Events to balance is a list of matches where the book leans to one side. One row shows the heavy outcome, what the house holds on the other two, and how much stake would flatten the worst case. What happens next depends on the phase of that skin.
A trading morning often starts with a fixture list. Events to balance starts with the hole.
Nash AI reads the open markets and puts the unbalanced ones on a single list. Each row is one match on one skin. The same fixture on another skin is another row, because it is another book. Filters cover all matches, matches in play, and the next 24 hours. A search box finds a fixture by name.
The badge on the sheet says the view is information only. Nash calculates the row and shows it. It does not message a customer from this screen.
How to read the row
Open the match. The market in view is the one that is most unbalanced, often the match result. Three lines follow, one per outcome.
Each line shows the current odds, the chance Nash is using, what the house holds if that outcome wins, and whether more stake is needed on that outcome. The chances are normalized, so the three of them add up to 100. The overround sits in the odds.
Under the lines, a short instruction names the stake to bring onto the lighter outcomes, at today’s odds. Next to it, the row can show an average: each outcome multiplied by its chance, then added up. That average is a picture of this quote. It moves when the quote moves. It is not a forecast, and Nash does not promise a result.
When a price is missing, the average cell can stay empty. The worst outcome can still be on the list. An empty average is a reason to look, and a reason to leave the row alone until the price is there.
A proposed limit on new bets can sit on the heavy side. It stays marked as proposed until someone confirms it. Bets already on the match stay on the match.
A sketch, so the columns are easy to read
This table is a drawing. It is not a book Sirplay ran, and it is not a target.
| Outcome | Odds | Chance | House now | Stake to bring in |
|---|---|---|---|---|
| Home | 2.20 | 40% | −100 | none |
| Draw | 3.40 | 26% | +180 | 45 |
| Away | 2.60 | 34% | +150 | 55 |
Home is the heavy side. Forty-five on the draw and fifty-five on the away add up to the hundred the house would lose if home wins. After that stake lands, home is flat. Draw and away are still ahead on this sketch, by less than they were. The row has spent part of the other two outcomes to close the hole in the first.
On a live match, do the sum again before you act. During a reading of this sheet the odds moved, and the stake to bring in moved with them. The instruction from two minutes earlier was already a different book.
What you do with the row
Read the worst outcome first. Then read how much stake on the other sides would flatten it, and what the average looks like after that. If the average is healthy and the hole is large, the row is asking for a decision. If the average cell is empty, or the match has just re-priced, the decision can wait.
The people already betting the heavy side are counted on the match. The customers Nash would offer the other side to are a second list. Opening a name shows the offer in the shape that customer already bets. It does not send the offer.
Where this row sits
Events to balance is one screen on the Nash desk, inside the sports betting software an operator’s team runs each day. It is not the screen that accepts or refuses a single bet. On that other work, Nash applies the limits the team has set. A low-risk bet can be accepted. A medium-risk bet can be offered a smaller stake instead of a refusal. A high-risk bet goes to a trader, with the reason. A bet stops only when the result is already decided, for example a goal just scored. Those are decisions on a ticket. This row is a decision on a market.
The choice of sportsbook platform still has to name who compiles the price and who may suspend a market. Events to balance assumes that person exists. The sheet can line the hole up. It does not appoint anyone.
Filters on the sheet cover every match, matches in play, and the next 24 hours. A search box finds a fixture by name. Sort the list by the size of the hole when the question is where the risk sits, and by start time when the question is what kicks off next. A small league match can sit above a famous one. The famous one can be the calmer book.
Read it twice when the match is live
A live price moves. The stake the row asks you to bring in is true for the quote on the screen at that moment. Leave the row open and the same match can ask for a different stake a few minutes later. The worst outcome can stay where it was while the other two prices move. Do the sum again before anyone confirms a limit, and before anyone goes looking for customers to take the other side.
An empty average belongs to the same habit. The list can show a hole before it can price the average across the three outcomes. That match belongs on the desk. It does not belong in an instruction until a price is there. A trader who treats the empty cell as zero has invented a number the sheet refused to print.
The sketch above uses round figures so the columns are easy to follow. On a real row the odds will not be that neat, and the chances will be the normalized ones Nash is using, not a guess from the price alone. Write the three house figures down, add the stake you would bring onto the lighter outcomes, and read the new house figure on every outcome before you leave the chair. If one of those new figures is a loss you did not expect, the instruction was for a different quote.
The phase of the skin
Each skin has its own phase. Managers set it in Nash AI, under Settings. The same match can be a reading on one skin and a decision on another.
- Off. Nash ignores the skin.
- Shadow. Nash evaluates and shows the box, and decides nothing. A skin not yet configured stays here.
- Pilot. Nash decides only the reviews no trader decides in time.
- Production. Nash also decides at verification, according to that skin’s switches. The pilot stays active. On those bets Nash decides, and the skin can always be moved back.
A new skin usually goes shadow, then pilot, then production.
This tab is a different switch, under Balancing. Observe shows the matches, the badge says information only, and nobody is contacted. Propose publishes the plan for marketing, which chooses whom to contact. Nash still contacts nobody here and changes no bet. Production decides at verification. It does not send this list.
Machiavelli would put the test in one line: who can still leave the skin in shadow. The managers can, and they can move it back from production.
Nash does not close an account, and it does not promise results. A manual override remains. The licence stays with the operator. Sirplay sells the software, as the about Sirplay page records, and does not take the operator’s place. A supplier’s own testing does not authorise the book. The sportsbook buyer’s guide is the wider list of what to settle before a desk is live: risk, live betting, and go-live. This article is only the row. A white label sportsbook is a way to deliver that desk. It is not a licence. The phase of each skin says who decides at verification.
On a demo, ask to open Events to balance on a sample book. Read one live row, watch whether the stake to bring in changes when the price changes, and check that a limit on the heavy side stays marked as a proposal until someone confirms it. Request a demo. A demo can be met within 24 hours. Go-live is about ten working days, and that count leaves the licence outside it.
Not from this sheet. Observe shows the row and contacts nobody. Propose hands the plan to marketing, and Nash changes no bet. In production Nash also decides at verification, according to that skin's switches.
No. It is today's prices, weighted by the chances on the row. A new price makes a new figure.
The worst outcome can still be listed. An empty average is a reason to look, and a reason to leave the row until a price is there.
No. Nash does not close an account. A person on the team remains able to suspend the market, and a confirmed action belongs in the audit log.