Quote: Originally Posted By zhawks on 1/20/2010

Quote: Originally posted by colonels19 on 1/20/2010 That's unfortunate.

What is more unfortunate is that you continue to troll our forums

Quote: Originally Posted By seble on 1/20/2010I'm not going to get into an argument with colonels as that's like talking to a wall, but to set the facts straight: Why do you bother responding to him. He's proven time and again that he is to stubborn to listen to anyone's perspective besides his own; once he has decided on a position he is the unmovable object; and his starting point for logic is skewed off center. He's also convinced himself that he is some type of genius who understands things others don't and has the debate skills of the intellectual elite. The former true when the standard deviation is kindergarten and the latter in bizaro World. Obviously none of these self delusions are recognized by the HD community and we’ve (most of us, some still for whatever reason get dragged into his nonsense) simply decided to ignore him. You have no need to defend yourself against his nonsensical points of view and it is appreciated that you’ve kept us appraised of things new to HD and are open to listening to alternative perspectives...Simply stated dont waste your time on that clown.

- I stand by the NBA game and especially the engine, because it's one of the most sophisticated of all of our engines

- There is a handful of users there and one in particular who created a mob mentality of negativity

- I was never "gagged" from any forums. That was my choice to avoid these very conversations that waste everyone's time.

- You've latched onto this "extreme randomness" idea without really knowing what that means.

- This is a simulation of a real life game that does include quite a bit of variance in results from game to game

Quote: Originally posted by seble on 1/20/2010Most of the points made already are correct so I won't delve back into that. The original engine used the standard VB RNG for quite a while and then at some point we switched to using a third-party RNG that was supposedly closer to true randomness and had some other useful random functions built into it. From testing we did, there was no noticeable difference between the two in game results, for the reasons given above.

For this release, the engine is being translated to C#, so that is the RNG I'll be using. I'm not switching because of any shortcomings, just because it's easier to use the built-in RNG.

Quote: Originally Posted By metsmax on 1/20/2010
do you mean too random or too highly variable - those are different things - a game can be run on totally random data with high or low variability

I think - from other threads - that your critique is the level of variability - the beta - or the size of the standard deviations - but am I right? what is the statistical nature of your critique?

Quote: Originally Posted By metsmax on 1/20/2010
I still dont understand.

I dont know how to post a graphic, but imagine three curves, three distributions of results.

One is a "normal" distribution, one has exaggerated tails, one has a very bulky middle and thinner tails. Do you mean that you think that the distribution of results from the engine has exaggerated tails - leading to more outliers, more "black swans" as one author might say?

Or do you think that the random number process for WIS is defective? I guess I'm more prone to believe that the RNG is messed up, because I'll accept ANY result with TRUE and/or great randomness.

Whatever your answer, how can you tell the difference between those hypotheses based on the observable data? what makes you think it is the RNG and not that the distro is set for more variability, bigger tails, than you would like? I could really care less about the distribution, it doesn't matter to me where the majority of the results fall as long as the randomness is good.

AND other than some anecdotal events, what evidence do you have in the observable data for the idea that there are too many outliers, too many black swans? too many compared to real life? compared to what in your opinion makes a good game? I guess I don't have any, but that's all I have to go on is feel because I'll never see the inner workings of their engine/rng. I'm not the only guy griping about randomness, most vocal/belligerent, ok I'll give you that, but I'm not the only one here. The fact that I've seen it in other games and others have sided with my claims of extreme randomness, leads me to believe that something is wrong. Its like if you get some stinky sushi at a restaurant...you don't want to eat it because it smells funny, if seble's your waiter, he's assuring you that its good enough to eat, you won't get sick, etc, etc...are you going to eat it or not? I don't know what happened to the sushi, where its been, how old it is, whose dirty hands touched it, but I don't necessarily need to know that to think/"know"/presume that something is wrong.

All any of us have to go on here is the "eye test" and the feel for text based sports games which I've had a considerable amount of experience with both. Couple that with WIS' "damage controlling" CS staff, why should I believe that the randomness is good when they're just trying to appease me and keep me playing? WIS' involvement isn't totally kosher because they have a vested interest (money) in whether I, you, or anyone else plays or not.

lets get specific about the critique

Quote: Originally Posted By gin_caesar on 1/20/2010
REPORT CARD: COLONELS ATTACKS RANDOMNESS, ROUND 4

Grade - C-

Comments - You need to be focusing on the abstract idea randomness from the end-user perspective instead of rng methods. I'm proud of you for not posting that link again, but I'm still deducting a point because you used the word "clustering." Still, you're showing improvement. You even got a few responses from seble!

Quote: Originally Posted By metsmax on 1/20/2010but basically then you are saying that you think the randomness is broken merely because your gut tells you that it is - hard to take that for muc

Quote: Originally Posted By metsmax on 1/20/2010but basically then you are saying that you think the randomness is broken merely because your gut tells you that it is - hard to take that for muc

Quote: Originally Posted By metsmax on 1/20/2010
I read folks who are peeved at a particular game or result - and I am ready to believe that the engine has some defects - most likely in some sorts of extreme situations that were not fully thought out, but the leap from peeved at a result to indicting the RNG is big for me

yup, going, say, 0-15 from the FT line should not happen - does it ever happen in a season of real life? dont know......might it happen in a simulation - yes - would it take a very sophisticated model to keep it from happening, yes.....heck if I know

but peeved at weird result does not tell me that the RNG is broke