[briz1046]

Quote: Originally Posted by adembroski

The real question is why. While it would be more realistic for teams to play a player who's productive over one whose young and slightly more talented, it would be realistic because reality is not knowing precisely who's better. In Madden, we kinda know who's better, right? This would put CPU teams at a disadvantage. So by doing as you suggest, we're treating a symptom and putting the CPU at a disadvantage. A far better solution would be to treat the disease and insert some uncertainty in talent evaluation. The way I'd do this would be to have ratings replaced by evaluation; your position coach and coordinator's opinion of a guy's talent. Those, btw, might be different, and you as the head coach are responsible for making the final call. It still translates into something like the archetype overall rating, but it's less certain. Your DBs coach, being a guy who favors big press-man type corners might say Richard Sherman's the best CB in the league, while your defensive coordinator, a cover 2 guy, would rather have a Ronde Barber. Now you have to decide; your DC is a more experienced, accomplished coach, but he's a former d-line coach and has a forest over the trees perspective, while your position coach spends more time with the particular player but might not quite have the same level of experience or isn't looking at the big picture and how he fits into the defense. So you get *mostly* accurate but uncertain information. It's very important it's close enough and enough information is available to make an informed decision, but it need not be spelled out for you as it is now. How the CPU would do this is to give the Head Coach the same kind of logic as the DC and DB coach, allowing him to break the tie. This was actually the intent of the "big decisions" feature when I first started talking about it back in '13 (I don't know if my idea was directly responsible for the '17 feature, I only know it resembles what I suggested back in '14 somewhat). The idea was two-fold; 1.) By finding a way to boil as much down to binary decisions, presented directly to the user, we'd increase retention in CFM by not overwhelming inexperienced players with information, and 2.) by presenting the CPU with the same binary choices, we can more easily create logic that would compete with User players. So the plan was to bake as much of the decision-making process into the talent evaluation system as possible. For example, players would ultimately have 3 "Overalls", based on right now, a 3-year projection, and a 5-year projection. This took care of teams drafting QBs in consecutive years because CPU teams would evaluate how close they were to compete for a title and use the appropriate projection to evaluate their roster. Clearly that's not where it went after I left, but that was the idea.

Yes excellent ideas and you're right my idea would have put CPU teams at a disadvantage if the user chose not to "play the game" , badly thought though on my part a consequence of too many years of using 32 team control I'd imagine