A Vision For a New Ratings System

[Jimbo12308]

Hi all,

So I’m sure this is something that has been suggested before, but below is an idea for a system of player ratings that would, in my opinion, make the game so much more interesting:

To me, one of the biggest issues with Madden is predictability. If you have a starter with even just slightly higher ratings than the backup - then it makes sense to play the starter all the time every time. Any “down” game is just pure coincidence - as is any “good” game from a low level player. It always makes sense to simply make the cold, calculated decisions. Including in free agency. Some guy has slightly better numbers than your guy? It makes sense, every time, to cut your guy and bring in the guy from FA (depending on the contract). It’s boring. I think a few big changes in franchise mode would make a world of difference:

1. Ratings are hidden within in ranges. Physical ratings may be in smaller ranges since those can be “tested” but all ratings would be hidden in a range. That range would shrink the longer the player is in the league (though it may start to expand again when the player gets older to simulate potential regression) and it also shrinks the longer they are in a system. The range would be smaller for the team with the player since the coaches see them in practice. This would drastically alter team building since it injects massive amounts of realistic uncertainty into player evaluation.

2. Ratings fluctuate on a game-by-game basis. Each game, a player’s true ratings - which dictate their play - fluctuate within a smaller range inside their main ranges. This means that “good” and “bad” games aren’t just a coincidence/fluke - they were actually good and bad games. This also means that players’ performances in games could indicate where they may actually be within their more broad range. Let’s say a young player has an overall range from 58-73. We may not know it, but their “true” rating may be a very good 72. In games they would fluctuate between say 65-73. So, if you have a good eye, you could determine from their performances in games that they are at the upper range of their 58-73 rating range. In addition, you could then make a call like benching a player who plays poorly in the first half because you make be determining they they are having a down game. This could even make practice useful. Perhaps for a certain amount of time per week you could practice with players and they will practice at the same level that they’ll play in the game. See a player lighting it up above their normal levels in practice? Might suggest they are on a up week.

These changes alone would:

1. Add uncertainty to the caliber of your roster
2. Add uncertainty to player development decisions
3. Increase the incentive to hold on to players since you have familiarity
4. Add significant amounts of uncertainty to adding free agents
5. Add variability to the performance of your team in games
6. Makes “up” and “down” games from players actually possible
7. Creates scenarios where benching or promoting a player actually makes sense
8. Makes draft picks true mysteries as to their quality until proven
9. Makes practice actually useful as a way to evaluate players (this could honestly branch into a whole system where you could observe practice yourself or you could sim it and you get a report on good and bad performers in practice - the accuracy and depth of that report could depend upon the quality of your staff).

This would obviously have to work around other changes to things like development and scouting, but I think the fixed and certain nature of ratings is one of the biggest detriments to this game.

[canes21]

I support the idea of having more fog of war, but I wish we would get rid of numbers. I don't look at Tom Brady in real life and think he has 47 speed, I look at him and think he is slow. I look at Tyreek Hill and think he is fast.

I'd rather Madden(all sports games) ditch the numerical rating system and simply give us detailed scouting reports that put players and their skillsets into tiers.

For example, Tom Brady as a whole is in the +++ tier of QBs. You can call it the 1 tier, the A tier, the +++ tier, whatever, that's semantics. Then when you click into his player card you'll see a list of strengths.

+++ Football IQ
+++ Leadership
+++ Work Ethic
++ Play Action
++ Footwork
- - - Top Speed
- - - Quickness
- - - Agility

And then a section detailing durability and past injury history and whether or not any past injuries are ones that could lead to further "re-injury" in the future. For example, a guy like Gurley would show he has knee issues and it would have a note detailing how he is likely to have future injuries because of the past injuries and what they've done to his knees.

This system would give the player a great idea of a players overall skill by having them in macro tiers and then breaking down their individual skills into further sub-tiers allows you to compare them head to head, looking at their strengths and weaknesses and making decisions based off of that.

