ELO Ranking (Full Version)

All Forums >> [New Releases from Matrix Games] >> Strategic Command Series >> Strategic Command WWII: World at War >> Multiplayer Sessions



Message


slickandjake -> ELO Ranking (3/14/2019 6:50:09 PM)

I wish to start an ELO ranking for this game. Here will be the basic formula modeled and slightly modified from the World Football ELO ratings formula:

Rn = Ro + K(W - We)
where
Rn = new rating
Ro = old rating
K is a constant, and I will use 4 for it
W = result of the match; I will use 5.5 as a decisive victory, 4.5 as a major victory, 4 as a minor victory, and 3.8 as an Axis Tactical victory (for 1939 campaign game)
We = expected result of the match, where We = 3.7/[10^(-DR/400)+1]
DR = Difference in ratings

Each new player will start with a rating of 1000. I will keep a rating as Axis and as Allied for each player, and also will take the average of each as a third rating.

As an example, if a player with an Axis rating of 1020 plays a player with an Allied rating of 990 and the Axis win a major victory, the player's Axis rating will increase to 1030 and the player's Allied rating will drop to 980. Conversely if the Allies player wins a major victory the Allies score becomes 1001 and the Axis score 1009.

Just send me the two players, who played which side, the result of the match, and I will keep the ratings updated. If someone quits or resigns, I will give the winning player a decisive victory and the losing player a decisive loss. I will also count the number of quits/resigns and if the number becomes excessive I will cease to allow a match with that player to impact ratings. This will be my discretion, but I am thinking three quits in some time period is unacceptable and grounds for disallowing games with them.

After a while as the ratings get sorted players can then determine equally matched opponents for better gameplay.

For different game campaigns, I can do the same if you play something other than 1939 campaign.

Edited on March 15, 2019 since the calculation did not penalize a player playing a lower ranked player very much. I updated the numerator in "We" from 1 to 3.7.




ElvisJJonesRambo -> RE: ELO Ranking (3/15/2019 12:28:53 AM)

That's cool of you to consider ranking/rating system. Personally, I don't feel there's anything wrong with resigning. When a game gets obvious that it's over, completely 1-sided, is there a point of continuing? Loss is a loss. I'd rather say,"Hey, good game you got me" and play a new one. Also, there's probably free online Ladder System that might to this work for you, if links are allowed to point to it.




slickandjake -> RE: ELO Ranking (3/15/2019 4:23:15 AM)

If it is a mutual agreement on resignation I have no problem with it. But if it is someone who just stops playing without contact, well that is different and I want to avoid playing players like that. I also had an opponent just resign when I was the Axis and I just took France well before Russia or the US entered the war. While things weren’t going well for him, that is way too early to just up and quit and without any contact too. I will never play them again.




ElvisJJonesRambo -> RE: ELO Ranking (3/15/2019 4:41:24 AM)

@slickandjake --- I hear you brother. You're bringing up the "Dodge Ball Dude", person who is on the ropes but won't take their beating. If you want to throw in the towel before the beating, so be it. But don't make me come find you. I'll admit it, I hate to lose. I've busted keyboards, and even a laptop (that was over Texas Holdem Poker). I can "feel a player" through the replays. Can sense frustration. That's what makes the game great. Can you get over a screw up or some bad luck? I get a bitterness, ever human does. You're gonna make mistakes and lose at times. That's the mental challenge of SC.

Far as your Dude went skirt after France, that's a pure loser. France is a lost cause. I get players who snap all the time. Weak minded souls. Ultimately that's between them and their Maker.





slickandjake -> RE: ELO Ranking (3/15/2019 1:02:41 PM)

Wow, talk about passion! I don't like to lose either, but not to the point I want to bust something! [:)]

My opponent was in bad shape. I think I destroyed his whole UK air force, he only slipped his HQ back to the UK from France, I had destroyed some UK ships including 2 carriers I think, I picked off a few of his units in North Africa and he was losing there, Italy had declared war on Greece and was about to take Athens, and I was close to already taking Lanchow. It was an ugly start for him.

I am not making excuses for him, he should have at least notified me he was quitting, but I hadn't even started to invade the UK due to bad weather and the war against Russia hadn't even started. Things could have change.




Tidavis -> RE: ELO Ranking (3/22/2019 1:39:18 AM)

You can put me down as zero wins and six losses in multi player. An aggregate score of 0.00 Yes I have quit in frustration. I'll give you an example, when it takes a medium bomber, two tac bombers, 3 infantry corps, and two attacks by two panzer corps, to dig out a special forces unit that just landed,
then go on to have a HQ cause 5 points in losses to two other panzer corps to 1 lousy point against said HQ, I guess frustraion might be a bit understandable.


Oh well. To each thier own. Don't need any more lesson from the experts.


iiIii




slickandjake -> RE: ELO Ranking (3/22/2019 7:01:26 PM)

Tidavis, the whole point of the ELO system is that you should be able to identify and play players at the same level as you, and know which players are experts that will give you a beating. But the information you gave me is not enough for me to give you an ELO ranking.




BillRunacre -> RE: ELO Ranking (3/22/2019 9:52:34 PM)


quote:

ORIGINAL: Tidavis

...then go on to have a HQ cause 5 points in losses to two other panzer corps to 1 lousy point against said HQ, I guess frustraion might be a bit understandable.



Hi Tidavis

This combat result sounds very strange, and certainly not within the realms of generally expected combat results.

I can only assume that the Tank units were very weak/lacking supply/without HQ support AND the enemy HQ was entrenched/behind a river/in defensible terrain?

Assuming of course that no amendments were made to the campaign before play began?

Bill




Tidavis -> RE: ELO Ranking (3/23/2019 5:21:46 AM)

This was the second turn of the Invasion of France in 1940. Supply was not a problem. I recall several predictives showing 2-0 results ending up 0-1 or 0-2.

Just way too many odd things happen in my games to make it enjoyable.


This was my last game in my multi player. Surely you have a way to access these files?




slickandjake -> RE: ELO Ranking (3/23/2019 6:08:23 AM)

If those were the predicted results then the actual results should be impossible since the worst you could have is 1-1. So if that was truly the case then that appears to be a bug. I have never seen that before at all.

I am playing a game as Axis and I scouted into the UK and found no units in an area, so I dropped a parachute unit and was going to try to take a port only for there to not only be a carrier interceptor which didn’t attack my scout plane like it should which damaged my parachute unit, but when it landed suddenly there are units everywhere where I scouted. Therefore my invasion was thwarted right away and drew the US and Russia closer to the Allies in the process. I have never seen a plane scouting on land come out to be wrong, and it was the most critical moment that it failed!




4100xpb -> RE: ELO Ranking (3/29/2019 5:16:37 PM)


quote:

ORIGINAL: slickandjake

If those were the predicted results then the actual results should be impossible since the worst you could have is 1-1. So if that was truly the case then that appears to be a bug. I have never seen that before at all.


You never have actual results differ by more than 1 from the predicted results? I know the manual says this, but it has happened so frequently with me I just assumed that it's normal. Are Tidavis and I the only ones?




slickandjake -> RE: ELO Ranking (3/29/2019 5:56:43 PM)

Correct, I have never seen results that deviate more than +/- 1 for each side that I can think of.




Page: [1]

Valid CSS!




Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI
0.90625