PDA

View Full Version : New match making suggestion



Soulseller
07-29-2014, 01:58 PM
The new match making system isn't live yet but Real World said something in another thread that made me think:

The matchmaking won't be done with relics any more. You are placed against a player similar in base strength to yourself though if your relics start to increase too quickly you will start to get stronger bases. Likewise if you start getting your ass handed to you you will get easier bases.

This about that if your relics increase to quickly you get matched against stronger bases is good since we were worried that the new match making would benefit bases with weak DR but strong offense. However, as I understand it this could still be a problem if a player is gaining relics in a steady but not to fast way.

Wouldn't it be better if the DR that's searched for would increase and get closer to the max the closer you get in relics to the top? It could be something similar to this:

DRsearch = DRcurrent * (1 - (Ryou/Rleader)2) + DRmax * (Ryou/Rleader)2

Ryou = Your relic count
Rleader = Relic count of leader of the rankings
DRsearch = Defense rating searched for (+/- some value)
DRcurrent = Your current defense rating
DRmax = Maximal posible defense rating or DR of leader of rankings

The difference between my suggestion and what Real World said is that any player that is very active can still rise fast in rankings and is only getting higher DR rated opponents when he gets close to the top. It also makes sure that the best ranked players have the best bases.

Of course the exponent could be changed if you want a "flatter" or "steeper" function.

What du you think??? :confused:

Soulseller
07-29-2014, 02:04 PM
I can give you some test numbers:

Ryou = 100
Rleader = 4000
DRcurrent = 250
DRmax = 1000

-> DRsearch = 250.5

Ryou = 500
Rleader = 4000
DRcurrent = 500
DRmax = 1000

-> DRsearch = 507.8

Ryou = 3000
Rleader = 4000
DRcurrent = 750
DRmax = 1000

-> DRsearch = 890.6


PS: I have no idea what DR numbers are realistic but thats not important, it's only the proportions that matter.

Eowyn
07-29-2014, 03:37 PM
This makes a lot of sense. I don't fully understand the equation at top, I will admit, but it makes sense.

Phileine
08-01-2014, 08:16 PM
I'm trying to understand what you are trying to say. Let's take your first example: DRcurrent = 250 and DRsearch = 250.5. Does that mean that any opponent with a DR between 0 and 500.5 would be a possible match? Or am I interpreting DRsearch wrong?

Soulseller
08-02-2014, 01:13 AM
I'm trying to understand what you are trying to say. Let's take your first example: DRcurrent = 250 and DRsearch = 250.5. Does that mean that any opponent with a DR between 0 and 500.5 would be a possible match? Or am I interpreting DRsearch wrong?

I mean that DRsearch is the actual DR value searched for. You can think of it as the mid value, i.e. if assuming search range is +- 20 % of DRsearch we get 250.5+-20% -> DRsearch = 200.4 to 300.6
Does that clarify?

Phileine
08-02-2014, 01:49 AM
Ok, currently I perceive the bandwidth that is searched for as a bigger problem than the middle of the search (the 20% in your statement) - I appearantly have such a high DR that there were just no matches for me to be found - but I can see what you are trying to do.

Currently the situation is assymetrical where higher players can attack further down than lower players can attack up. This is caused by using a percentage: 20% of a higher relic number gives a bigger relic bandwidth than 20% of a lower number. Your formula appears to cause to opposite: lower players can attack up more (because the DRsearch is always higher than the DRcurrent unless you have zero relics, than they're equal) than players can attack down. Is this intentional? If so, what are you trying to accomplish with that?

goly
08-02-2014, 02:06 AM
seguimos esperardo la bandera de argentina otra actualizacion y nada

Soulseller
08-02-2014, 02:26 AM
Ok, currently I perceive the bandwidth that is searched for as a bigger problem than the middle of the search (the 20% in your statement) - I appearantly have such a high DR that there were just no matches for me to be found - but I can see what you are trying to do.

Currently the situation is assymetrical where higher players can attack further down than lower players can attack up. This is caused by using a percentage: 20% of a higher relic number gives a bigger relic bandwidth than 20% of a lower number. Your formula appears to cause to opposite: lower players can attack up more (because the DRsearch is always higher than the DRcurrent unless you have zero relics, than they're equal) than players can attack down. Is this intentional? If so, what are you trying to accomplish with that?


My biggest point was that I didn't like how Real World said the new system works: that you get harder bases if you get relics fast. I rather want a system where you can climb fast with no limitations except that you have to have a good base (high DR) as in my system.

Yes it's true my system is assymetrical. It's not that easy to get a symmetrical system when the function is non-linear. Need to think some more then :rolleyes:. But my point wasn't to give an exact function, more to give an idea of what I think would be good. It would need some tweaking and I need to know more about the prerequisites to come up with a more final and ready to use function.