Draft Logic and More

[NBryant]

I apologize, I somehow started this in the Colege thread when it belongs here.

First, and it may have been corrected in '22, but on the draft screen in '21 the initial text where the announcer states a team is on the clock - it says ""Team" is now on.... Here "Team" is the team now which makes "is" wrong and needs to be changed to "are".


Logic. Still digging but what is the AI logic when it comes to the draft. I watched a team with an 88 OVR QB on the roster wit ha 5 year contract draft another QB with an OVR of 91 in the first round. Makes not sense at all. As you go through you can find other similar examples. Look, I know the heart of the PF series is it's design being more multiplayer in league settings. I am sure you have noted there really arn't that many leagues though, a handful yes, but for something in this genre one would/could expect more. Even so, in a multiplayer setting there may/will be teams that are set to CPU which could lead them to having the AI draft for them and continue to use a confusing logic. At this stage of the games development is it even feasible to try and strengthen the quality of the AI in this and other instances or is it what it is and thats it?


With this in mind can/would Wolverine share the football genres development plan and timeline for the foreseeable future (over the next 3-10 years). What areas of the game are planned to be upgraded or sytems developed and/or enhanced. What is the market focus. I only throw this out there becasue at this moment you are the only game in town besides the mogul series. With the right mindset and plan this type of game in this genre could be yours much like OOTP has captured that genre and without crumbling internally it is theirs for as long as they pursue it. With my simple minded vision and having seen this game progress from it's earliest vision I, personally, think with a focus on more of the "real" world aspects of the game it would easily explode to an even higher market share. It will never be Madden much like OOTP will never be The Show but it could easily draw much more of an audience interested in not only league play but solitaire as well, if not more.

[brooks_piggott]

Thanks for the feedback. Typically the AI has a lot of factors that it uses for draft picks including coach grades, scouting, age, existing contracts, existing backups, etc. It's also looking at position fit and individual attributes, not necessarily OVR. The coaches could have had a higher grade on that QB than some of the other players in the first round.

[NBryant]

Thanks Brooks. I understand your feedback but would then also ask if there would/may be away for the coaches to also grade the current roster by position prior to the draft and if it fails a "value" then there is a draft roster bounce that then identifies potential improvements deemed a "need". So in the example above a "roster check" by the staff value(s) would deem the QB position as complete because of the current position value AND contract and move on to identify a true need and weakness. As is, again in this example, a QB was drafted and is either ultimately retained and given an extended contract with the other being released and dependent upon your financial setting you incur a financial penalty or the AI either retians this QB with a corresponding contract based on his ratings or may cut him thereby having wasted a draft selection.

Look, I hope I am not coming across the wrong way. I can't do what you can do. I can repair helicopters systems and powerplants and did take a Fortran and Cobol class a million years ago as a kid but can't remember any of it. I also know its basically you and your time is likely split with other demands. Add to that trying to appease what can be a demanding communty of enthusiastic members each with a wish or want. You have an amazing product here but parts of it still feel incomplete, from an outsiders view. With that same view I see what can be an amazing future and pray that it becomes a reality with an end game being that the NFL reaches out to Wolverine and wants to recognize this product because of its representation of their product. Hey, one can dream.

[brooks_piggott]

It already kind of works that way. It weights positional value, needs, salary, existing roster makeup, coach grades, personality, scouting, etc. to come up with a draft value. In this case it chose the QB upgrade. I'd need a lot more information and several different decisions across different coaches, different draft, different teams, different league files, etc. to determine whether there is an issue and if so how to resolve it. In my testing the AI for the draft logic is reasonable within the parameters I've given it, but if you happen to notice a pattern where the logic breaks down consistently then I'm happy to tweak things.

[Gabe778]

Pigybacking off this I'd like to see more logic when it comes to positions that are picked, I think in the 50 season I ran a sim on the position picked most often first overall was CB (almost 30% of the time which is quite alarming). Would like to see some logic/value be added to guys like QBs in QB heavy drafts and pass rushers and such as CB 1st overall that many times seemed insane (esp since looking at the players most were pretty bad)

[AKH]

It is really an two-part issue.

First: Part of the draft logic is built to make competitive teams in the context of the game (This is especially important for multiplayer leagues, since otherwise CPU teams will fall behind and tough to take over for new owners). In the context of building a winning roster getting elite defensive backs [EDIT: that might look bad overall, but I am willing to bet they are generally speedy] is often preferable (in the short term) to getting a slightly better QB (even if he is younger). So, these decisions are partly driven by trying to get as good as roster as possible - mostly as a service to the multiplayer leagues.

Second: the CPU is very conservative trading wise, so you dont see the QB needy teams selling the house to get to the top picks (where they would generally select a QB over a DB)