Do Player Ratings Matter in All-Madden? Here’s an Experiment

[Vanilla_Gorilla]

So during my Texans franchise, a ran a zig route with Will Fuller against Man Coverage that resulted in an 8 yard gain against perfect coverage by Jonathan Joseph. I wondered how good Joseph’s ratings were to be able to cover the zig so perfectly, and they weren’t too shabby, 84 man, 84 press, 81 play rec.

This kind of got me curious. I wanted to see just how ratings stack up in All-Madden with User defenses and CPU defenses.

Unfortunately, the roster I have loaded doesn’t have Joseph on the Titans roster, but Adoree Jackson does have similar ratings physically and coverage of wise with slightly less play rec and less press (which doesn’t matter due to the play call and route).

I ran 25 versions of the same play as the offense and the defense. Same players. No interactions. Same plays. All Sliders set to 50.

The way I judge whether Jackson won the play was either
A) Immediate Tackle
B) Coverage Animation but Caught
C) Coverage Animation PBU
D) Close Coverage
E) PBU
F) Interception

Obviously, Fuller wins if he catches the ball and gets solid RAC yardage (I.e. he escapes away from Jackson, or outright beats him).

*User Offense*
Jackson won 14 out of 25 snaps. Good for winning the play 56% of the time.
1 play resulted in an interception.
4 in PBUs (total.
The rest were immediate tackles.

Upping CPU Pass Coverage and Reaction Time Sliders to 75 Jackson won 17 out of 25 snaps good for winning 68% of the time.
1 interception
4 PBU’s

Moving up to 100 CPU Coverage and Reaction Time, Jackson won 15 out of 25 Snaps. Winning 60% of the snaps.
0 interceptions
6 PBUs

*User Defense*
When controlling the Titans defense, Jackson won just 6 out of 25 snaps. Good for winning the play just 24% of the time.
No interceptions.
No PBUs.

When upping the User Pass Coverage and Reaction Time Sliders to 75 each Jackson won 9 out of 25 snaps, which is a bit better at 36%.
1 Incompletion
2 PBUs
No interceptions

When upping User Pass Coverage and Reaction time Sliders to 100 each Jackson won only 5 of 25 snaps at just a 20% win rate.
1 Incompletion
1 PBU
No interceptions

*Conclusion*
This is just a brief, but nonetheless, disheartening experiment. I’m sure there’s more ratings at work than the one’s I believe are the main factors, but User defenders are at a clear disadvantage within the game’s engine compared to the CPU. This is likely why playing defense against the CPU can just be...just plain not fun at times. Your ratings aren’t mattering in the grand scheme of things. The sliders also don’t seem to be as effective of a bandaid either.

[khaliib]

Re: Do Player Ratings Matter in All-Madden? Here’s an Experiment

1) the whole point of AM is to create disadvantages (manipulate ratings/die rolls) towards the User as a way of producing greater AI difficulty.

2) different throw/release timing between User and AI QB, which impacts Break-on-Ball triggering, thus outcomes

3) greater Fatigue applied to User players than AI, again to produce greater AI difficulty

4) different times of TUP penalty possibly being applied due to Pass Rush not being exactly the same every snap, thus impacting outcomes

**If this is Next-Gen, new Player Movement will also impact outcomes also

[Vanilla_Gorilla]

Re: Do Player Ratings Matter in All-Madden? Here’s an Experiment

1) the whole point of AM is to create disadvantages (manipulate ratings/die rolls) towards the User as a way of producing greater AI difficulty.

2) different throw/release timing between User and AI QB, which impacts Break-on-Ball triggering, thus outcomes

3) greater Fatigue applied to User players than AI, again to produce greater AI difficulty

4) different times of TUP penalty possibly being applied due to Pass Rush not being exactly the same every snap, thus impacting outcomes

**If this is Next-Gen, new Player Movement will also impact outcomes also

1) That should not be how a Sports game’s difficulty works. What’s the point of developing players if their ratings are handicapped in the actual game? Higher difficulties should mean a more difficult AI, not worse user ratings. Sports is not dice rolls.

2) Throw variation was minimal. I made sure to only include plays where it was an accurate bullet pass, this is including pass rush as well. The zig route against man is quick, so pass rush wasn’t much of a factor. If it affected the play, I didn’t count in. I was looking purely at the DB/WR interaction.

