Warscore Manipulation (not matchmaking)

[Gryphonknight]

@Garanwyn

Sartana and Aeron staring at me, hungry for feeders. It takes a lot of willpower…

So shiny. . .

@Gryphonknight

He can join the club. My Frida, Marjana, Isarnia, Vivica, Misandra are currently not speaking to me.

[Anaea]

I’ll just throw my 2 cents in here. I’m in a smaller alliance as well (only 12 participating in war) and we routinely see exactly the same sorts of mismatches described in the original post and I can typically look at the match-up during the prep phase (before the opponents teams show up on the board) and have a pretty fair idea whether we are going to be competitive or if one alliance (usually us) are going to get blasted just by looking at the player levels. If the opposing side has several players whose levels are 10 or more levels higher than our highest, we are going to get beaten pretty badly unless those players opt out.

As @Kerridoc suggested the reason is almost certainly depth. These players best team is probably no better than any of ours, or only marginally so, but their deeper benches mean that their 3rd - 6th teams are way, way better. Unfortunately the algorithm intentionally devalues this difference by weighting heroes 6-30 less than heroes 1-5. As for this being a more significant issue in smaller alliances this is because in a war between small alliances one or two high-level players with deep benches are capable of one-shotting the top part of the opposing alliance and are going to have a much bigger impact on the outcome than they would in a 30-on-30 war where the top players on each side can only reliably take out a small percentage of the opposition. If one player gets 3 one-shot kills against an alliance of 12, this is 25% of the opposition, whereas the same player doing the same thing against an alliance of 30 is only taking out 10% on one set of flags.

This depth issue probably extends to troops as well.

[krainsy]

I actually feel the opposite and think the warscore calculation doesn’t apply enough weight to the defenses (top 5 heroes multiplier)…

Perhaps it is different in 12v12 war but in my full alliance, the sticking point for us seems to be # of 3k+ defenses, as it becomes very difficult to find targets for flags 4-6 and points per flag plummets.

So while I get what you are saying about opponent players w higher levels and thick benches being able to earn more points, that imbalance doesn’t limit your ability to score points for your team. I think it’s a bigger problem when your alliance only has 13 3k+ defenses and your up against an alliance with 25 3k+ defenses that your teams will have to face with all 6 flags. Even if you theoretically have more depth than your scary looking opponent with higher defenses, you probably aren’t able to capitalize on the team kill point bonuses as much as your opponent.

[AKSU108]

Thank you for your reply. I understand what you said and I had a look at the warscore, this was the main reason why I started this post. The warscore was the same, whilst the opposing alliance was clearly stronger than ours, not just stronger, they were on a whole different level. The additional player on our side with the strength of 1800 could do NOTHING against their team. We were steamrolled, they cleaned out field twice, whilst we could only manage one complete destruction of all their defences with coordinated teamwork, after which we were left empty. They won by a 1000 points! And during all that our warscores were the same. My question was exactly that, how were they able to attain such a low warscore whilst their heroe’s were clearly stronger and more numerous than all of ours put together??

[AKSU108]

@Wharflord

Despite your lower average team power, you have an extra player which makes a very significant difference in score. Not only do you get 6 extra hits, but also each enemy team is worth 17/16 = 106.25% of your teams on average. These two factors give you an effective score boost of 13% over your opponents, which should go a long way to counteracting the difference in team power

True enough, we had one extra player, but does that seem fair to you to warrant extra 5 players with 3000+ defence when we had an extra player with 1800 top team?? He could do absolutely nothing during the war.

[AKSU108]

@princess1

Before the war scores were figured in (and after the updates in-between), our alliance had 25 LOSSES in a row. I completely understand and commiserate with any alliances experiencing this scenario.

Since the war score has been factored in, our matches have been very fair (touch wood). We win about 60% of our wars, which we are fine with. We enjoy them all, even the challenging ones.

Our alliance always fields 29 - 30 teams and it’s my understanding that it’s alliances with fewer teams who are having an issue?

