Elim 2.0 Discussion

[milosh]

Here's the results of the formula I am proposing:

Scenario:

Player 1
------------
Rating: 1280
Rec: 12-10
R: +6

Player 2 ==WINNER==
------------
Rating: 960
Rec: 14-9
R: +32

Player 3
------------
Rating: 1110
Rec: 7-10
R: +/-0

Player 4
------------
Rating: 860
Rec: 5-10
R: -3

Player 5
------------
Rating: 900
Rec: 11-10
R: +16

Average Rating: 1022
Num Players: 5

As you can see for this formula the zero mark is at a 70% win ratio. I can take that up to 80, 90, or even 100 if desirable though. I think 80 would make a happy median.

[MetalKid]

Might be nice to have the bot PM you when you have 2 deaths left, so you can start playing smarter. You shouldn't have to say 1 because if you say 2 and you die, you'll know you are at one now.

Also, I agree that there should be a new formula. In this case, new = better.

[milosh]

The statistics for the elimbeta bots have been added to our home website. These statistics are simply for beta testing and do not reflect any permanent information until the final release of Elim 2.0. You can get to them easily with http://www.trenchwars.org/elim or by clicking on any of the elim links on the home page. I've designed the stats display so that you can organize statistics by any field you'd like(to see who has the most wins, kills, etc.). There are also navigation controls to page forward and backward as well as jumping to the end and beginning.

EDIT: I added in a player profile page with a search option. It can search for partial names as well.

[Lofty]

1:Foreign> old rating system was everyone starts out at 300 and your rating is k/d*average rating of pilots you killed

I think I remember this as well to an extent

It seems most of you want the rating calculation to be the same as it used to be. That's fine but does anyone know exactly what that was?

NO. so this would be an opportune time to make and shape things to be different. DIF. FER. ENT. So screw them.

Also, in order to prevent specing for rec, if a player has a 2:1 ratio or better without finishing the game, that game shouldn't count for them.

Better than 2:1 ratio actual sounds like a good idea:
Can't spec at all with zero deaths without getting penalized unless of course you've got zero wins. Then you got 2-1, 4-2, 6-3, 8-4, 10-5. None of those recs sound any worth to be speccing for if there is a penalty in store. 9-5 isn't that great a rec to be speccing for rec and since the 2:1 ratio is so specific (1-1, 3-2, 5-3, 7-4, 9-5, etc, and worse) it'll be difficult to even try to obsessively obtain something like that every round.

Oh and screw the old ranking system. OUT WITH THE OLD IN WITH THE NEEEEEEEW SHIT! SHIT! SHIT! SHIT SHIT SHIT SHIT! SHIT SHIT!

[Lofty]

To be honest if I remember correctly I think they used pretty much the ranking formula Blizzard uses for battle.net for StarCraft etc.

I think there was a formula given on the forums at some point...will look for it

[milosh]

It seems most of you want the rating calculation to be the same as it used to be. That's fine but does anyone know exactly what that was?

Lofty, I'm sorry I haven't replied to your post. I've been busy lately and I'm being spread pretty thin.

[Zeebu]

For rankings I think games should only count if there are 5 or more players in at the start of the game. Also, in order to prevent specing for rec, if a player has a 2:1 ratio or better without finishing the game, that game shouldn't count for them. Other than that I think the old ranking/rating system should stay the same.

i like that

[Static Burn]

For rankings I think games should only count if there are 5 or more players in at the start of the game. Also, in order to prevent specing for rec, if a player has a 2:1 ratio or better without finishing the game, that game shouldn't count for them. Other than that I think the old ranking/rating system should stay the same.

[Lofty]

Milosh, you never gave any feedback about this - http://forums.trenchwars.org/showthread.php?t=35072

[MetalKid]

Ah, true. The wonders of Math.

[milosh]

To simplify that...

Just take your KILLS divided by your PERCENTAGE and multiply by 100 and that will tell you how many shots you took.

[DankNuggets]

When showing the % of hits, could you also show the total shots fired? I know that my % would be very low because I often stray at people. Thus, my % will be lower, but the number of times I fired should be a lot higher than average, too.

This might take a bit of thinking but you could just figure out how many "hits" you got by counting your kills.

(hits)/(hits+misses)=%

(hits) = (%)*(hits)+(%)*(misses)

(hits)-(%)*hits = (%)*(misses)

[(hits)/(%)]-hits = misses

So take your kills per match, divide it by the percentage, then subtract the kills from that to get your misses.

[MetalKid]

When showing the % of hits, could you also show the total shots fired? I know that my % would be very low because I often stray at people. Thus, my % will be lower, but the number of times I fired should be a lot higher than average, too.

[RednaZ]

If I understand it correctly that this new elim will also be facilitating casual javs play, I strongly suggest you remove every single bot from ?go javs. Shit killed base-elim with the streaks and barwarp and other non-sense.

quoting this till it happens