Connect with us

College Football 25: Adaptive AI Explained

EA Sports College Football 25

College Football 25: Adaptive AI Explained

One of the major “under the hood” features of College Football 25 is the adaptive AI, known as the Real Time Coaching (RTC) engine. It’s not that complicated, but it’s worth running through how it works.

Recommended Videos

This is not “comeback code” built into the game, and it’s not coach-specific, it’s just a way for coaches/AI to adjust as games progress. It’s also overarching, so it encompasses things you maybe wouldn’t think of as relating to “adaptive AI” in the normal sense.

I’m actually going to include EA’s exact words about the RTC engine below because it’s probably worth reading again now that the game has been out for a couple months:

Making its debut in this year’s edition, the Real Time Coaching (RTC) feature is an adaptive coaching engine designed to authentically enhance both offensive and defensive CPU gameplay.

In this RTC engine, coaching aggressiveness possesses unique adaptive adjustment levels. This means that different coaches will react and adapt differently to successful run or pass games, thus creating a more authentic, dynamic, and unpredictable game environment.

Additionally, the AI now has access to the full range of adjustments, just like a human player would make. As a result, you can expect to see teams employ an array of adjustments such as shade techniques, showing blitz, run commits, QB contains, and Spy’s, to name just a few situational adjustments.

The AI will also recognize and adapt to your offense personnel. For instance, if you’re dominating a team with the inside zone out of 11 personnel, you’ll notice the CPU tweaking their defensive line, providing a different look to potentially alter your play call. If you’re having significant success on the ground, you may find some teams loading the box earlier than others.

Furthermore, the AI is designed to be aware of your players’ ratings and abilities. If you have an elite wide out, some teams might double him from the jump or match him with their top corner. This is an example of the ‘Best on Best’ matchup strategy enabled by RTC.

The defensive AI is also capable of disguising coverage, adding another layer of challenge to the gameplay. Situational adjustments are another key feature of this new engine. Despite being in its first year, this engine significantly enhances pre-snap adjustment chess matches. For example, in scenarios where you’re inside the 10 with man coverage, expect to see more press and inside shade technique forcing the ball out and over the top.

Another example of situational adjustments is when the opposition is leading by three scores late in the game. In this case, you’ll notice them keeping the ball inside and tackling conservatively to keep the clock ticking.

In conclusion, this is just the first year of the implementation of this foundational feature. The groundwork for the ultimate strategic experience has been laid, and we can’t wait to see where it goes from here.

The big things to take from that are the RTC engine exists to try and make AI teams play more like humans. So they have access to all the pre-play adjustments, coaching adjustments in the playcall screen, and will use all the gameplay features during plays as well.

I think what goes overlooked is how you should think about AI coaches. To me, the best thing you can do to understand adaptive AI is look at the previous play in the playcall screen:

You use the right stick to flick between the pages in the top right corner (assuming you aren’t in a formation/screen where it would instead change your Coverage Shell or whatever) to see the previous play. Early in games, this is a screen that I find especially useful to check because AI defenses will generally be somewhat consistent. They will lean towards man blitzes, zone blitzes, man, or specific zone coverages, but they don’t bounce as much between different looks if something works. They want you to prove you can beat their looks before they start adjusting.

One other aspect that I think is worth calling out is the “aggressiveness” cited by EA in the explanation of the RTC engine. There are “offensive aggressiveness” and “defensive aggressiveness” sliders you can set for your coach, but those sliders don’t have anything to do with on-field gameplay. Those exist for Super Sim/simulating games in general. They don’t relate to how passive or aggressive your coach or other coaches will be as games play out on the field when you’re in control.

I would guess “coach schemes” might matter some for on-the-field plays because certain defensive and offensive playbooks might just have more blitzes or deep pass concepts than others, but I think most of the aggressiveness is tied into game situation or things playing out under the hood that we can’t control or influence with specific coaching settings.

All in all, I think the RTC engine is one of the better creations EA has come up with for its football games in recent times, whether or not it’s wholly unique or new. The AI is more unpredictable, and the variance between teams is more pronounced this year. At least part of that has to be the RTC engine. Still, it’s worth understanding what levers this engine is and is not controlling, so hopefully this has helped in that regard.

About the author

Chase has written at Operation Sports for over 10 years, and he's been playing sports games way longer than that. He loves just about any good sports game but gravitates to ones that coincide with the ongoing real seasons of the NBA, NHL, MLB, NFL, and so on. As of now, he's gearing up for EA Sports College Football 25 and what should be a wild summer while still dabbling in the latest Top Spin and MLB The Show.

More in EA Sports College Football 25


Related

To Top