Posted by oldresorter on 9/26/2010 7:58:00 PM (view original):
I have won several titles with the new engine - 3 I think - I have not changed anything - just like everything else in this engine, matchups represent a small fraction of success, it is far more important to make sure your best player is playing his best position and is getting touches ... real life would you play john wall at center and give him 45% disto because the other team has a weak center defender, or would you play him at PG and let the other team see what they can do with him? this whole matchup issue slays me, thankfully seble seems to have gotten it right, when he announced it, I thought he was going to make a sham out of the engine, but thankfully he did not.

a bit extreme of an example there, no? i was thinking more like, switching my pf and c, or pg and sg, or maybe a starter and backup. those seem useful to me.

billy yea - purposefully extreme - I'll bite - so if you coached a real NCAA team, you would move your pg to sg because the sg on the other team is a weak defender?  Tell me exactly how many times you think that has happened to john wall or chris paul?

Posted by oldresorter on 9/27/2010 7:31:00 AM (view original):
billy yea - purposefully extreme - I'll bite - so if you coached a real NCAA team, you would move your pg to sg because the sg on the other team is a weak defender?  Tell me exactly how many times you think that has happened to john wall or chris paul?

OR, that's not a fair question or example -- that does not exist in real life. You can't "move" someone somewhere, because it's the defensive team that controls who's guarding who. But it's extremely common to have your top perimeter defender guard the other teams top perimeter scoring threat ... so maybe we should be talking about individual defensive matchups instead? 

Forget about moving a player all around from his "natural" position, personally I'm talking about games where you simply have a player being guarded by a weaker defender, and your ability to exploit that properly on a consistent basis. (Not every game basis, but significantly more than if he was being guarded by a strong defender.)

girt - I'll answer that too - if you simmed 1000 games - the weak defender in your example would give up lots more than a strong one - but this game is founded on random number theory, a bell curve of results if you will - about 16% of the time, you will get some pretty negative results from a matchup you should win, that is a pretty significant number - but,  the moment you guys convince seble to change that, the engine and the game stops producing real life results, and becomes something else.

Also, just like the diff between +5 def and -5, the diff between weak and strong defense is probably not as high as people would like, I am almost sure tarek once told us (maybe me in a ticket), the diff between +5 and -5 on 3pt shooting was only 12% - 28% on avg vs 40% on avg, my guess is the diff between being guarded by a 1 def player and a 100 def player is also not that high, maybe 12% also, maybe 20% but no higher.  The variance in the bell curve is much higher than that real life, hence often you will see some pretty bad games in games where the off has a huge edge, since the shape in the games bell curve is pretty much like the shape of the curve in real life.

Girt I know you know all this and we have debated this many, many times.  But it is interesting, seble seems very willing to do whatever the majority says, yet even he has not listened to what I am almost sure is the majority opinion in this game.  Tarek as you know, was adamant such a change was not in the games best interest. Those two are in a position to best comment on this subject, the rest of us are just guessing based on pieces of facts, pretty typical of most 'hot' topics in this game.

debate on my friends.

OR, no question that there should be a % of the time that a seemingly good matchup doesn't work out. The question at hand is whether the % of the time that happens currently is too high, too low, or just about right. And whlie we can all supply our educated and semi-educated guesses based on what we observe, there's really no way that any of us can know definitively.

I've always done everything in this game on feel, I've never crunched any #'s or tried to figure things out from that perspective. My feel is that it's too common for a matchup that should be exploitable not to work out that way. All of us are working with a sample size that isn't truly large enough to reach an appropriate conclusion, so I'm open to the possibility that I'm off base here. But that's my feel right now.

My perspective on this: one of the reasons I felt I had success early in HD was that I was very good at exploiting defensive weaknesses, not just in M2M matchups but also in finding holes in Zone and attacking the vulnerabilities in the Press. As time went on and the game evolved, the strategies I used to exploit those weaknesses become less and less effective, and nothing new I tried seemed to work either. Partially I know this was due to the top players becoming too similar (i.e. 90/90/90/90/90 syndrome) but it also seemed to me that the results of mismatches I used to feast on with consistency had become too random.

The big factor that influences me coming back to the game was the promise that this aspect of the game would be restored. And while I wouldn't say my ability to attack those mismatches and weaknesses is what it was in my National ******* Champion of Naismith days, it's definitely a hell of a lot better than it was when I quit.