The Longbow Experiment Post #4: Conclusions

[InsaneGoblinEater]

Yah, real wars starts at th9/10. When you meet all those maxed th10 and heroes. That where the excitement kicks in whoever flop lose!

Yes, but thats wasnt what i was trying to establish. Totally missed the point.

[Superfinch30]

Yes it did, but the expertise in th6/7 ive achieved, combined with the incompetence of the enemy teams meant i could beat them. I won at least two 'hard' ones i should have lost entirely due to the lack of attacks by the other side. But they could easily still have won had they applied themselves - i was unable to star the top two in one instance, and could only rack up about 22 stars, and I still won cos they only made half a dozen poor attacks. You can still see the the spike in points every 9/10 wars (and you can see the drop in scores).

Right, that's what I was after.

I am genuinely intruiged, despite the suspicious nature.

I just fail to believe that the MM algorithm actually cares about clans runs. What happens when (impossible hypothetical situation) the only 2 clans suitable searching at same time are both on long win streaks and due the cold deck?

I'm only saying this because, weight/hero/mismatches seem to be pretty damned random to me.

In space of 3 wars, all three were unwinnable by one side or the other, given any normal attacking skill level for both sides.
That's quite regular for us in a 75% th10 25% th9 clan.

I'm talking about a 15 star win, 5 star loss, and a 34!!!!! star win. I would probably have put numbers close to all of those if had to bet on scores shortly before battle day.

[InsaneGoblinEater]

Right, that's what I was after.

I am genuinely intruiged, despite the suspicious nature.

I just fail to believe that the MM algorithm actually cares about clans runs. What happens when (impossible hypothetical situation) the only 2 clans suitable searching at same time are both on long win streaks and due the cold deck?

I'm only saying this because, weight/hero/mismatches seem to be pretty damned random to me.

In space of 3 wars, all three were unwinnable by one side or the other, given any normal attacking skill level for both sides.
That's quite regular for us in a 75% th10 25% th9 clan.

I'm talking about a 15 star win, 5 star loss, and a 34!!!!! star win. I would probably have put numbers close to all of those if had to bet on scores shortly before battle day.

ah but it probably works differently at high levels, where the skill level of the players is a far more significant factor. I deliberately picked a low level account set to try and minimise the effects of player skill, and so as not to swamp out any game effects.

Just a random stab, you probably get more mismatches at high level because there a more limited pool of clans of th9/10, than there is og clans of say 6000 points consisting of th5/6/7 bases. The game has more choice at matching. At high level it probably goes to pot, due to limited choice.

Needs testing somehow.

Right, that's what I was after.

I am genuinely intruiged, despite the suspicious nature.

I just fail to believe that the MM algorithm actually cares about clans runs. What happens when (impossible hypothetical situation) the only 2 clans suitable searching at same time are both on long win streaks and due the cold deck

It matches them together, cos thats the best match it can make. You cant have an algorithm whereby it gives up in the end. In the end it makes a match no matter what.

[InsaneGoblinEater]

I do enjoy the level of nerdiness you've taken this experiment to. I read through the other threads as well.

Why stop at the low levels? Continue the push into the upper levels of war and see if your theory holds weight while you figure out more complex bases and more complex attack strategies.

Are you able to three star any TH8 that comes up with your wars with your dragons?

EDIT:
Adding in first three posts

Winning Wars for Dummies

Winning Wars for Dummies part 2: The Longbow Experiment

The Longbow Experiment Post #3

It never threw a TH8 at me.

BUT it HAS thrown a couple of clans with THREE TH7 at me in a desperate attempt to beat me, which was thwarted by the sheer incompetance of the players owning those bases. By rights they should have won. One of those matches i was unable to star the top two, so by rights thats a 6 star lead for the enemy , usually enough for a good win. But the ♥♥♥♥ed it up. They had in effect six drag rushes, they used only one. The other two th7's tried to barch me.

[DirtyWolf]

Oh geez, could have saved you some time if I found this sooner, I've been doing this for a while now. So have others.

One comment, the best army composition for th6 and under is loons with 25-30 archers. This is what is typically used in these clans.

