@Pushover: ratio of armies controlled averaged over all turns of a game may be a good one? If a game stays close for a long time, the ratio is closer to 1. This obviously takes into account the number of armies killed by the opponent, but also (efficient) expansion and cards.

Games are won for 2 reasons: (1) having more income than the opponent (2) having more armies on the board than the opponent.
Both of course influence each other.
Killing more armies of your opponent does give you an immediate advantage in (2), but not necessarily in (1), which in turn could lead to a disadvantage in (2).

I agree that armies controlled is quite a bit better. Taking the average over all turns is more robust than any particular turn, but will be annoying to calculate. But I suppose such is the price of league admin -_-

From the example game:


I won't be able to spend much more attention on this until Wednesday night. But I am leaning towards Math Wolf's suggestion.

I won that game, not Oh Noes :P

I'm interested to join as the fourth new player

Beren is right, looks like kill rate is back on the cards.

ratio of armies controlled averaged over all turns of a game may be a good one?

I don't like this one either - a player can have better ratio despite losing, especially in close low-income games. I dug up some examples:
https://www.warlight.net/MultiPlayer?GameID=4257446
https://www.warlight.net/MultiPlayer?GameID=4333886
https://www.warlight.net/MultiPlayer?GameID=5183340
https://www.warlight.net/MultiPlayer?GameID=5628414

The only fair way to break ties is to make tiebreaker tournaments as long as needed.

GROUP E
Wenyun (from F)

Holy crap. I haven't promoted for a long time.

When I won group H I was so happy I'd got rid of buns for a while... This happiness is now gone, and I am empty inside.

So promotion from K4 goes to J or to K3 ? If it goes to K3 couldnt you use something else than random to arrange those groups ? Like order of registration in the league ? Anyway nevermind, it will be a good practice on WR and if anything goes wrong, I cant go down lol.

Just, how did MOD join group C from E for example ? Or I'm missing something.

Edited 4/26/2016 22:05:54

I think you're confused by the fact last season had 4 bottom groups (K1-K4) and this season has 4 bottom groups (K1-K4) aswell. Blame the dropouts :)

bassoon player added to group K4. all other new players will be waitlisted.

@timon92: I agree that tiebreaker tournaments are the best tiebreaker, but I will not grossly postpone the entire league for a single tiebreaker tournament. Therefore, to decide promotions and relegations under extreme circumstances, something ought to be used as a last resort. Under Chokecherry, average turn speed was used. I think we can so something a little better than that. Keep in mind that tiebreaker tournaments are the first recourse of a deadlocked tie!

The example games you provide show close, come from behind victories. The idea of the tiebreaker is that close victories count less than blowouts. I will calculate the numbers later to get an idea of what we are dealing with. But perhaps using army ratios of the final turn only would be good enough? Again, I'm trying to come up with a tiebreaker that will be marginally good enough to break ties for promotion and relegation purposes, as a backup to proper tie breaking, to avoid the situations that occurred under Frank's administration.

@Mike, all the K groups last season and this season are filled with newbies and therefore have no rank vs each other. Therefore I have to treat them as equally as possible, and I have to heuristically decide who promotes where. Super-promotions were heavy due to dropouts. If you are a good player, you will eventually filter up the groups. The alternative to having newbie groups of equal rank is to either arbitrarily rank players who have not played this league before (which is unfair to players who are not known to me) or tell 3/4 of the new players that they can't play this season.

As to super promotions, there are currently eleven tiers to the league. Eleven!! So the super promotions are designed to allow good players at the bottom to more quickly filter up the ranks to fill spots left open by vacating players if they do well enough. A player dropped from the league in group B opened up an extra slot in group C. The top two players from D are promoted. The third slot should either go to 3rd place D or 1st place E. Frank prefers the following sort of order: 1st D, 2nd D, 1st E, 3rd D, 2nd E. In cases where there are LOTS of dropouts, I try to make sure that either the bottom player in a group relegates or at least doesn't promote. Then I try to prioritize promotions for winning records. But in the end the slots need to get filled. This is why everyone in the lower groups got promoted so much.

The idea of the tiebreaker is that close victories count less than blowouts.

I agree. However, the problem with proposed tiebreaker is that quite often a loss is worth more than a win.

But perhaps using army ratios of the final turn only would be good enough?

It certainly would be much better, maybe somehow combined with income ratio?

You could simply take the ratio of Fizzer's formula to determine seasonal winners.
That is: armies on board + 2*income (see https://www.warlight.net/wiki/Seasonal_ladder ).

This would also promote not dragging out games too much which is always a good thing, I'd say.

Another benefit to Math Wolf's formula is that it is easier to calculate. You don't have to mess around with the stats window. Reducing the overhead for the organiser is definitely a good thing.

Im not too crazy about the super promotion. In your example, the guye from E should go in D and an extra guy from E should go to D to fill the room left by an extra guy from C replacing the dropout from B. Because there s no guarantee top player of E is better than everyone in D and his super rpomotion to C could therefore be unfair. Now you are in charge so you can change rules that make more sense :-)). But nvm.

Edited 4/27/2016 16:31:00

Well, i can't see myself in the tournament even though you told me i was in pushover?

@Rouxburg, how did you contact Pushover?

I checked the players for this season. You didn`t participate last season, nor wrote in the season 18 thread that you wanted to join. (https://www.warlight.net/Forum/130085-promotionrelegation-league-season-18?Offset=180)

If you contacted Pushover by any other means (PM, Skype, game chat) then I couldn`t have known. If that is the case, then my apologies. We (and when I say we, I mean Pushover) will see what we can do.

EDIT: I think I know what when wrong. You have signed up to WSOW (https://www.warlight.net/Forum/147689-wsow-season-3-rules-call-participants?Offset=60). Pushover has confirmed your participation there, and you`ll be in once it starts.

This thread on the other hand is for the Promotion/Relegation League. It would seem you have to two confused (they are both run by Pushover)

Edited 4/27/2016 19:38:03

Maybe he is getting mixed up with the other thing pushover is now running, wsow?

Buns posted as I was editing. Seems most likely indeed ^^

@Min34, ah thanks for going over all of that. i thought they were the same thing because of pushover.