User Comments

When you activate "Use web page login" in Application/Settings, it means that when you log in to Betfair via the bfexplorer app, a dialog will appear displaying the official Betfair web page login. This allows you to complete the second step of your two-factor authentication (2FA). To be frank, I have not tested 2FA yet, but the standard web page login works. My only hope is that after a successful login, the 2FA process redirects back to the same webpage, as bfexplorer needs this location to read the session token and authorize any of your Betfair API operations.

Set up bfexplorer agentic support in your AI chat tool, and then simply ask bfexplorer your questions.

This phenomenon is called internet connection latency. While internal operations within bfexplorer are measured in nanoseconds and microseconds, most of the time, sending any request to the Betfair servers is dictated by this connection latency. Therefore, if your strategy relies on low-latency execution, you should run it from a data center located close to the Betfair infrastructure.

Yes, it is possible to code using the Betfair Streaming API. When a new market is added to the Betfair exchange, the streaming API subscription will automatically receive that market or markets. The only limitation is a 200-market limit per single subscription from Betfair. You can find relevant code examples on my GitHub.

The bfexplorer app is an agentic application that provides AI agents with access to extensive data from both bfexplorer and betfair; therefore, you can simply prompt it with 'For favorite event "Horse Racing - GB and IE", filter markets with the racecourse: Catterick and open only those markets in bfexplorer', and bfexplorer will automatically open all markets for the Catterick racecourse.

You previously trialed bfexplorer for free, and your Betfair API access expired on July 24, 2026.

If you run bfexplorer with the -mcp localhost:10043 arguments, the MCP server runs within the bfexplorer application. This allows bfexplorer to function as an AI agentic tool for you.

By adding the MCP settings to your AI Agent, you can prompt bfexplorer and execute AI-driven strategies using available data context.

I haven't updated the bfexplorer app for some time because I have had no interest in it and faced issues releasing the Bfexplorer Studio tool.

Therefore, until I find a solution, I have no plans to update the current version.

The Execute Strategy Settings With Parameters tool is available; its purpose is to allow AI agents to execute preset strategies using parameters that can be changed by the agent.

The REST API exposed by bfexplorer was originally an experiment used at the beginning of my AI-driven strategy development, primarily because most AI chat clients did not support MCP at the time. It was simply a way for me to test some of the MCP tools I was building. Its main purpose is to expose data context to the user.

Furthermore, you can change any parameters of the preset bots you configure, as well as write your own bot strategies to execute preset strategies and dynamically change any parameters.

http://bfexplorer.net/Content/Bfexplorer/BfexplorerAI.png

I do not know if you are aware that Bfexplorer is the first app on the Betfair App Directory to support AI agents. Here is the AI's answer to your question, which you could have found yourself:

Nice work — your setup is logical and follows the common “Lay the Draw at half / hedge if a goal” pattern. Below I’ll walk through what you’ve done, point out where it’s correct, suggest improvements (hedges / risk controls), show options to auto-select matches, and explain how to keep bots running continuously.

1) Quick review of your current bot choices (correct / sensible)
- Place Bet bot ("Lay draw"): correct for placing a lay in-play (AllowPlacingBetInPlay = True, OfferMyBet = False, BetType = Lay). Good.
- Close Selection Bet Position at Odds ("Back and cash out"): sensible choice to hedge/close the lay position automatically when conditions met (HedgingEnabled = yes). Using CheckingLastPriceTraded is a common way to trigger on price movement.
- Football bot to start the hedge when score/time changes ("Back draw when 0–0 breaks"): correct trigger logic (Score or MatchTime).
- Sequence Execution to run the lay then waiting/trigger bots: correct approach to keep a single strategy that performs the sequence.
- Execute Bots (entry price filter) to only run when draw odds < 2.9: good pre-entry price filter.

So overall your choice of bot types and the sequencing is appropriate and follows typical practice.