We're at level 7 in the clan we do this in, 82-2, with 48 perfect wars, longest consecutive streak was 65 wins. The two losses came from other Noob War clans (that's what I call them).

Well, we are now at won 53, lost 2, drawn 0. I have 35 consecutive wins, 13 perfect 30. The clan is now lvl 5 The 'unbeatables turned up around wars 35 and 45, but they have stopped being unbeatable.

This is because of two reasons. First I realised the skill level was going up, because the individual accounts were levelling up. SO i dropped the top th7 account, made the top th6 into th7 and put a new th5 at the bottom. This caused the matching to downskill back to start levels. I suspect every 50 wars this is what you would have to do, to keep the clan pinned at about 6000 points.

My alts are frozen at 6.5 when I noticed this occuring, and th8 regular players are moved into the main clan.

The other reason ive stopped losing is because my personal skill level with th6 and th7 drag rushes is now bordering on superexpert. Consider, ive done about 600 attacks with the same th6 army (40-50 archers, 8-10 giants, 8-10 ballons, 5 wallbreakers) or about 150 dragon rushes (8-9 lvl 1 dragons+ 1 lvl 3 drag). Almost none of the opposing th6 or 7's will have anything like this level of specific expertise

As mentioned, that is a sub-optimal army composition for th6 and lower.

[The fundamental flaw is that personal skill level , account level, and clan level are not tied together. Therefore an anomalous hybrid of high personal skill, high clan level, single person multiple account population, and low account levels produces a clan that the clan matching rules cannot deal with in a normal way, and thus always ends up producing an easy win for the hybrid clan.

That is an incorrect assumption, obviously, as both of our clans have lost wars. They will almost always win, so long as the population of opposing clans comprised of multiple accounts remains low.

Now some of you might say you already knew this. Ive no doubt. But the point here is ive demonstrated beyond doubt, and precisely how it works in practice and what to expect as a result. I am confident you coudl get a clan to level 10, 130 exp at a time, and it would get easier and easier, and the material cost in gems, elixir and gold would be a fraction of the usual costs. The timescale , however would probably be much much longer.

Why would you ever only go for 130xp at a time? You only need 25 strong players, as mentioned in a previous post, and the rest can be filler for 50v50. It is indeed the cheapest and quickest way to level a clan.

Im done with the Clan Longbow experiment, it took me 6 months and was very interesting, but i need to find another game to take apart now.....

I do want to mention that this is a special case, where we take advantage of superior game knowledge to win wars at the game's most basic levels. There is a reason we don't allow th8s to war with us. It's a whole different game from there on up.

[InsaneGoblinEater]

Oh geez, could have saved you some time if I found this sooner, I've been doing this for a while now. So have others.

One comment, the best army composition for th6 and under is loons with 25-30 archers. This is what is typically used in these clans.

We're at level 7 in the clan we do this in, 82-2, with 48 perfect wars, longest consecutive streak was 65 wins. The two losses came from other Noob War clans (that's what I call them).





My alts are frozen at 6.5 when I noticed this occuring, and th8 regular players are moved into the main clan.



As mentioned, that is a sub-optimal army composition for th6 and lower.


[COLOR=#000000][FONT=Arial]

That is an incorrect assumption, obviously, as both of our clans have lost wars. They will almost always win, so long as the population of opposing clans comprised of multiple accounts remains low.



Why would you ever only go for 130xp at a time? You only need 25 strong players, as mentioned in a previous post, and the rest can be filler for 50v50. It is indeed the cheapest and quickest way to level a clan.



I do want to mention that this is a special case, where we take advantage of superior game knowledge to win wars at the game's most basic levels. There is a reason we don't allow th8s to war with us. It's a whole different game from there on up.

Yes, you really dont understand whats going on do you. You need to read all the threads from the start, then you'll realise why most of your comments are irrelevant or wrong. Im not even going to bother ploughing through them, sorry.

[malikmoore26]

what device did you use for this? assuming android but what kind, or what are all the devices you used for this, kinda wanna try it myself