Hopefully there will be a solution soon so every player has the opportunity to enjoy this part of the game!

Thank you for your reply! We are all in favour of a challenge, even a difficult one! But it is a completely different matter when we come up against an alliance which we have no chance against, absolutely no chance. To be completely annihilated without any chance of even retaliation is demoralising, we have junior alliance members who ask us repeatedly “Why is this happening?” and unfortunately we have no answer for them yet.

[Kerridoc]

Thank you for the extra info about the matched war score—it’s refreshing to have a thoughtful and well-researched complaint raised. What we have here is not some broken matching process–the opponents had nearly equal war scores and nearly equal sizes. What’s broken, therefore, is whether the war score, as currently calculated, is actually a robust measure of an alliance’s tools to wage war.

@mhalttu should think about this issue. My suspicion is that the core cause is four-fold:

1. Disparate strength of members: two mid-tier players are deemed equal to a senior player + a junior player. In practice, the two mid-tier players will do worse
2. Over-valuing weak offenses when aligned against strong defenses
3. Under-valuing troop depth. When stacking heroes, have multiple good troops in a color matters
4. Placing no value on heroes 30-50, even though these are situationally useful.

Anyone else have thoughts about how the war score should be tweaked to reflect actual strength of opponents?

[AKSU108]

Dear Kerridoc,

I firmly believe that it’s not about the heroes depth in backup teams or troops strength. I believe that some alliances have found a way to fool the system. We did tests within our alliance and we are stumped. We gradually took off participation in war ticks and measured how it affected the warscore. It behaved logically and reduced according to strength. However, the opposite alliance teams were clearly superior so the only expanation I have (it’s only a theory) is that participation is the loophole. Loosing a war totally and without any participation may affect the warscore so much that the next match is easy pickings (i.e. us). However I may be wrong on all points expressed as I have nothing to back these theories up with. We have agreed to do a test with zero participation next time we are faced with an overwhelming opponent and see how it affects the warscore. I will update on the results.

[AKSU108]

And I would like to say that it’s not so much a complaint but rather a cry for help and an attempt to understand what is going on. We are trying our best to keep the morale up within the alliance but this atmosphere of “no chance for victory” is starting to affect everyone in our ranks.

[princess1]

@AKSU108

And I would like to say that it’s not so much a complaint but rather a cry for help and an attempt to understand what is going on.

I totally hear you and commiserate. I wish I could offer more than that. We were in the same position before it was fixed (for us), with the war score addition.

I know my many “complaints” were seen just as that when in reality, I loved the wars before they were broken with the previous updates, and was also just begging to have them fixed.

Please don’t give up! Keep reminding everyone that it’s not fixed for everyone yet.

[Gryphonknight]

Elo

@Kerridoc

Anyone else have thoughts about how the war score should be tweaked to reflect actual strength of opponents?

Elo is time tested

Click for very long, very boring, discussion

==Reusing code or Individual war trophies==

I still don’t understand why you don’t just use the computer code for raid programming with a new variable name.

Just like every player in an alliance has a raid trophy score, give them individual war trophy scores.

Example:
If Petri’s war defense team is attacked three times, and the attackers fail all three times, Petri’s individual war trophies go up three times based on Elo’s math. If Petri’s defense team is attacked three times and the attackers succeed all three times, Petri’s individual war trophies go down three times based on Elo’s math. If Petri’s defense team is attacked three times and the attackers succeeded once, it doesn’t matter to Elo’s math if it took one player to succeed, two players, or all three. Petri’s individual war trophies go down once based on Elo’s math and up twice based on Elo’s math. Instead of conceptualizing it as three different players attacking Petri’s defense team three times, think of it as one player attacking with three different teams.

The same math applies to the attacking teams even though they are only used once per war.