It gives you plenty of information to not be lost, but it rids of the issue you highlighted in your post where you simply know with 100% accuracy who is better than who in a numerical rating system.

The system could and should also implement consistency like you described. That would be a part of a player's scouting report in their player card. Brady would be a very consistent player. His attributes behind the scenes would have a small range of numbers they'd fluctuate between game to game, drive to drive. Other players would have notes saying they are somewhat consistent, not consistent, very inconsistent, and so on. These guys would all have bigger and smaller windows their attributes could fluctuate in behind the scenes.

Also, to keep the scouting report from being an overload of information, it should really only highlight true strengths and weaknesses and note that all else is assumed to be average compared to the player pool at that position. So some very average players would simply have small scouting reports that maybe highlight one or two things and it'll also tell you the player is about league average as a whole.

A system that combines these ideas is one that instantly adds variety to franchise mode for the user even if the CPU logic remains terrible. The user is no longer omniscient and position battles will be a thing again and you'll have valid reason to rotate players in and out to see who does perform best for your team when you don't have a standout player who is the obvious starter.

It provides enough information for the casual gamer, but keeps enough information hidden in the fog so that any hardcore franchise player can find longevity in the mode.

Sent from my SM-N950U using Tapatalk

[jfsolo]

I'm with any of these ideas about changing the ratings systems, but I will say that players with higher OVR and even individual ratings aren't actually always better players in games, not just a game here a there but for even for multiple seasons.

The traits are part of the reason, but also many, many players just over and under perform their ratings. It's one of the reasons I love playing the Preseason games so much. Some 73 OVR players greatly outperform 80 OVR players in terms of production in game.

[canes21]

Quote: Originally Posted by jfsolo

I'm with any of these ideas about changing the ratings systems, but I will say that players with higher OVR and even individual ratings aren't actually always better players in games, not just a game here a there but for even for multiple seasons. The traits are part of the reason, but also many, many players just over and under perform their ratings. It's one of the reasons I love playing the Preseason games so much. Some 73 OVR players greatly outperform 80 OVR players in terms of production in game.

While true, the CPU is always going to start players based off of the OVR rating and nothing else. Having a system that led to the CPU starting players for different reasons like production, consistency, etc. would be much realistic and add more variety to the mode than what we have now where the depth chart practically stays the same weeks 1-17 which we all know does not happen in the league.

[TheButter]

Here a pic of something similar to the OPs idea8d946b0-1.jpeg

[Jimbo12308]

Quote: Originally Posted by jfsolo

I'm with any of these ideas about changing the ratings systems, but I will say that players with higher OVR and even individual ratings aren't actually always better players in games, not just a game here a there but for even for multiple seasons. The traits are part of the reason, but also many, many players just over and under perform their ratings. It's one of the reasons I love playing the Preseason games so much. Some 73 OVR players greatly outperform 80 OVR players in terms of production in game.

While I agree that “overall” can definitely be misleading because some individual skills are definitely better the others, I disagree that certain players over and underachieve. If you take 2 players with the same traits and have one of them be 1 point better in every attribute, in the long run that player will outperform the other player. That’s just how the programming works. Any instances where the lower ability player plays better is simply luck/coincidence/wishful thinking.

That’s what bores me to death. I’ll play a preseason game and have some young player light it up. I go an look at their ratings and traits compared to my starter and can easily see that it was simply a fluke and that there is no logical/mathematical (since that’s all the programming is...math) reason to suspect that the success will continue. When we have the ratings, we have the answer key. The rest of it is just dice rolls in the programming. And the ratings literally outline to us exactly which dice roll is more likely to win.

