A very simple change that would improve CFM a lot

[adembroski]

Quote: Originally Posted by Sphinx

This would be fantastic if it were implemented. I'm assuming during your time at EA you worked with one or both of these guys and built friendships with them. My question is, are John White and Ben Haumiller open to ideas that weren't their own? Are they willing or even able to implement something like this via a patch?

I missed your second question. I think they could do it with a patch, but I honestly don't know for sure. All I really remember was some vague idea that you can patch code but not data, but you could tune-file data but not code. This change might require both, I'm not positive. I don't know if you can patch and tune-file for a single issue, simply because users can delete the tunefile and you have data that doesn't match what the patch expects.

So... TLDR version, I really don't know. That's an engineer question. I'm confident it can be done. What risks are associated with it, I do not know.

[Sphinx]

Quote: Originally Posted by adembroski

I missed your second question. I think they could do it with a patch, but I honestly don't know for sure. All I really remember was some vague idea that you can patch code but not data, but you could tune-file data but not code. This change might require both, I'm not positive. I don't know if you can patch and tune-file for a single issue, simply because users can delete the tunefile and you have data that doesn't match what the patch expects. So... TLDR version, I really don't know. That's an engineer question. I'm confident it can be done. What risks are associated with it, I do not know.

It would be great if they could implement this type of change and get it tested before the release. But it's probably too late in development to shoehorn it in. Hopefully, they take your input as someone who worked on the product in the past and run with it. Whether it's your exact idea or something similar. We need to start seeing cpu teams building their teams around their own schemes. In last years game I was constantly editing players to fit the scheme of the team they were on, within reason anyway.

[adembroski]

Quote: Originally Posted by Sphinx

It would be great if they could implement this type of change and get it tested before the release. But it's probably too late in development to shoehorn it in. Hopefully, they take your input as someone who worked on the product in the past and run with it. Whether it's your exact idea or something similar. We need to start seeing cpu teams building their teams around their own schemes. In last years game I was constantly editing players to fit the scheme of the team they were on, within reason anyway.

I don't go that far, but I do have a cursory look through the rosters and see if anything's glaring. If a 4-3 team picks up a Von Miller or something, I'll move him to end. Obvious moves like that.

[4thQtrStre5S]

Awesome idea!! Liked and RT..

[Sphinx]

Quote: Originally Posted by adembroski

I don't go that far, but I do have a cursory look through the rosters and see if anything's glaring. If a 4-3 team picks up a Von Miller or something, I'll move him to end. Obvious moves like that.

In the early seasons I go a little crazy trying to keep things clean and making sense. But after a few seasons it becomes too much work. I hope they implement your change. It will save me tons of time and headaches.


Sent from my Pixel 3 using Operation Sports mobile app

[adembroski]

Quote: Originally Posted by Sphinx

In the early seasons I go a little crazy trying to keep things clean and making sense. But after a few seasons it becomes too much work. I hope they implement your change. It will save me tons of time and headaches. Sent from my Pixel 3 using Operation Sports mobile app

I'm going to try to be a little more active in the wishlist arena this year. There are so many minor changes they could make under the hood to make the whole thing feel more real. Most of the big, sweeping features people ask for would do little to improve the mode because the foundation is shakey. This is just an example of that... and it's a very simple fix that could address a ton of issues.

It'd be nice if the CPU was smart enough to make appropriate position changes, but another way to address it would be to built the depth chart so it fixes the problem without complicated AI.

Imagine if Aaron Donald became a free agent; 4-3 teams would want him to be DT1, 3-4 teams want him to be RE1. What's the common thread? He's a 3-tech interior lineman. Why not a depth chart that makes 3-Technique a position instead of either RE or DT? Then he'd be in the right spot no matter where he signed. Same applies to edge rusher for RE/ROLB, or Stack Linebacker for MLB/ROLB in 4-3 Under, MLB1 and MLB2 in 3-4.

The idea is some "AI" can be handled with common sense structural fixes rather than trying to find the right combination of nested if-else statements that are prone to buggy behavior.

[stinkubus]

Would it be easy to get a majority of teams to prioritize players who would be slightly lower potential OVR? If the AI min-maxes the development game then the only way the human could keep up would be to do the exact same.

Also would expect all rosters to converge to a very narrow band of OVRs unless some teams were intentionally making mistakes.

Edit - Team OVR could still see some variance because the archetypes/schemes are definitely not all balanced, and some archetypes aren't used by the AI teams at all (Scrambler QBs, Receiving HBs, might be others). Would make for a fun test to see if any of my theories about the utility of various schemes are correct.

[adembroski]

Quote: Originally Posted by stinkubus

Would it be easy to get a majority of teams to prioritize players who would be slightly lower potential OVR? If the AI min-maxes the development game then the only way the human could keep up would be to do the exact same.

Are you saying you'd prefer a franchise mode in which your goal is NOT to build the strongest possible roster? I'm a bit confused. Kind of a "you say that like it's a bad thing" sort of situation.

Quote:

Also would expect all rosters to converge to a very narrow band of OVRs unless some teams were intentionally making mistakes.

I don't think so. The CPU isn't singing ONLY it's scheme fits, it's signing the best players to fill a spot. If you go through a draft or free agency and look at the number of players who are the best fit for your scheme but are not "scheme fits" in the Madden sense, you'll see what I'm saying. If the CPU prefers Elusive HBs, but there's a power HB available in the draft who ALSO happens to have the highest Elusive rating, he'll take that guy. The CPU will be getting the best player available, but wont be loading up on scheme fits to the expense of best player available.

Quote:

Edit - Team OVR could still see some variance because the archetypes/schemes are definitely not all balanced, and some archetypes aren't used by the AI teams at all (Scrambler QBs, Receiving HBs, might be others). Would make for a fun test to see if any of my theories about the utility of various schemes are correct.

In the long run, you'd like to see some sort of acquisition philosophies.

Both in terms of "how close are we to our window" (putting teams in "Win Now" or "Rebuild" mode) and how they prefer to seek people out (Build Through Draft, Free Agency - this would require that the CPU can analyze both the draft and the free agency period at the same time and make an educated guess as to who'll be available at their draft spot AND be able to target and move up to get those guys).

What I'm offering here is a quick-and-dirty solution that will significantly improve the AI from what it is... not make it perfect. The highest potential peak scheme OVR IS NOT necessarily the optimal play.