MizzouRah
07-24-2004, 12:31 AM
ESPN NFL 2K5 Trade Issue Analysis
There have been multiple reports of high-volume, poor trading by the AI in 2K5. I did some investigation into this issue, mostly because I'll be weeping uncontrollably if I find a gamekiller.
Why? because this game is MAGNIFICENT.
I'll tell you why next week, but please remember in the meantime that I absolutely pounded ESPN last year for abysmal 2-minute A.I. Madden has clearly had far superior gameplay (with proper sliders) for the last two years. This year, though, ESPN is just wonderful. So far.
The most serious issue has been this trading problem. It has two components: one, poor player valuation leading to lopsided trades, and two, high trade frequency, which spotlights the sometimes questionable A.I. even more.
I ran quite a few tests today, and I believe I've identified causes and corrective action. If you start a franchise with both pre-season games and weekly prep turned on, you will see 10-11 trades during the course of the season. If you turn one feature on and leave the other off, you'll see about 6-7 trades during the season. However, if you turn both off, the number of trades during the season will fall to only 2-3. The numbers are remarkably consistent each year.
There will still be offseason trades (5-6, usually), but with the greatly reduced in-season frequency, you'll only see one or two deals a year that look suspect. I can easily live with that.
From what I can determine, the suspect trades involve uniformly valuing players by their ranking regardless of position. As an example, you will occasionally see a position player (a defensive end, for example) traded for a quarterback. In the NFL, this never happens, because a quarterback has unique value due to his position. The game A.I., though, apparently only gives a small bonus to the quarterback, so an 86 rated quarterback traded for a 91 rated defensive end looks like a reasonable deal. In actuality, the quarterback position is so difficult to value versus other positions that quarterbacks are almost always traded for draft picks.
I've also occasionally seen a kicker traded for a position player. Again, same difficulty with valuation due to the uniqueness of the kicking position. Kickers are rarely traded, and when they are, it's always for a draft pick.
Now if you see five or six of these deals a year, it might drive you crazy. It certainly would me. But seeing one a year? No problem. Plus I have no problem giving up pre-season games and weekly prep. I didn't use them, anyway.
If you want to retain weekly prep, I believe that you could just leave it turned off until after the trading deadline in Week 8, then turn it back on (remember that you have to also select trading deadline 'on' in the options menu as well for the Week 8 deadline to apply). I don't know if it would have any undesirable side effects, though, so do that at your own risk.
Again, thanks to Bill Harrris!
Todd
There have been multiple reports of high-volume, poor trading by the AI in 2K5. I did some investigation into this issue, mostly because I'll be weeping uncontrollably if I find a gamekiller.
Why? because this game is MAGNIFICENT.
I'll tell you why next week, but please remember in the meantime that I absolutely pounded ESPN last year for abysmal 2-minute A.I. Madden has clearly had far superior gameplay (with proper sliders) for the last two years. This year, though, ESPN is just wonderful. So far.
The most serious issue has been this trading problem. It has two components: one, poor player valuation leading to lopsided trades, and two, high trade frequency, which spotlights the sometimes questionable A.I. even more.
I ran quite a few tests today, and I believe I've identified causes and corrective action. If you start a franchise with both pre-season games and weekly prep turned on, you will see 10-11 trades during the course of the season. If you turn one feature on and leave the other off, you'll see about 6-7 trades during the season. However, if you turn both off, the number of trades during the season will fall to only 2-3. The numbers are remarkably consistent each year.
There will still be offseason trades (5-6, usually), but with the greatly reduced in-season frequency, you'll only see one or two deals a year that look suspect. I can easily live with that.
From what I can determine, the suspect trades involve uniformly valuing players by their ranking regardless of position. As an example, you will occasionally see a position player (a defensive end, for example) traded for a quarterback. In the NFL, this never happens, because a quarterback has unique value due to his position. The game A.I., though, apparently only gives a small bonus to the quarterback, so an 86 rated quarterback traded for a 91 rated defensive end looks like a reasonable deal. In actuality, the quarterback position is so difficult to value versus other positions that quarterbacks are almost always traded for draft picks.
I've also occasionally seen a kicker traded for a position player. Again, same difficulty with valuation due to the uniqueness of the kicking position. Kickers are rarely traded, and when they are, it's always for a draft pick.
Now if you see five or six of these deals a year, it might drive you crazy. It certainly would me. But seeing one a year? No problem. Plus I have no problem giving up pre-season games and weekly prep. I didn't use them, anyway.
If you want to retain weekly prep, I believe that you could just leave it turned off until after the trading deadline in Week 8, then turn it back on (remember that you have to also select trading deadline 'on' in the options menu as well for the Week 8 deadline to apply). I don't know if it would have any undesirable side effects, though, so do that at your own risk.
Again, thanks to Bill Harrris!
Todd