Live Tennis API Ultra for a Personal Bot

Disclosure: Links to Live Tennis API, BF Bot Manager and BetBot are affiliate links. BotBlog may earn a commission if you buy, at no extra cost to you. Live Tennis API gave BotBlog Ultra access for testing. Code botblog is 10% off at their checkout. Full disclosure.

What this personal tennis bot actually does

I’m Stephane. I play tennis, I trade the tour, and I got tired of refreshing a page to see whether the server was at 15–40. So I wrote a small desk that stays on one PC and watches my own account. It reads two things at once: a Live Tennis API Ultra stream, and prices from the exchange. When a score I care about appears, a rule can open a paper order. It does not go live unless I turn that on myself.

If you only want the score next to a ladder, you do not need any of this. Open the free Tennis Trader Board. It checks about every 8 seconds, and the board guide shows the filters. The desk below is for the next step: your own rule, waking up when the point commits, not when a web page happens to refresh.

The top bar is the part worth reading. PAPER · MATCHBOOK means nothing has been sent to the exchange as a real bet. Ultra push is the key doing its job, with the lag shown in seconds next to the Matchbook clock. Under that, the cards are the afternoon’s paper book: a working bank of £994.20, the day down £5.80 after commission, 3 winners and 20 losers. I am showing it because it is the real screen, not because those numbers are an edge. One afternoon cannot tell you whether a tennis trading strategy is any good.

Why Ultra is the feed this bot needed

I did not pick Ultra because it sounds faster on a pricing page. I picked it because the bot I wrote has a hole where the score should be, and the other plans do not fill that hole in the way a point-level rule needs.

A break-point rule cares about one change on the board. The server goes to 0–40 or 15–40, the Match Odds market starts to move, and a bot that is still waiting for the next poll is already late. The free board is a timed check, about every 8 seconds, shared by everyone looking at the page. That is a good second monitor. It is a poor trigger. Ultra opens a WebSocket and pushes the point when it commits, along with who is serving. The desk can wake on that message instead of asking “what is the score?” on a timer.

The same plan includes a model win percentage. That is what lets a rule compare the model with the price on the exchange, which is the idea in model win % versus Betfair odds. You can see a “Model vs market” row in the rules screenshot further down. Basic and Pro will give you scores if you keep requesting them. They will not give you this push, and they will not give you that model on the stream. For a bot whose whole point is “do something when this point lands,” that is the difference between a feed and a page you have to keep reloading.

Two things Ultra deliberately does not do, and I am glad of both. It does not log into Matchbook or Betfair, so the score subscription and the betting login stay separate. And it does not place the bet. The order, if there is one, is my desk talking to the exchange. Costs, the Delayed key versus a Live App Key, and what you should not try to sell are on the Betfair tennis bots and API cost page. A personal project and a product you sell to other people are different licences. This is the personal project.

What one afternoon looks like

The open row is a paper back. You can see the score, the price the order was sent at, and what greening up would be worth if it were closed now. The four rows under it are matches the desk is sitting on, waiting for a score that matches a rule. “Waiting for a slot” is the overflow: either Ultra has not joined that match yet, or the desk is already watching as many as I allow. It is bookkeeping. It is not a selection, and it is not a tip.

If you want the same score on a phone, without any of this order book, the Telegram 0–40 / 15–40 alerts are a smaller project. If you want a browser scoreboard you built yourself before you think about orders, start with the live tennis scores API dashboard.

The rules are switches, not a system to copy

Each row is a door. Break-point lay, for example, is allowed to look when the server is 0–40 or 15–40. It can be On or Off. The stake on these rows is £2. The P/L beside them is paper, and most of the samples are too small to mean anything. The badge says Learning for that reason.

Please do not try to recreate this list and run it. I am not publishing the settings, the code, or a claim that any row wins. The project stays on my machine. The useful part of the picture is the shape: a live score comes in, a switch decides whether that score is allowed to open a paper order, and the result is written down so you can see it was a bad afternoon.

BF Bot Manager already has a tennis strategy builder

Most people who want a tennis bot should not write one. BF Bot Manager already has the builder, and it is the one the tennis guides on this site are written against.

You build the strategy in Bet Rules rather than in Python. A typical tennis entry is a score condition, an odds band, and a reason to stay out (a tie-break, a late match, a price that is already too short). After the bet, a second set of rules greens up, stops out, or closes when the score changes. The in-play tennis guide walks through loading Match Odds, and the Simulation mode guide is the part I would not skip. BFBM logs the hypothetical bets against live prices before you point the same rules at real money. There is a stop-loss guide for the exit side.

One honest difference: BF Bot Manager takes its tennis score from Betfair’s in-play data, which you can see updating in the Markets tab. It does not need a Live Tennis API key. I use Ultra because I wrote my own worker and I wanted the push and the model inside that worker. If you are happy for the software to bring the score with it, the Windows builder is the shorter path. Map a manual idea such as the 15–40 strategy into a score condition, run it in Simulation for a full day of matches, and only then think about real stakes.

BetBot’s builder goes down to the point

BetBot is the same job in a browser tab, which matters if you are on a Mac or you do not want a Windows install. The BetBot review covers the product. The part that belongs on this page is the strategy builder.

You can start from a template, answer a few guided questions, open the full builder, or let it draft a strategy that you then read line by line. For tennis, the conditions go down to the point on the board: 15–40, 30–30, break point, serving for the set, not only the set score or the game score. A market can reprice on a single point, so a builder that only notices “end of game” is late for the moment the break-point playbook is about. Those triggers use BetBot’s own live score. They do not use my Ultra key, and they do not use yours.

It starts on paper. That is the right default, and it is the same discipline as the desk above. BetBot is not my program, it does not run on your PC, and a clean builder is not evidence that a strategy makes money. It is evidence that you can test a point-level tennis idea without writing the feed yourself.

Where to start

• Watch first, no key: Tennis Trader Board and the board guide.
• The manual idea behind the break-point rows: 15–40 tennis trading strategy and the wider break-point playbook.
• A phone ping, still not a betting bot: Telegram 0–40 / 15–40 alerts.
• A private scoreboard before any orders: live tennis scores API dashboard.
• Windows rules, no code: BF Bot Manager tennis.
• A browser builder, paper first: BetBot review.
• Keys, licences, and what a personal bot is allowed to be: Betfair tennis bots, live scores and API cost and the Betfair API key guide.

Buy Ultra when you are about to put a score stream inside a program you run yourself. Until then the free board is the right tool, and a strategy builder you did not have to write is the right shortcut.