Simming Games
Sat May 13, 2006 11:47 am
How does Live 2000 determine the outcomes of simmed games in franchise mode? Is it the ratings, and do the statistical results even have a bearing on future performance? Also, are the stats generated independently of the actual game results?
My questions stem from a franchise I started using a recent roster update. I wanted to see if I could gain a comparative advantage over the AI GM's by inputting an estimated Player Efficiency Rating as the "custom" stat, and then basing all my personnel decisions on that stat. In other words, since the CPU evaluates trades by Overall Ratings, I would trade guys with high ratings but low PER's for guys with low ratings but high PER's, repeating ad infinitum. As Dallas, I won the first four titles of the franchise mode that way... but just as quickly, my team went to lottery-bound hell. Yet in terms of PER, I had a bunch of guys with good numbers, and our point differential was not commensurate with how bad a record we had -- year after year. Are simmed games based solely on comparing the two teams' ratings? Because if so, my methodology is inherently flawed, and no wonder it backfired so badly. Yet the individual stats seemed pretty consistent with ability -- if I traded for a high-PER guy, he'd continue to post a high PER for my team. Do the individual stats have any connection to results, or does the sim engine simply compare overall ratings and spit out the game score (and the stats) later?
My questions stem from a franchise I started using a recent roster update. I wanted to see if I could gain a comparative advantage over the AI GM's by inputting an estimated Player Efficiency Rating as the "custom" stat, and then basing all my personnel decisions on that stat. In other words, since the CPU evaluates trades by Overall Ratings, I would trade guys with high ratings but low PER's for guys with low ratings but high PER's, repeating ad infinitum. As Dallas, I won the first four titles of the franchise mode that way... but just as quickly, my team went to lottery-bound hell. Yet in terms of PER, I had a bunch of guys with good numbers, and our point differential was not commensurate with how bad a record we had -- year after year. Are simmed games based solely on comparing the two teams' ratings? Because if so, my methodology is inherently flawed, and no wonder it backfired so badly. Yet the individual stats seemed pretty consistent with ability -- if I traded for a high-PER guy, he'd continue to post a high PER for my team. Do the individual stats have any connection to results, or does the sim engine simply compare overall ratings and spit out the game score (and the stats) later?
Sun May 14, 2006 4:26 am
Hi
So many questions in your post, nevertheless you raise many interesting topics!
I'm too tired to make a constructive answer, so here are some random facts:
1/ like in every other versions of NBA Live, overall ratings are pretty much fucked up. They can give you an indication if the player is a superstar, a starter or a role player, but don't give it too much importance. In general, Live 2000 tends to overate PF and C, while SG and PG are underated
2/ you're talking of a "recent roster update", are you talking of mine? If yes, well, overall ratings are fucked up. I have to update ratings with an external editor because if you update ratings in the game, it will automatically fuck up the salary value. So with an external editor, overall ratings are no longer updated... I update it manually, but it does not necesarily reflect what the game would give, all the more as I try to compensate the overated PF/C problem
3/ players stats are also calculated with DSTATS values. Check also this page, you may learn interesting things
So many questions in your post, nevertheless you raise many interesting topics!
I'm too tired to make a constructive answer, so here are some random facts:
1/ like in every other versions of NBA Live, overall ratings are pretty much fucked up. They can give you an indication if the player is a superstar, a starter or a role player, but don't give it too much importance. In general, Live 2000 tends to overate PF and C, while SG and PG are underated
2/ you're talking of a "recent roster update", are you talking of mine? If yes, well, overall ratings are fucked up. I have to update ratings with an external editor because if you update ratings in the game, it will automatically fuck up the salary value. So with an external editor, overall ratings are no longer updated... I update it manually, but it does not necesarily reflect what the game would give, all the more as I try to compensate the overated PF/C problem
3/ players stats are also calculated with DSTATS values. Check also this page, you may learn interesting things
Sun May 14, 2006 5:45 am
Thanks!
Sun May 14, 2006 6:06 am
Wow, I've had this game since the day it came out in 1999, and yet I knew so little about how it worked! Looking at that article, it seems to me that Dstats dictate what stats a player puts up -- and therefore what his PER will be. Too bad these numbers are totally independent of my team's W-L results! Now, if I remove Dstats, causing the numbers to be produced strictly by the ratings... will that make the stats more reliable to base decisions on? Because I was having guys with ratings in the 60's put up All-Star-type PER's (which made me happy at first because I was paying them peanuts), but when it came time to play games, my team of (statistical) All-Stars would have a .125 winning %. Now I realize that the stats are generated independently of actual results... silly me for actually expecting players with All-Star-levels of production to actually contribute to wins! Thanks for the link, maybe eliminating Dstats will help.
Oh, and I say "recent roster update" because, sadly, I can't remember when or where I got it. Although they seem to be the rosters heading into 2005-06: Diop is on Dallas, as is Adrian Griffin.
Oh, and I say "recent roster update" because, sadly, I can't remember when or where I got it. Although they seem to be the rosters heading into 2005-06: Diop is on Dallas, as is Adrian Griffin.