Example:
I have at least 3 okay 4* 3.60+ heroes of each color. When I am raiding, I adjust my attacking team based the defense. I am currently fluctuating between Gold and Platinum ( thank you 4* 3.60 Hansel ). Elo’s math does not care that I am attacking different defense teams or that I am using different combinations of attackers. It only cares if I win, lose or tie.

==Using the individual war trophies==

Players that have not participated in the last two wars, have a temporary war trophy score of zero for matchmaking purposes only. Some sort of war trophy decay may also be needed for very long absences.

For an alliance war trophy score, add all the individual player’s war trophies together then match alliances for war using this variable.

Elo’s original math is designed not to care WHY you got a won, loss or tie. If it did, it would never work with MMOs which are influenced by energy meters and heroes available. So alliances that work together would still have an accurate score when adding all the individual scores together regardless of membership, activity level and hero roster.

Example:
Petri High School in Atlantis has a chess club and a darts club. To match players for chess they use Elo’s math. To match players for darts they use Elo’s math. The only difference is there is one score for chess and one for darts since players who are good at chess may not be good at darts and vice versa. They are not creating a new algorithm. They just keep track of chess wins, losses and ties with a chess variable and keep track of the darts wins, losses and ties with a darts variable.

As a benefit the code is already written. You are just using the raid trophy code and assigning each individual player one war variable ( war trophies for all wars ) or three war variables ( war trophies for each war rule ). I prefer the three war variables since I think raids, war rule #1, war rule #2, and war rule #3 all require different defenses and attackers ( Thank you whoever decided to let us change defense teams during the war ).

This solution would also account for players leave, joining, being inactive, or becoming more active.

==War rewards based on losing alliance’s bases==

If deliberate tanking becomes a problem, the Devs can use the losing alliance’s total food and total iron storage to adjust war loot similar to the way the losing player’s watch tower capacity is used to adjust raid loot ( food and iron ). In fact, this may be a feature the Devs want to add even if none of the alliances try to tank this system and even if the Devs decide to use another matching system.

Example:
If the losing alliance all has Food and Iron storage at 10, the loot, for both winner and loser, would be lower than if the losing alliance all has Food and Iron storage at 20. ( Sorry, I don’t know the actual amount of total food and iron for food storage 10/ 20 and iron storage 10/20 ).

Just to add to this, I don’t think using points would be a good factor.

==Factoring in War Points==

I agree with you.

Points are not used with raid trophies, and raid trophies works fine if the attacking player is not deliberately tanking. Elo’s math is designed to predict wins a majority of the time ( if it knew who would win, there would be no need to have this discussion ).

With Empire’s new Elo math based raid loot system ( food and iron ), War points are effectively “raid loot”. With team HP being watchtower food and iron storage. At the end of war, the alliance with the most “raid loot” wins, and then both alliances turn in their “raid loot” for their actual war loot.

In the “raid loot” example above, reviving teams on a timer almost works like a ladder system, but that is another discussion entirely.

This would help with unknowable asymmetrical like Elpis:

@Elpis

This way i’m up to 20 wins straight in war.

The current matchmaking is hampered by his hero roster. Elo’s math does not care. It would just keep bumping him up until he fought a losing war then knock him down.

While a losing alliance would keep getting knocked down until it eventually won.

Opt out

@AKSU108

We have agreed to do a test with zero participation next time we are faced with an overwhelming opponent and see how it affects the warscore. I will update on the results.

Nothing will happen.

Matchmaking is based on your participants at the time of matchmaking. If you all opt out, it effects the NEXT matchmaking.

Abort abort abort

To save moral, you can abort. But aborting will not influence your next matchmaking since it does not count as a win or a loss AND you lose you 1x ascension item roll for a lost war.

The Devs have been doing this a long time and have seen it all.

@Gryphonknight

The Devs do not trust us - which let us be realistic, is smart on their parts.

[AKSU108]

Erm, probably I did not make myself clear. I did not mean taking off the tick from the war. No, we will be in the war, we will just do nothing. None of us. Just let them beat us with 100% of our war tickets remaining.