3) Practice mode, so fatigue not a factor.

[SolidSquid]

Re: Do Player Ratings Matter in All-Madden? Here’s an Experiment

All madden is designed to challenge the user not be a fair ratings driven mode

[Vanilla_Gorilla]

Re: Do Player Ratings Matter in All-Madden? Here’s an Experiment

All madden is designed to challenge the user not be a fair ratings driven mode

That’s just silly though. What’s the point of developing a guy to 90 overall or signing that 90 overall player in FA if he’s never going to play to his rating?

All Madden should challenge the user, absolutely, but not by just blanket reductions in how a player’s ratings work. Why isn’t the AI just better and smarter with maybe some minor penalties towards the user?

[Darth Aegis]

Re: Do Player Ratings Matter in All-Madden? Here’s an Experiment

Ratings matter maybe 1% of the time imo. It's all the settings under the hood that make things happen or not.

[khaliib]

Re: Do Player Ratings Matter in All-Madden? Here’s an Experiment

That’s just silly though. What’s the point of developing a guy to 90 overall or signing that 90 overall player in FA if he’s never going to play to his rating?

All Madden should challenge the user, absolutely, but not by just blanket reductions in how a player’s ratings work. Why isn’t the AI just better and smarter with maybe some minor penalties towards the user?

Every video game does this my friend, not just sports games.

Have you ever heard these terms applied to Madden...?
- Skynet
- Robo QB
- Psych DB

Its a fine line that’s difficult for a Game Dev, as what’s a sweet spot for you, would terrible for 10,000 other gamers.

Madden just needs to bring their Gameplay Sliders up to par of the genre.

This is the only sports game in the genre that remains so restrictive towards its users.

[PhillyPhanatic14]

Re: Do Player Ratings Matter in All-Madden? Here’s an Experiment

So what's the verdict? Is it better to play All Pro with tough sliders? Regular all pro is just way too easy. All Madden is a challenge but defense feels a bit like the user has no control over anything at all.

[OJsakila]

Re: Do Player Ratings Matter in All-Madden? Here’s an Experiment

So what's the verdict? Is it better to play All Pro with tough sliders? Regular all pro is just way too easy. All Madden is a challenge but defense feels a bit like the user has no control over anything at all.

I'm leaning towards AP with adjusted sliders... In years past, AM leaned on the side of magical moves by the cpu... Maybe this is different now... And I'm on new Gen and really enjoying the feel of the game... I would start at something like Matt10's AP sliders and work from there... Just an idea as opposed to trying to figure it all out from the get go. Read all the slider threads with a ton of replies and try and understand how the sliders work before you start messing with them... Some work opposite of what you think and others tremendously effect things you wouldn't think they effect... In particular, penalties.
Do your research and try and find something you can work with..

[Vanilla_Gorilla]

Re: Do Player Ratings Matter in All-Madden? Here’s an Experiment

Every video game does this my friend, not just sports games.

Have you ever heard these terms applied to Madden...?
- Skynet
- Robo QB
- Psych DB

Its a fine line that’s difficult for a Game Dev, as what’s a sweet spot for you, would terrible for 10,000 other gamers.

Madden just needs to bring their Gameplay Sliders up to par of the genre.

This is the only sports game in the genre that remains so restrictive towards its users.

That’s terrible on the Madden Devs part then, because with just a little bit of sleuthing like I did, you can just find out that there really isn’t a point in playing the game anymore.

What’s the point of playing the game when your user controlled defense is essentially reliant on the pure chance that X play beats Y play or you get pressure on the QB? That restriction of your players does nothing of service to the game and only hinders the player’s meaningful interaction with the game.

The best game’s I’ve played don’t interfere with the player’s functions in the core aspects of the game, or do so very minimally and instead manufacture difficulty in the accessories to the main game, and of course the AI.

[Landonio]

Re: Do Player Ratings Matter in All-Madden? Here’s an Experiment

All-Madden was proven to be unfair like 20 years ago. Why do people still insist on playing on it. Pride? Nobody cares what difficulty you play video games on. Games are meant to be fun, not exercises in frustration.

