Imagine a pair of farmers who plant the same crop every year. Imagine it because analogies are fun. Or imagine it because agriculture is the basis of society. Imagine that one of them (the "May farmer") likes to plant his crop in May and the other (the "June farmer") likes to plant in June. Some years the rains come at a time that helps the May farmer produce the most crops, some years the rains come at a time that helps the June farmer produce the most crops. Now imagine that crop produce records are kept, and that as it turns out the May farmer has more good years than the June farmer. Consistently.

Now suppose that the June farmer is ALSO a meteorologist. They are smart. The June farmer understands a thing or two about weather and rainfall and maybe even about climate. The June farmer has lots of arguments about why it should be expected that the rains should come at a time that helps crops planted in June, and he loves telling these arguments to people, especially during the years when he actually does produce more crops.

And suppose that the May farmer doesn't know a damn thing about clouds. The may farmer doesn't know what "humidity" is or "air pressure" or even "evaporation." All the May farmer does is measure the amount of rain fall every single day and record the amount. This has led him to plant in May.

Warzone is full of June farmers. These players have ideas about how things should work out, and often these ideas sure do seem to make sense, but in the end these June farmers have more bad crop years (they lose more games) than May farmers. I am here to tell you that people who clusterpick in 3v3 games (such as putting all of their picks in the same megabonus) are June farmers. Clusterpickers lose more games in the long run, regardless of whether they are able to make good arguments about why clusterpicking is in theory useful. In practice, spreading out each players' picks on your team wins more games.

"You have your way of playing and I have my way" is a fine argument for those who enjoy losing just as much as they enjoy winning. For everybody else, results should matter.

Wait a sec... what you trying to say about meteorologists here?

Good analogy but I don't get what studying meteors has to do with Warzone.

good post, but sadly, your intended audience will never see this.

I'm a late May planter. Spread'm out in Spain and drop one or two in Britain or Africa.

since you only play ror let me give an ror example: https://www.warzone.com/MultiPlayer?GameID=15593757. Clustering can be great if done right. My clustering almost won us the game despite the fact that the other team consisted of much much much better players than my team (I had randoms, they had a set team - to see how poor my team was just watch egypt..).

90

EDIT: Clustering gets a bad reputation because people dont do it right, usually because most players who do it are inexperienced. But if experienced players do it, it can be great

Edited 7/10/2018 14:59:25

Sorry, 90, I just don't buy it. For every 20 games you show me in which clusterpicking worked, I'll show you 40 where spreading out worked. That's sort of what my main point was. Single instances don't matter, statistics do.

Even in that example you show, I could easily argue that your clusterpicking did nothing for the team. Yay, you got germany. AND your absence in illy/mace (your clusterpicking) also gave magna macedonia to enemy. Yes, your cluster pick gained you something, but clusterpicking also leaves something open.

In fact it's more certain that clusterpicking leaves open land than it is that you will complete the megabonus where you clusterpicked.

Edited 7/10/2018 19:08:11

I got germany and broke 2 bonusses on turn 1.. that is pretty great if you ask me. And if i had teammates nearly as good as the enemy team it was an easy win. In fact, the opponents told us they thought it was gg after turn 1. My absence in macedonia didnt give it to the enemy. Rather, my teammates noobiness in Alexandria gave it to the enemy. If he was a more experienced player he would have covered macedonia just fine.. Also, John was overcautious in UK, and ought to have cleared pichu while she was going for macedonia... but that is not the point. The point is that the cluster in that game was a grand success and i dont think it is fair to suggest otherwise.

I clusterpick in about 50% of my RoR games, and most often it works just fine.. this is not an isolated case

90

Edited 7/10/2018 19:36:23

@ 90 \(ºº)/ regarding your game:

You didn't clusterpick in the classical sense, a "real" RoR clusterpicker puts all his picks into one bonus, preferably Partha.

Apart from that your style of "clusterpicking" is documented in my M'Hunters strategy guide, although there I talk about clustering a first turn bonus with autogame settings so you get the bonus not initially after picks (https://docs.google.com/document/d/1NyhCpIQKShAbWGXicO_ph9whV_UyMS-3a7_re0C7R8Y/edit#heading=h.g1s8mzy7wor0 under "First turn bonus play").

Basically your strategy comes down to killing the opponent with your initial momentum which you gain from your income + army advantage after picks. Therefore you place your picks in areas where you find your opponent immediately or at least after turn 1. I can certainly understand how this can win games.