[Jimbo12308]

Quote: Originally Posted by canes21

I support the idea of having more fog of war, but I wish we would get rid of numbers. I don't look at Tom Brady in real life and think he has 47 speed, I look at him and think he is slow. I look at Tyreek Hill and think he is fast. I'd rather Madden(all sports games) ditch the numerical rating system and simply give us detailed scouting reports that put players and their skillsets into tiers. For example, Tom Brady as a whole is in the +++ tier of QBs. You can call it the 1 tier, the A tier, the +++ tier, whatever, that's semantics. Then when you click into his player card you'll see a list of strengths. +++ Football IQ +++ Leadership +++ Work Ethic ++ Play Action ++ Footwork - - - Top Speed - - - Quickness - - - Agility And then a section detailing durability and past injury history and whether or not any past injuries are ones that could lead to further "re-injury" in the future. For example, a guy like Gurley would show he has knee issues and it would have a note detailing how he is likely to have future injuries because of the past injuries and what they've done to his knees. This system would give the player a great idea of a players overall skill by having them in macro tiers and then breaking down their individual skills into further sub-tiers allows you to compare them head to head, looking at their strengths and weaknesses and making decisions based off of that. It gives you plenty of information to not be lost, but it rids of the issue you highlighted in your post where you simply know with 100% accuracy who is better than who in a numerical rating system. The system could and should also implement consistency like you described. That would be a part of a player's scouting report in their player card. Brady would be a very consistent player. His attributes behind the scenes would have a small range of numbers they'd fluctuate between game to game, drive to drive. Other players would have notes saying they are somewhat consistent, not consistent, very inconsistent, and so on. These guys would all have bigger and smaller windows their attributes could fluctuate in behind the scenes. Also, to keep the scouting report from being an overload of information, it should really only highlight true strengths and weaknesses and note that all else is assumed to be average compared to the player pool at that position. So some very average players would simply have small scouting reports that maybe highlight one or two things and it'll also tell you the player is about league average as a whole. A system that combines these ideas is one that instantly adds variety to franchise mode for the user even if the CPU logic remains terrible. The user is no longer omniscient and position battles will be a thing again and you'll have valid reason to rotate players in and out to see who does perform best for your team when you don't have a standout player who is the obvious starter. It provides enough information for the casual gamer, but keeps enough information hidden in the fog so that any hardcore franchise player can find longevity in the mode. Sent from my SM-N950U using Tapatalk

I would be perfectly happy with a system like this and agree that numbers are kindof dumb. However, the player ratings are a huge selling point/marketing tool for the game so I feel they would likely never go anywhere. If we stick with numbers then in competitive modes and for marketing purposes the players can simply have numbers like they do now. But when a user enters franchise mode it could create the ranges and could randomize where the players’ “true” ratings are in those ranges at the start of the franchise (so us franchise players couldn’t simply refer to the “true” ratings from the competitive side). Again, there would be some logic to the ranges. Due to consistency and the league’s familiarity with him, Tom Brady’s range (maybe 87-95) would be tiny compared to Kyle Trask’s range (maybe 50-80).

But I do think that with a number system that is shrouded with ranges, it can still really create that uncertainty - even though there are numbers. A rookie with a 50-80 could very well be better than a vet with a 76-85. The rookie’s actual rating could be an 80 while the vet’s could be a 77 (and this is obviously simplifying things because of course every attribute would have a range, but for the simplicity of the discussion I’ll just refer to overall ratings). Close observation in practice and in games could make the rookie’s talent clear to a user. Or the dice rolls could play havoc, in a good way. The rookie could have a good game just by luck and the user could assume that the rating is high...then they realize the hard way that they were wrong. Or maybe the good game was true and then a few bad games are flukes. Will the user give up too quickly? Or will they trust what they saw in the good game and ride out more games? Meanwhile, if we had some semblance of locker-room management the vet player could be annoyed that they were benched for the rookie. It creates whole storylines and difficult decisions just based on uncertainty - even if there are still numbers at play.

So while I do like your system, and in many ways I like it more, I do also think it would be more difficult to implement since Madden will want to keep using numbers for the competitive crowd and for marketing purposes.

[forme95]

I love the ideas and would love to see something along the lines of a different rating system. Or, ratings lowered to make better gameplay......... OH wait......... todays simple minded, short attention spanned, want everything easy society.......

Well, it is still a good thought at least