Score calculating in not finished ladder games

[SoulMan]

Hello. I read about score calculating in ladder games, and I guess it's not fair. I think Heroes are underrated there, and cities overrated. I think game should consider cities wall level, income and production as well. Some values should depend on map size as well. Here is my proposition (for next season ofcourse):

Unit = (1 point per tier) divided by sqrt(K/10)

Heroes = (4 + 2 per level + 1/2/3 per item tier 1/2/3 (items score cap to 3 items per hero)) multiplied by sqrt(K/10).

Gold = (1 point per 100 gold) divided by sqrt(K/10)

City = 6 + 1 per 5 defense(wall) + 1 per every 25 income over 15 + 1 per tier of the most expensive in city production

Other buildings = 0.5 + 0.5 per 5 defence (wall) for towers, 1 point per watch tower, 1 point per 20 income from sites.

This points distribution gives better balance for me and considers map influence (the bigger map is, units and gold are less important and heroes are more important). It is estimated for 1v1 games (not sure if it should be different in FFA or Team games, rather not).

I think i feel this game quiet good, if this values are less balanced than actually used I shouldn't be able to play so effective, because my priors are simply bad.

Now its kinda weird that even game with cities distribution 45/55 % on big map can be considered for win by 55% owner even if he has no heroes and the other player has 2 level 5 heroes full of items which stacked have 25 leadership 60% group ambush, 7 chaos and +6 group move for example. Now Heroes (3 points + 1 per level!!) - if hero value is so low - why u even consider hero offers?

[smursh]

I know all this was discussed at the time the calculator was made. Some of it was done to simplify the system(so for example all cities are equil in the calculation)

I agree heroes are underated though. As the ladder season ends if you have a close game and are looking for a championship passing on hero offers and hording gold is a tactic to win the battle calculation. I dislike changing my strategies just for ladder points, but there it is. Plundering a city you take and not buying production also can help win the calculator.

Basically any system will have flaws, this is in place mainly so that if you are overwhelmingly close to victory your opponent can't just stall so the game goes to the next season.

[KGB]

I still don't understand why *any* code / time is being spent on this.

Let the ladder season end as normal. 99% of in progress games won't matter who wins/loses so just end them as is done now. The only games that actually matter are ones that decide 1st place. How many of these can there be? Let those few games continue and if that delays crowning a champion by a few weeks (or a month) so be it. My guess is there might be 1-2 games tops that are in progress that might decide who is 1st because I bet in a vast majority of times 1st place is already decided.

KGB

[gil]

i think KGB said it all.

if you ask me this should be canceled.

when i join a 80% castles need to win fight then i didn't win till i got that 80%.

any rule you make to bypass what the players agreed on before the game if unfair.


soulman i disagree with your assumption that on big maps heros are what counts. i think gold will make you win. and i will remind you instant hero+allay offers can appear on the front lines with no warning.

[SoulMan]

fresh level 5 hero + 3 dragons and fill with trashes can take down 0 or 1 trash unit from stack with 2 strong heros which makes gold really useful. Whats more, 2 strong hero stack has over 30 movement each turn. I think best idea is let the games to be finished, but if it is a rule which is counting game score - it should count it relatively balanced way.

I'm not very experienced player, this is my first season. But I think my values of in game things are relatively good, otherwise i couldn't do so well so far. Btw. In game I usually overrate hero even more than I did in suggested score counting scheme.

[gil]

I was thinking 3K gold lvl 5 assassin with 3 ghosts and 4 garbage units killing your hard earned hero stack.

i figure all maps with more then 20 castles per player are set on neutral castles once you have more castles you won.

i've been looking for a replay to prove me wrong since i started playing a year ago please let me know if you find any.

[smursh]

KGB may have it right that we should do away with this and only count games that have actually ended.

Main reason for this was to keep people from stalling at season end, but it still seems to me that that is what happens. Not just players looking for championships, but others seem to do so just for the pride of finishing a little higher. All that has happened is people are now stalling games sooner.

[SoulMan]

ok Gil. Im actually playing the FFA game, where opponent has like 2 times more cities and we are only 2 in game. Opponent seems to play every turn as long as he can, although he was playing really fast before. Now I have stack with 2 maxed barbarians, 89% group ambush, over 200% group ward (ok its capped to 70%), barbarian has the hunter (so his warding is capped to 90%). They also have great archon and red dragon stacked, over 20 leadership (over 30 in defense), over 10 fear and over +30 group anti air. Stack has over +10 group move as well. Generally i feel opponent has no idea how to stop this. When I attacked his "strong" city with about 20 units, 2 heros, with red dragon, green dragons, archons and assassin behind this - i lost spider and yeti - if i was unlucky I could lose catapult as well. Newly hired level 6 hero + whatever 7 units (assassin, great archon, red dragon, 5 medusa?) really can't take down more than 3 units from this stack because of 79% rule. It's just a bit annoying because if opponent will hold every turn for nearly 3 days i will not win this before end of season. Actually this game could influence FFA top 3. He is trying to sabotage and backdoor my cities, but im producing h.infantries or spiders in over half of them. I have strong enough hero to defense as well and about 5000 gold so if he raze city one its not a problem.

[gil]

SoulMan if that's a big ffa map with a 100 castles and he's got 2X your castles then your dead and he's doing you a favor when he delay.

[SoulMan]

67 cities map only, distribution is like 15-35. Its sure he has much more armies as well. So his 15 cities x 15pts = 225 pts. My 5 heros 3+1 per level each, I think about 70 points in total. I think this value at this point should be on favor of heros, or, armies value should be divided by sqrt(amount of cities) or something like that, because when you have more cities then you need more armies to guard it.