What do you mean by abort? There is a way to stop the war once we were matched??

[DBC]

If someone with access to the data, examined alliances that have long winning streaks and compared them to alliances with long loosing streaks. Once obvious factors like low participation have been factored in (if they ain’t trying I don’t care about their war experience any more than they do) Then surely the way that heroes are measured towards war scores could be adjusted based on the reality of what is actually happening?

[Gryphonknight]

Members skipping war

@AKSU108

Just let them beat us with 100% of our war tickets remaining.

If you do this, remember to have all members check their war opt in. If they forget and this happens twice in a row, they will be automatically opted out. This is the real reason the Devs added Individual Opt outs. For members skipping wars.

Low investment wars

When I am working long hours at work I just use find opponent, auto fill and attack for all six war energy. Takes almost no time and I do not feel as demoralized because “The computer picked my opponent and teams” but my alliance still gets war points.

This might be a better compromise.

CONFIRM ABORT (Y/N)?

@AKSU108

There is a way to stop the war once we were matched??

There are several ways. Some have real consequences.

But a loss will match you with a weaker opponent next time.

An abort just skips this week. Think of it as a retro active opt out.

Are you SURE you want to abort?

[AKSU108]

I am just interested in how it is done? Do you mean taking off the tick from alliance wars on leader/co-leader level? Is that what you call an abort?

[Elpis]

I’m definetely against it.
All is fine guys, all is fine

Nothing to see here.

[General_Confusion]

I think you got most of them, I also think:

5. make sure that the algorithms scale correctly from alliances of 1 to alliances of 30

Along with my suggestion in the APL for dealing with #1, I strongly believe that the weight of the top 5 heroes needs to be increased. Most/All of these are used in defense, and then again in offense…and a 3100 teams trying to take out a 4100 team needs a lot of bench strength to even that out.

[Gryphonknight]

Power

@General_Confusion

strongly believe that the weight of the top 5 heroes needs to be increased.

Book of heroes tried SGG war matchmaking for 1v1 Raids but eventually switched to Elo because Elo works. I will not bore you with the details but it boils down to no man is a number. Kashhrek is substantially stronger in some situations than others. Giving him a flat value is almost guaranteed to under value him versus some opponents and over value him versus others.

Elo’s biggest strength is it knows nothing it arrives at its values by empirically testing it values and correcting them when it is wrong.

Run away

@AKSU108

just interested in how it is done

This abort is the fastest and the easiest.

Leader opts out of the next war.

After war starts ( edit: thanks @General_Confusion ) , to abort, each teammate in the war leaves the alliance and rejoins.

This should automatically end the war ( the Devs may have changed it ) but resets everyone’s time in alliance.

Illegal Formation

Leader opts out of the next war.

All teammates use a 1* 1.1 or 2* 1.1 as their right corner war defense hero.

After matchmaking, or after war starts, to abort, each teammate in war uses the 1* 1.1 or 2* 1.1 right corner war defense hero to level a hero in their roster, removing it from their war defense team.

This definitely ends the war but puts you as a big disadvantage if you decide not to abort.

You mother was a hamster

Everyone attacks with a single 1* 1.1 for all six war energy.

This abort requires you to wait out the full 24 hours of war, but does the least weird stuff if you decide not to abort.

-=

[General_Confusion]

@Gryphonknight

Elo’s biggest strength is it knows nothing it arrives at its values by empirically testing it values and correcting them when it is wrong.

And it also has some pretty big issues. No system will be perfect, but there are things that can be done with what we have to tune it to better than it is. The hard thing is to improve the edge cases while not harming the improvement for the majority of cases, but I firmly think it can be done.

[General_Confusion]

@Gryphonknight

After matchmaking, to abort, each teammate in the war leaves the alliance and rejoins.

This should automatically end the war ( the Devs may have changed it ) but resets everyone’s time in alliance.

They did. As long you leave after matching starts (someone said 1 hour after) you can rejoin and fight.