There is a simple solution for this problem. Bottleneck and 90 each assemble a team and play 10 games to establish a reasonable database on the success of their strategies with nonrandom teams on ROR .

@Norman : I strongly disagree with the comparison of cluster picking in a 1v1 autogame to a 3v3 on an expansion template like ROR.

Well, 90 used his momentum to kill his opponent and my that's what my guide states to look out for when considering a clusterpick.

It’s a full dist lottery too

how was the picking on 3v3 dror in cl9?

@Norman, aoc - On 1v1 Auto-games it might be worth noting that with ftbs you can often outexpand your opponents as well.. you dont necessarily need to play super aggressive: For example, if you have 2 spots taking a +3 and 1 spot in a +4, then you can complete both by turn 2 to get 12 income and this income momentum can be put into expansion making it hard for opponents to keep up. In contrast, i think in RoR it is a bit harder to pull off an "expansion" strategy with clusterpicking. So I do agree with aoc's point that comparing these templates is not entirely fair. However, I do agree with Norman's observation that in most cases, an essential ingredient for the success of this kind of RoR clustering is to take early fights and make them count. I think that the "aggressive" strategy w.r.t clustering on both templates has a similar spirit that Norman is trying to point out.

We could do this experiment as aoc suggests: I would take my team as aoc and Juan, but Juan hates RoR so much i doubt he would play :p

90

Edited 7/10/2018 22:06:52

May farmers farming in June is good.

Cool! Discussion ongoing.
Well the first thing I want to highlight is the fact that in my original post the thing I call cluster picking is... where is my quote... okay, I say "such as putting all of their picks in the same megabonus." 90's picks are reasonable because picks 1,6,7 are on the OPPOSITE side of the map. Thus the non-spreading-out weaknesses of clustering are largely absent. The remaining weaknesses of clustering are just the risk involved in leaving some map open and having so many picks in same place. Like if the lone partha pick had encountered nonzero resistance... 90's clusterpick remains totally doubleteamable, and it's a doubleteam that is potentially devastating since 90 has no other area where he may grab a bit of income.
But anyway yes many of the benefits of spreading picks are still present, so while I'm still extremely skeptical i'm somewhat open to people trying to convince me that that style of picking can work as often. In fact I employ it myself when I feel like taking a risk and having fun.

such as on the RoR map, a version of "clusterpicking" I enjoy doing sometimes is one pick partha one pick asia and 3 picks completing that little bonus in gallia by UK. When it works it tends to work great. But it has also failed spectacularly a couple of a times if enemy within that 2 bonus. Even when I play intelligently by putting myself in different parts of the map, when I cluster 3 picks together it is a risk.

"I clusterpick in about 50% of my RoR games, and most often it works just fine.. this is not an isolated case"
If it works out fine most of the time, then you probably aren't going all picks same megabonus, are you? Access to map regions matters.
Alternative response I could give: I spread out almost every single game, which works 80% of the time.

"There is a simple solution for this problem. Bottleneck and 90 each assemble a team and play 10 games to establish a reasonable database on the success of their strategies with nonrandom teams on ROR ."
Since arrogance is fun: Nah, my team would simply have better skill, so there's no telling whether it would be our superior picks that win it. :p
More seriously though I think this would be slightly unfair to the team testing 90's form of clustering picks... since a primary benefit of 90's strategy is SURPRISING the other team with the cluster.

Cluster picking can be great if you're stuck with 2 awful players and need a secret win condition. You overwhelm your nearest opponent and take Parthian Empire while you use your 2 teammates and a distraction then 1v3 to victory.

@Bottleneck - that is a good point that the element of surprise is important for my style of picking :) - so I agree that such a test might not be a good indicator.

"a version of "clusterpicking" I enjoy doing sometimes is one pick partha one pick asia and 3 picks completing that little bonus in gallia by UK. " - haha yeah those picks are exactly the kind of thing which I enjoy too, and it often works great as you mention :).

Nox's comment is also insightful - spreading yourself out can be great when your allies are good players but if your allies are all noobs, then maybe it is better to be a bit selfish and go for income so that you are the strongest player on your team. "going for income" could either mean going for places you dont think the enemies will be, or it could mean clustering :)

90

Edited 7/12/2018 13:39:50

All in Armenia or you're a noob.

Good post m8; the May and June idea can be expropriated to all the important fields of study.