Spring RTS Engine

Regardless of significance of some skill-value what else is the point of a ladder? People who play in a ladder try to be at top of that ladder.

My skill isn't that good that i would be on top, but i dont care. But imho a ladder should represent a roughly value of the skill of a player at a certain position. I don't say it's a perfect indicator of skill, but i think a ladder should try to rate skill. At least all ladders i ever participated worked by that principle.

Essentially, I don't want it to be used as a tool of elitism, so I'm making it very clear that the rating system cannot be treated as the be all and end all.

Why would anyone even play a ladder if it wasnt a tool of elitism

You can still play non-ladder games, what is the problem?

Ishach wrote:Why would anyone even play a ladder if it wasnt a tool of elitism

I R BETTAR THEN U!

That is what competition is all about.

Missing the point, lads, missing the point. I may be addressing this very topic in a lecture next year, I'll get it recorded if I do.

neddiedrow wrote:Essentially, I don't want it to be used as a tool of elitism, so I'm making it very clear that the rating system cannot be treated as the be all and end all.

I don't think anyone's invented a ladder in which everyone is equal, for obvious reasons...

i aim for the bottom.

proposal:

make two different ladders!

A monthly reseted ladder and one that remains.

We could create awards, for example the "Addicted Springkid" for the most ingametime (of the month) or the "Porcupine" for the most battles fought on metalmaps.

Its just a question of creativity.

For the overall ladder we could do it this way:

(ingametime (in min)/10) + ((1.5*wins)-losses) + (Games com kept alive - Games com died) = Overall Spring Rating Points

The calculations for the monthly ranking just should result in greater numbers.

Needed requirement:
Database that stores all the player data (i presume that only ingametime is recorded atm )

But there should be the possibility to be not mentioned in the rankings.

DISCUSS!

LordMatt wrote:

neddiedrow wrote:Essentially, I don't want it to be used as a tool of elitism, so I'm making it very clear that the rating system cannot be treated as the be all and end all.

I don't think anyone's invented a ladder in which everyone is equal, for obvious reasons...

Not my point at all.

It is very difficult explaining anything to this community. I'm just going to drop it. As Det said, I place too much in relationships - but really, that is humanity - if something negative happens, I will raise the issue, but I feel akin to the one eyed man among the blind here, at least in understanding people, social relationships, and social constructs.

Caradhras wrote:...(Games com kept alive - Games com died)...

No. Bad.

The ranking should NOT give more points to certain strategies over others.

Kill-Death ratio for commanders

Lindir The Green wrote:

Caradhras wrote:...(Games com kept alive - Games com died)...

No. Bad.

The ranking should NOT give more points to certain strategies over others.

Yep... but I like the initial idea of Caradhras

Snipawolf wrote:Kill-Death ratio for commanders

Good idea, but what about mods with coms who cannot fight?

Ok, what about this one:


(ingametime (in min)/10) + ((1.5*wins)-losses) + 0.1*(Games com kept alive - Games com died) + (~destroyed metal value (tru com) / (com death* com metal value)) = Overall Spring Rating Points

BTW, if you disagree, dont just say "that suckz", be productive and show a better alternative!

Caradhras wrote:BTW, if you disagree, dont just say "that suckz", be productive and show a better alternative!

Well, an ELO system would I think be the best for a competitive ladder, but for general competence we could just keep track of total # of wins with each seperate mod.

Obviously this has been discussed before, but the advantage of a ranking system for people who are not at the top of the ladder is not wasting time getting hammered by someone far better than they are when that isn't what they are looking for. I played OTA on the TEN matchmaking network, where I generally had a 1400 rating or so (the ratings started at 1000, and the best players had a rating just over 1600). If I played someone with a rating within 50 points of my own, we generally had a fun game. I went much outside that range, either I walked over them or they walked over me.

The only people inconvenienced are those who are at the very top of the rankings, few people want to match up against them (because they know they won't have much of a chance) so they were constantly having to start new accounts, play until their rating was too high again, and start over.

It can cause some other problems, it probably encouraged rush-friendly maps because those games would be over in 10-15 minutes and help boost the ratings for people in the 1500+ range who could only gain a point or two per match, and there were problems with people deliberately disconnecting to avoid a hit to their W/L ratio. But it's better for the community overall than just random matchups, more people have a better chance to get better games.

--Dave

@Dave, or anyone: Do you know how this rating was calculated?

It seems, that the system functioned.

But in my proposal a win has 50% more effect than a loss, while the w/l-ratio has a pretty little effect all in all.

Btw i would like to read more productive posts and i would be very happy, if someone would more go into my proposal.
Or at least make your own calculation and tell why it is better.

I tired to post a link to the Wikipedia article on the Elo Rating System, which I believe the TEN ladder was based on, but I set off the spam alarms on the forum. So you'll have to look it up manually.

--Dave

very interesting, but i would need an alteration, because of the non-existing limtitation to 1v1 games.

option: Summarize and allocate the rating points dependant on player number.

option 2: Summarize and allocate rating points dependant on player experience (noobs get more)