Donny Moore, the 'Madden Ratings Czar', Leaving EA

[Gman 18]

Quote: Originally Posted by DCEBB2001

I have so much progression data already built into my model that it would have to at least be discussed. I literally can do an analysis to tell you how much players increase or decrease on the average. I can break it down any way I want to and get the 30,000 foot level or go player-specific, but it just depends on what they want and what they want me to do. Honestly, most of the change happens at the year-end evals. That is where the steepest changes are seen. Weekly is rare (maybe 100 players at most on a given week and rarely a player has a change two consecutive weeks), and those are never really drastic.

It would be a breath of fresh air to have someone who uses actual PLAYER DATA to determine player progression/regression in CFM. Progression determined by XP is certainly not the way to go and I hope Dan's suggestions make EA think otherwise and hopefully they decide to implement his ideas


Sent from my iPhone using Tapatalk

[Skyflame21]

Not sure that this isn't a positive move for the game. Unless they get someone even worse. Rating players is a huge undertaking and obviously not an easy task to do. However over the years it seems ratings for players were based on the ESPN and fan hype and less about their actual on-field abilities. Sometimes tough live is needed to players all can't be great. I wish him all the best on his new endeavors.

[bucky60]

Quote: Originally Posted by DCEBB2001

This is the whole "if a tree falls in a forest and nobody is around to hear it, does it make a sound?" argument. The scouting data is the confirmation of a player's ability. Of course the player changes his ability at the time of or just prior to it being recognized (and subsequently recorded), but that is not how the data works in scouting. There is always time passage between the evaluations. I believe you are confusing the scouts actually changing their abilities with my previous statements of the scouts changing the GRADES to mimic observed changes in ability.

With progression In Madden, what observed changes in ability are you talking about that the scouts see? Stats? Stats should never drive progression for a realistic progression model. Allowing a scouting dept in madden to drive progression is unrealistic. Other factors should be determining progression/regression. Realistic factors. Scouts should ONLY be evaluating a players perceived abilities.

I think your progression system confuses re-rating with progression. When you are defining a roster for a Madden release, I agree with your method. But not for progressing players. It's unrealistic to use stats for progression. It's unrealistic to use scouting evals for progression.

[bucky60]

Quote: Originally Posted by Gman 18

It would be a breath of fresh air to have someone who uses actual PLAYER DATA to determine player progression/regression in CFM.

I happen to disagree with using player stats to drive progression. It's unrealistic. I personally think it would be a breath of fresh air to use a realistic model of progression.

[DeuceDouglas]

Yeah, Madden 05 had some of the worst progression I'd ever seen. I remember one season I had Gary Stills go up like 15 OVR in one postseason and basically go from a special teamer to pro bowler because he recovered a couple fumbles and returned one for a TD. IMO progression should be based on a combination of potential and coaching. There are certainly other factors but I think at its most basic form you could create a very good progression model based on those two elements.

[DCEBB2001]

Quote: Originally Posted by bucky60

With progression In Madden, what observed changes in ability are you talking about that the scouts see? Stats? Stats should never drive progression for a realistic progression model. Allowing a scouting dept in madden to drive progression is unrealistic. Other factors should be determining progression/regression. Realistic factors. Scouts should ONLY be evaluating a players perceived abilities. I think your progression system confuses re-rating with progression. When you are defining a roster for a Madden release, I agree with your method. But not for progressing players. It's unrealistic to use stats for progression. It's unrealistic to use scouting evals for progression.

No, not stats. I would create a progression model based on the amount of progression we see for players in the historical scouting data. I addressed this on post #185 in this thread. I am not a big fan of basic stats at all. I think we can debate HOW it should be implemented, and believe me I am all ears on that, but I think that the data can give a reasonable control for the level of progression we would see in the model. Find the average amount of progression we see in both the negative and positive directions, find your standard deviations, and make it fit in a bell curve. That could be a solution to make sure that things don't get out of hand within the model.

I would think that we would have to tie it something that the user does, however, instead of making it random. Win/Loss record? Coaching? Efficiency? I don't really know what, but I think we at least need to put some of it in the users hands.

[DCEBB2001]

Quote: Originally Posted by bucky60

I happen to disagree with using player stats to drive progression. It's unrealistic. I personally think it would be a breath of fresh air to use a realistic model of progression.

What is your idea of a "realistic model"? You posted something to this effect in two consecutive posts so I am wondering if you already have an idea outlined.

[DCEBB2001]

Quote: Originally Posted by DeuceDouglas

Yeah, Madden 05 had some of the worst progression I'd ever seen. I remember one season I had Gary Stills go up like 15 OVR in one postseason and basically go from a special teamer to pro bowler because he recovered a couple fumbles and returned one for a TD. IMO progression should be based on a combination of potential and coaching. There are certainly other factors but I think at its most basic form you could create a very good progression model based on those two elements.

I totally forgot about the potential grade to this. That may be a good idea considering that it is up to the user to find players with high potential. The higher the potential, the higher the likelihood that the player will positively progress. Throw confidence into that and you may have a nice combination there.