Subject: Questions & Answers

Well, "that's life" is an acceptable response to a team with multiple GREAT youth pulls, but for the rest of us, it's not.

One question to Vlady:

Is it possible that in game engine, in same match, same action happens twice ?
If yes, the result of action will be the same ?

It is possible, but the result might not be the same. If Marian Gaborik has 2 breakaways in the same game, it is possible that he will score on the first one and he won't on the second one.

I didnt say there are no random figures entering the calculation. I said, there are no random results. Every game result is a sum of calculations. these calculations must use random figures to determine their outcome.

Let's say you have a scoring chance and you have 10% chance of scoring a goal based on the attributes. A random number is generated (lets say from 1 to 1000). If that number is from 1 to 100, you score, if it is 101 to 1000, you don't.

Ok, so the % that you might score does in fact increase if you have better attributes?

....
show me one team that has had results with 50% of upsets over a long period of time. You cant distribute upsets evenly because every game is a separate entity. there is no record who had how many upsets in his games.
....

why 50%, is that your internal number to decide that something is wrong?

You even find 20%, but that doesn't prove everything is OK...
And there should be record how many games are with unexpected result...then you'll either shut our mouths or see that there is something wrong

....
same here. every pull is a separate calculation. it is impossible to ensure the same amount of "luck" for everyone. that's life.
....

Yes, it is possible. There is huge range on each level...that's reason why there are so many differences on same level...Just make that range smaller, you'll ensure "luck" or "disappointment" for everyone

of course

"why 50%, is that your internal number to decide that something is wrong?"

because you said "but why are there teams that never have these "upsets" and "surprises" and teams with them every second game?"

"And there should be record how many games are with unexpected result...then you'll either shut our mouths or see that there is something wrong"

We can simulate any game hundreds of times, we have the tools to judge if the percentage of upsets is within normal or not.

"Just make that range smaller, you'll ensure "luck" or "disappointment" for everyone"

This is a matter of opinion. What range is small enough?

Thanks, now I get it !

Correct me if I'm wrong:
Every scenario is calculated independently, and based on some parameters, the result of the scenario can have different values (in percentages), but it is still choosed randomly (based on those percentages) !

I don't understand those complaints. Roll a 6-faces dice 10 times, it's possible you get 10 times '1'. But roll it an infinite number of times, you will have almost an equal number of '1', '2', '3', ..., '6'.

Just a matter of basic probabilities and statistics. If you don't believe it, spend 15$ on this book www.amazon.com/Statistics...

It's not really like that !

'Roll a 6-faces dice 10 times' - this happens when both teams are equal.
'Roll a 6-faces dice, with same value on 5-faces, 10 times' - this happens when 1 team is stronger !

yeah, that's grate, count on words not the meaning behind them...and big picture is so nice
...
you have tools, but who program those tools? If you think that something is good, you'll program tools to check that in same way
...
you can't tell which range is small enough, but you can say, this range is good, funny

Your point is?

I'm as outspoken as the next person here, but i can tell you this... the guys at PPM are pretty damn dedicated to making this the best game they can, and if Vlady says they have run tests and found the system is working best, then I believe him and suggest we all do the same.

nowadays... usually... at first, test tools are implemented and only then the code that will be tested... meaning... you dont know what results from your algorithms will be and hence it is quite difficult to influence test results

"but you can say, this range is good" => say...

I don't question their dedication, I'm sure they try to do they best, but that don't mean that everything they done is good and should be that way

I'm programmer myself and I know that programmers never do the checks as good as testers do...
and programmers should never program test tools, because he will program them in same way he programmed the application

"you have tools, but who program those tools? If you think that something is good, you'll program tools to check that in same way"

i'm not sure I understand this. You think we manipulate our own tools to fool ourselves?