Amaroq
Posts: 1100
Joined: 8/3/2005 From: San Diego, California Status: offline
|
My first reaction, from the title, was "Oh my god no!" - as a game developer myself, I take great pride in the fact that none of my AI's have ever cheated, and they still provide a competitive opponent. That said, your idea definitely has some appeal without feeling like 'cheating' to me. I can certainly see value in having the AI draft 'household names' in the first round, and insist on holding on to them even if they aren't panning out; I follow that strategy myself, with an irrational emotional attachment to somebody just because I know who he was IRL. It could potentially backfire, if the AI weighted 'real life' more highly than 'game development', e.g., if Williams simply isn't developing, and the AI rates him too highly, he's costing the team a roster spot. More worrisome might be impact on the trade AI - if it was willing to give you players who have developed into stars in the game for somebody who was (only) a star in real-life, that would be a competitive disadvantage. That might be solvable by having the AI use your proposal for 'draft', 'release', and 'trade out' decisions, but not for 'trade in', 'free agent', and 'contract renewal' decisions. . . . On the other hand, this does nothing for fantasy-player GM's; maybe a better solution would be to increase the value the AI places, in the draft, on super-young, super-high-potential players. I mean, an 18-year-old with a 90 potential, I don't honestly care what his current stats are, and if my farm system is in good shape, I'm willing to wait six years while he develops through A, AA, and AAA. The AI, on the other hand, seems to look only for major-league and AAA-quality players (appropriate for a 35-man roster, but not doing it much service on a 60-man roster). Maybe the proposed 40-man-roster changes will give the AI the sense that anybody outside of the 40th man should be evaluated *solely* on future potential, weighting 'current ability' only as a step towards determining how good the player is likely to be (so that it can tell the difference between two 18-year-olds of 90 potential, one who is 40-40-40, and the other who is 5-5-5); it should also count the things which don't improve (speed, hands, arm) in the evaluation at full weight as well. Likewise, the AI needs to start signing all of those young, 23-and-under high potential players, to a long-term (5- or 7- year) contracts, or six-year minor-league-salary contracts in the 40-man-roster world.
|