[Hooe]

Re: Do Player Ratings Matter in All-Madden? Here’s an Experiment

1) That should not be how a Sports game’s difficulty works. What’s the point of developing players if their ratings are handicapped in the actual game? Higher difficulties should mean a more difficult AI, not worse user ratings. Sports is not dice rolls.

A computer cannot digitally simulate the continuous range of results of human athletic performance without dice rolls. The ratings, difficulty, sliders, and user inputs all load the dice in different ways to produce dynamic outcomes. All simulation sports video games work this way, Madden is no exception.

An example - without randomness, an offensive lineman with RBK 90 always wins against the defensive lineman with BSH 85. It rarely happens in the NFL that one player so effortlessly and consistently dominates all of his competition, and even the best players make mistakes. Randomness allows that to happen. I would be curious as to how you think this should happen without randomness.

That said, I agree Madden needs to get smarter. Much smarter. If you know how to ID coverages pre-snap, how the different coverages work, and how to work hot routes, even All Madden difficulty is trivial with just a handful of offensive formations.

We need smarter CPU coaches which understand fronts, coverages, and formations. We need CPU coaches which can not only learn your play calling tendencies and call plays appropriately, but predict what plays you will call based on how you tend to call plays and what personnel you are putting onto the field. We need the game to surface opponent play calling data during games and intelligently suggest play calls based on recent and historic opponent tendencies.

Coaches in Madden 21 still run plays based on a very simple script originally built for Madden 11 (GameFlow). The CPU's lack of understanding of how to effectively use its own personnel by far the weakest component of the game. We can do better.

[Vanilla_Gorilla]

Re: Do Player Ratings Matter in All-Madden? Here’s an Experiment

A computer cannot digitally simulate the continuous range of results of human athletic performance without dice rolls. The ratings, difficulty, sliders, and user inputs all load the dice in different ways to produce dynamic outcomes. All simulation sports video games work this way, Madden is no exception.

An example - without randomness, an offensive lineman with RBK 90 always wins against the defensive lineman with BSH 85. It rarely happens in the NFL that one player so effortlessly and consistently dominates all of his competition, and even the best players make mistakes. Randomness allows that to happen. I would be curious as to how you think this should happen without randomness.

That said, I agree Madden needs to get smarter. Much smarter. If you know how to ID coverages pre-snap, how the different coverages work, and how to work hot routes, even All Madden difficulty is trivial with just a handful of offensive formations.

We need smarter CPU coaches which understand fronts, coverages, and formations. We need CPU coaches which can not only learn your play calling tendencies and call plays appropriately, but predict what plays you will call based on how you tend to call plays and what personnel you are putting onto the field. We need the game to surface opponent play calling data during games and intelligently suggest play calls based on recent and historic opponent tendencies.

Coaches in Madden 21 still run plays based on a very simple script originally built for Madden 11 (GameFlow). The CPU's lack of understanding of how to effectively use its own personnel by far the weakest component of the game. We can do better.

I can get behind this.

I do believe the blanket ratings decrease needs to go ASAP and the way they program the All-Madden difficulty needs to change.

Like I’ve previously said, there’s just no point in playing the game when the players you spend time and effort into developing wind up not mattering to the game’s mechanics. It’s disheartening and just frustrating.

Like you said, I want the AI to beat me by recognizing my play calls, understanding what plays and formations I use in certain situations, and mixing up their plays and coverages to confuse me. I don’t want them to beat me because my 93 overall CB can’t cover an Out-Route because reasons.

[stinkubus]

Re: Do Player Ratings Matter in All-Madden? Here’s an Experiment

Ratings matter on All-Madden, but they need to be relatively high in most of the crucial skills. If you want a pass rusher to show up you can either learn a blitz that gets them free or have > 80 PMV or FMV. 90 is better. Without an ability to complement it 90 pass rush moves alone are no where near overpowered on All-Madden.

[kennylc321]

Re: Do Player Ratings Matter in All-Madden? Here’s an Experiment

So what's the verdict? Is it better to play All Pro with tough sliders? Regular all pro is just way too easy. All Madden is a challenge but defense feels a bit like the user has no control over anything at all.

Funny you should ask. I find myself asking this same question. Heck... I am trying to figure out if I should make a tough 'pro' or an easy all pro level. Just this weekend I was working on that.