2) Improvements and suggestions (risk control, robustness)
- Add an explicit liability / stake cap: make sure your Place Bet bot has a known maximum liability exposed if possible. If the Place Bet bot doesn't have a direct max-liability field, calculate the lay stake by formula so you never exceed desired liability.
- Add a late-time stop earlier or scale: 75' is reasonable, but consider:
- A staggered stop (e.g., stop new entries at 60' and close remaining exposure at 70') or
- Lower stakes for matches that remain 0–0 after 60' (because late goals become more likely).
- Add a price-based stop-loss/exit: if the draw price drifts up significantly (e.g., > X ticks above entry) you may want to close or partially close to limit loss. Implement with a "Close Market Bet Position" or "Trailing Stop Loss" strategy:
- Example: if draw LTP > 4.0 (or X ticks worse than your entry) then close position (partial/full).
- Or use a trailing stop so that if the odds retrace by Y ticks you lock-in partial loss control.
- Partial hedging / scaling out instead of single all-or-nothing:
- Hedge part of the stake when small favorable price change occurs (lock small profit) and keep remainder running to target larger hedge — reduces variance.
- Handling unmatched bets / latency:
- Consider adding a check to cancel/replace unmatched bets after N seconds or only place bets if last traded price within small window of offer price. This avoids getting matched at a very bad price.
- Use "CheckingLastPriceTradedDifference" carefully:
- A difference of 3 ticks is common, but check how your odds ticks map and do some backtesting in Practice Mode to avoid premature or late hedges.
- Reaction to in-game events (red cards, injuries):
- If you want more safety, add criteria to stop/close if Cards or exceptional events are available in your data feed — but note not all matches provide reliable card/event feeds to the bot.

3) Hedging formulas and practical hedge setup
- Full green-up hedge formula (to calculate back stake for zero profit):
backStake = (layStake * (layOdds - 1)) / (backOdds - 1)
- Use the Close Selection Bet Position bot or a “Green Up / Hedging” bot that calculates and places the required back stake automatically.
- If you prefer target profit: calculate back stake to leave a desired profit instead of zero.

4) Automation: selecting a series of matches in advance
Yes — you can automate market/selection choice using these approaches:
- Favourite events and saved filters:
- Use Bfexplorer’s favourite events or event groups (your “My Favourite Bet Events”) to group competitions you want to trade.
- Market scanners / filters:
- Create a scanner or market filter (competition, kickoff time window, pre-match draw odds < X, in-play availability) and bulk-add results to monitored markets.
- Execute a strategy over multiple selections:
- Use “Execute strategy on multiple selections” (ExecuteBfexplorerStrategySettingsOnSelections) or an equivalent app action to run your sequence on many markets/selections at once.
- Scheduling:
- You can pre-select markets (by market ID or by adding events to favorites) and have a single ExecuteBot run across that list at kick-off time.
If you want, I can help design the exact filter rules you need (competitions, kickoff window, pre-match draw range, bookmaker implied probability filters, etc.), and show how to turn them into a list that the Sequence Execution can run across.

5) Running bots when your computer is off
- Bfexplorer must be running and connected to Betfair for bots to execute. If your PC is off, bots cannot run.
- Two practical options to keep bots running 24/7:
1. Run Bfexplorer on a VPS (Windows VPS) or always-on remote PC — this is the recommended, common solution. Use a reputable VPS provider (low latency to Betfair endpoints), keep Betfair session authenticated, and ensure Bfexplorer auto-starts on login.
2. Leave a dedicated local machine on (e.g., an old PC or mini-PC) with unattended operation and remote access.
- Running minimized/in-background: Bfexplorer can run minimized; the app doesn't require the UI to be visible. But it still requires the machine and network to be on.
- Consider automation precautions on VPS:
- Secure the VPS (login credentials, 2FA),
- Keep Bfexplorer and Windows updates in mind,
- Monitor logs and have remote alerts for failures.
- There is no way for Bfexplorer to continue to run if the machine and service are completely powered down (no cloud execution without a VPS).

6) Testing, monitoring, and risk management
- Always test in Practice Mode first (you already are — good).
- Start with very small stakes when moving to real money.
- Keep a monitoring dashboard or alerts (email/SMS) to notify if something unexpected happens (bot error, disconnection, large loss).
- Keep a historical record (GetAllBetResults) and review performance periodically, tweak parameters and re-test.

7) Potential alternative/advanced bots to consider
- Trailing Stop Loss bot: to lock profit / limit loss dynamically if the odds move after you’ve placed a lay.
- Close Market Bet Position: more flexible ways to set profit or loss thresholds.
- Sequence Execution across selections: to run the same full sequence across multiple matches automatically.
- AI/ML or selection-filter bots: if you want to pre-filter markets based on historical features or machine learning ratings (advanced).

