CPU Attributes

tzoky07 · Post by **tzoky07** » 02 Dec 2010, 19:51

Hello, i am new here and i would like to ask you guys a question about the CPU attributes, i mean skill is self-explaining but what is assurance, and how does it influcence the CPU jumps?
I am trying to put together this season's jumpers database and need some data about this.

Thanks in advance

mis · Post by **mis** » 02 Dec 2010, 22:38

Correct me if I'm wrong.
If a player has low assurance but high skill, then it is likely this CPU player is able to do superjumps but also bad jumps, a lot more inconsistency.
A CPU player with both good skill and high assurance is more likely to be on top of the table in each competition.

EyTschej · Post by **EyTschej** » 03 Dec 2010, 00:04

I've tried to create a database when the game was released, but most cpu players will have mostly random results unless you give them maximum values, I guess. It's basically a waste of time trying to create a good, realistic database. Since calculation of the cpu jumpers' attempts takes aaaaaages in DSJ3 anyway (was super fast in DSJ2 - never understood why DSJ3 was such a huge step back in that aspect), there was no need to adjust the attribute system, I guess.

Just wait for DSJ4 and see if cpu jumpers and their attributes will be of better use than in DSJ3.

Jani Oravisjärvi · Post by **Jani Oravisjärvi** » 03 Dec 2010, 00:27

I hope that in DSJ4 CPU jumpers with 100% attributes will have better level than in DSJ3. In DSJ3 they do competitive lenghts in back winds but with good winds they can't use winds at all. In flying hills with front wind I often jump 30m longer that best possible CPU-jumpers so it's too easy to take dominant victory in the cup.

I can imagine that if 100% CPU jumpers would compete in online winter cup against real players, they would be quite high in back wind competitions, but would nearly never be in top-30 in front winds.

Post by **Jussi Koskela** » 03 Dec 2010, 16:38

Mis's explanation about the skill & assurance levels was correct.

With regards to CPU players in DSJ4, I might have to leave them completely out from the game (at least in the first versions). Due to more complex physics it would take much longer to simulate the jumps and it would be pretty difficult to make the cpu players jump well enough (without any cheating of course).

Severin · Post by **Severin** » 03 Dec 2010, 22:55

I definitely would support that. Better to have a good game without CPU players than to have a game, where you have CPU jumpers, who perform completely out of bounds.

Anyway I have to say I thought the DSJ addiction would have been gone, but with the bits of info comming in of DSJ4 my right hand's index and middle finger already start to itch

EyTschej · Post by **EyTschej** » 04 Dec 2010, 12:30

@Jussi:
You don't need to simulate the cpu jumpers for real if it takes as long as in DSJ3 (it takes aaaaaaaaaages in DSJ3). Simply randomize their results based upon the wind. I'm thinking of a formula that might look something like this:

K-point * wind (1.xx for headwind, 0.xx for tailwind - xx = wind speed without the comma, e.g. 1.12 for 1.2 m/s headwind) * jumper skill (values from 0.50 to 1.50 for example)

So instead of calculating the jumps, you only do a more schematic and less realistic calculation, but that's much better for single player mode than what went down in DSJ3.

@Severin:
Wenn es in den Fingern kribbeln sollte, dann kannste dich auch immer noch an DSJ3 versuchen - hast in den letzten Tagen eine Menge Punkte auf mich verloren, MUHAHAHAHAHAHA

Nebenbei bemerkt: Wilder 2. Durchgang in Kuopio ... wild, wild, wild ...