Bf Bot Manager Simulation Mode – How accurate is it?

Simulation Mode in Bf Bot Manager

Simulation mode, according to BFBM, is only to be used to check if the bot functions and triggers as planned. They say it is not meant to be used in any way as a method to research or find if a strategy makes profit. I think they say that to cover themselves against complaints. I always use simulation mode as an important step in the research process. We just need to understand how it works.

 

Key points:

 

Simulation mode uses DELAYED data. The data is delayed 30-120 seconds. So the odds data is delayed but the inplay stats data is not.

So, if we bet at 2:30pm, we actually bet on the 2:28pm odds.

This means that pre-match simulation mode bets are going to be fairly accurate since odds don’t vary that much in a 2 minute window.

Pre-match 2:30pm and 2:28pm odds don’t change so much.

But in-play, If we bet on a 0-0 score at 30m in the game, then we are betting at 30m game time on 28m odds. This means draw odds might actually be a little lower at 2:30pm than at 2:28pm.  If we are backing the draw when it is 1-0, then the odds will be drifting upward, so the 2:30pm odds would be a little higher than the 2:28pm odds. Not much, but a little bit. So in-play simulation bots still have meaning but are not as accurate.

One problem might be that if a goal goes in at 2:30pm, and the bot bets in simulation mode, we might be betting on the 2:28pm odds when the match was still 0-0. We can counter this by adding a “delay bet by 150 seconds after market status change” rule. Depending on the bet the bot is making, there might be ways to mitigate the data delay.

 

Simulation mode matches ALL of the stake, even if there isn’t enough liquidity there in real life. So for smaller markets, it may be inaccurate, but for larger markets it is probably okay. It is still useful for establishing positive ROI at level stakes for a core trend though. The liquidity issue just means we don’t know how scalable a bot is. If you wanted to us a simulation bot but avoid low liquidity bets then you can add rules to let it bet only when a certain amount of money is available on the 1st price, or even a combination of the 1st, 2nd and 3rd amounts of money by using the Selection Condition “Only if available amount >/<” rule.

 

 

Simulation mode only matches unmatched bets placed at higher odds IF the odds get to that level. In the old BFBM it matched bets instantly, even if the odds had not reached that level yet, but in the current BFBM it is better, and it only matches unmatched bets at higher odds when the market actually gets to those odds. So with my bots placing bets at the opposite odds +4 ticks, it will wait until the back odds actually reach that level in the market and then change the simulated unmatched bet to matched. But if we have an unmatched bet, and the odds reach the required level at 2:29pm, but a goal goes in at 2:30pm, the simulation bo will cancel that bet, because it has the 2:28pm odds, and thinks a goal has been scored, thus cancelling the unmatched bet because of a suspended market. With hedging and trade out bets, the results can become even more inaccurate since there is more interdependency on odds, timing and livescores.

 

If we bear these 3 things in mind, then we can still get useful information from simulated bets, even though BFBM themselves discourage it.

 

Once you have a strategy that works in simulation mode, the next step is to try it in real mode with SMALL stakes.  

Author: Mary Riley