8) Next steps I can help with
- Review your exact strategy settings and suggest concrete parameter values (e.g., ticks for trailing stop, price thresholds).
- Build a market-filter for auto-selection (competition, kickoff window, pre-match draw odds, etc.).
- Create sample hedging calculations and show how to configure the Close Position bot for partial/targeted hedges.
- Recommend VPS specs and setup checklist to keep your bot running 24/7.

If you like, share:
- an example match/market ID you trade (or the competitions/time window you target),
- your exact lay stake/liability rules,
- whether you prefer full green-up or target-profit hedges,

and I’ll produce concrete parameter suggestions and an improved sequence (including sample stake/hedge calculations) you can test in Practice Mode.

You must have switched bfexplorer to practice mode; switch practice mode off and you will place real bets.

BotTriggers are not distributed with bfexplorer app instalation. You can either download them from my gist, or when testing some of strategies on Betfair AI Trading community.

HorseRacing/BookmakersOdds.md

LastPriceTraded and you must run Cancel Strategies on Selection on all selections. 

The other alternative is write your own custom trigger bot script.

When compiling you certainly have got errors, so just read the errors and try to fix the problem.

The same project is included to my bfexplorer app solution, just to check possible compatibility issues if I had changed anything in the code base of the SDK. I have got no compiling errors for this project.

I did not yet released SDK for preview version which is build on .net 7/8.

Yes, use web page to login, in the Application / Settings, then when login you will use betfair web page login dialog with two factor authentication. 

There are two versions of bfexplorer apps. The old version released on 28th March 2023. This app version is built on .net framework 4.8

The latest version I call the preview is relaesed o .net 7, and you can install it from the link above, from my dropbox drive.

This preview version contains the latest version of new bot stratregies like the one described in this post: Execute Strategy Rules Bot

Open Task Manager and check how much CPU and Memory your bfexplorer app uses on your desktop pc. Bfexplorer uses really small amount of CPU but used memory depends on how much markets you open, or how many bot strategies you executes concurrently.

When using Strategy Bot Executor tool, markets are monitored only from set time and when bot strategy ends its execution the market monitoring is automatically stopped.

Quite strange:

I am looking for an experienced F# programmer to assist me with a project involving Betfair. Specifically, my project requires a script that can be run on Betfair’s bf explorer. The script needs to have the capability to place bets, based on a detailed strategy I have in place. My end goal is to efficiently place bets and manage them without too much hassle. I am expecting that the total timeline from start to finish should be less than two weeks for this project. The successful freelancer will have extensive knowledge and experience in F# programming as well as Betfair.Applicants with prior experience with using bfexplorer on Betfair are strongly encouraged to apply. Do you think you are a good fit for this project? I look forward to hearing from you soon!

https://www.freelancer.com/projects/php/script-run-explorer-betfair/details

When open Execute Trigger Strategy dialog, either by adding or updating the strategy, select TriggerParameters and in the right column click on …

Trigger Parameters dialog appears and there type:

CsvFile

and to other column:

c:\run.csv

Select BotName/StrategyName and in next column click on …

Here you can choose your strategy, of course you must create such strategy before, using any of existing strategies like Place Bet and so on.

The strategies represents exact betting or trading strategy you can parameterized by entering different values to available strategy parameters.

Execute Trigger Strategy just triggers execution on dedicated selection/s, and what is executed is defined by the parameter: BotName/StrategyName

HorseRacingCsvSelectionBotTrigger loads csv file where you define selection name (horse name) and strategy to execute on this selection/horse.

CSV file must use ; as the field separator.

Yes, if no data are in csv file or there is no selection with set horse name then bot execution does nothing, ends execution.

MySelectionsBotTrigger script is simpler one. The csv file contains just selection names, so loads the file defined by parameter CsvFile, you put there the full path name of the cvs file.

If on a market there is your selection/s then strategy is executed on those selections. The strategy is set by BotName/StrategyName in the Bot/Strategy dialog.

I do not remember for whom I made those two scripts, but the functionality of any script/code can be known by reading the code. If you wan to learn exactly what is going on when the script code is executed, you should build bot assembly from the script, then you can debug the code, like I do here:

http://bfexplorer.net/Products/BfexplorerBotSDK

All scripts were programmed to the old version of bfexplorer.

The preview release have a problem with compiling the script, so you actually have to use the second path of bot execution, so making from the script the bot assembly you can execute with the Execute Trigger Strategy.