Recorded results · all bookmaker options

Historical estimates for this date are shown below. They are excluded from recorded results.

2 historical estimates shown below. Their calculated returns appear only in the calendar and all-date display totals.

All dates

Overall results

5,130 settled picks

One unit per pick. Includes historical estimates at reference odds (1,974 settled). Display only; recorded performance and benchmarks exclude estimates. Recorded results: 28 Sept 2026 – 18 Oct 2026.

ROI+2.7%

Profit+138.37 u

W / L2,521 / 2,609

Avg odds2.12

How to read this board

A simulated selection is one match, statistic, scope and direction. Bookmaker prices are alternatives for comparison, not additional stakes.

Results use one virtual unit per recorded selection at its recorded line and odds. The comparison bookmaker is labelled on each selection. These are simulated returns; no real money is wagered or won.

Earlier matches without a saved pre-match recommendation are labelled Historical estimate. They remain visible in the calendar and are excluded from daily and overall results.

Free and VIP use the same process. Their labels reflect the benchmark rules; one day's return cannot determine which will perform better.

Telegram reads these same recommendations and applies your tier, bookmaker preferences and delivery time. A message does not create an additional pick.