Unrealistic pass completion rate in lower leagues - the death of direct game.

On 25/02/2022 at 10:30, Kyle Brown said:

Everyone, it's not the news you're looking for, it's not been tweaked for this update, nor do we think we will have anything concrete to implement in FM22. 

It boils down to the complexity of the issue and the amount of testing we would feel comfortable with before releasing a change like this into the game. 

Something to keep in mind is that flatly comparing real passing date to FM isn't reliable:

• There are different ways of tracking a pass, and different data providers will provide different stats. We use SciSports and they typically have higher completion rates that the others. 
• We will always know the AI's intention in game, unlike in real life. Think of a fluke cross that ends up in the net - pundits would debate this post match, but we know the true intention and could track it accurately. The same goes for clearances that end up at the feet of a teammate, for example, we can say with certainty if the AI meant it as a pass or a hit and hope. 

This isn't to say we don't feel there needs to be changes - we're not saying everything is fine.

We will always look to make improvements, we just have to err on the side of caution with this particular issue. Simply put, more time is needed to make a change this underlyingly big. 

I can only apologise, but the issue is right up the top in terms of priority and we will do everything in our power to ensure it's improved upon for next years edition. I was only informed yesterday that this wasn't quite making the cut or else we wouldn't have kicked the can this far down the road. 

 

 

This is insane...

FM23 isn’t due out for another 7/8 months, but apparently there is nothing that can be done to fix FM22 in that time? 

This bug is completely game breaking. Unrealistic possession stats, pass completion % and overall number of passes. 
 

 

On 02/03/2022 at 12:30, kepz said:

somewhere I could read more about their metrics compared to others

Disappointed that I haven’t received an answer to this. 

Unfortunately its not really something we're going to have on hand. I would suggest using the SciSports website Improving Football Performance - SciSports. 

To the best of my understanding, when it comes to the actual nuts and bolts so to speak, there's going to be proprietary information on SciSports side and information from an FM perspective that may not be permissible to share publicly. 

I’m not trying to be difficult, but you’re saying that the reason the passing numbers are so high is because of Scisports. So I ask for some insight into this, and you tell me that you don’t have anything that explains this. So where did you come up with this idea in the first place? 

Well there's a difference between us providing general information, which will be based on observations and/or known factors and then providing specific breakdowns and details. It's not the case that there isn't anything that explains this, but rather its not information readily available and there can be other considerations involved in terms of what can be disclosed about SciSports own system. 

With almost every aspect of the game we provide general information and its much rarer for exact methods/specifics/calculations to be shared. 

I will mention this internally to see if there is more information we can provide, but I don't want to give any false expectations.

On 27/03/2022 at 17:43, Michael Sant said:

its much rarer for exact methods/specifics/calculations to be shared.

I’m sorry, but I am not asking you for SI’s under the hood methods. 
 

I am asking you for some literature or anything from SciSports that explains why they themselves have higher completion percentages than, say, Opta.
 

That was the official explanation that was given, now all I am asking for is *how did you come to that conclusion?* I am not asking for the 0’s and 1’s here, I am asking something very basic and straightforward in response to what has already been brought up. I don’t really care how it’s happening in the game or whatever, I simply request that you show  me how SciSports comes up with an inflated number of pass completions in their data. 

On 29/03/2022 at 17:31, kepz said:

I’m sorry, but I am not asking you for SI’s under the hood methods. 
 

I am asking you for some literature or anything from SciSports that explains why they themselves have higher completion percentages than, say, Opta.
 

That was the official explanation that was given, now all I am asking for is *how did you come to that conclusion?* I am not asking for the 0’s and 1’s here, I am asking something very basic and straightforward in response to what has already been brought up. I don’t really care how it’s happening in the game or whatever, I simply request that you show  me how SciSports comes up with an inflated number of pass completions in their data. 

We wouldn't be able to speak on their behalf. In relation to the issue raised specific to this thread, it's something we've responded to and given there's nothing further we can gain from feedback on this at this time, we'll close by saying the issue is something we're very much aware of and will be looked at for future versions of the game. Thanks.