Startpos in 1vs1 games

I do not think there is a system that will make 1v1 games fair...first person that takes a turn has the advantage... this is why I do not like 1v1 games here or in RL.
Then there is a great drop also, where as in more players in a game, then they work together to take it away from that person. Fog may even the field out better than anything else.

Have the neutrals always take at least 1 territory in a continent upon start, then I think as stated, each continent gets one player of each color until the continent is full. This of course will give a player an advantage in one continent, but a disadvantage in another - thus evening out. Also, this ensures no one starts with a continent.

This would never be fairer, the lesser Neutrals to kill is even more of an advantage for the first player.
Putting 3 reinforcements on a starting 3 and trying to attack 3 Neutrals (6 v 3) regularly kills 0 or at least causes almost guaranteed attacker losses.
There is a huge difference in the average attack results when 6 v 2 or 6 v 1.

My opinion only: I like SethHrab's idea of first assigning 1 random Neutral territory to every region, then the usual random assignment of the rest of the troops.
This completely eliminates any start bonus!

There will always be player majorities in most of the maps regions. However, there will never again be any extra reinforcements to advantage a player unless they play for and win a bonus region.

Exactly. I don't think there's a real solution to having neutrals blocking paths, heck, I actually like that added gameplay feature, intentional or not. However, this does make a player have to actually conquer a continent to have the bonus, not be gifted it by luck. Hope it's a feasible coding solution!

I'm afraid Sethrabs solution won't work on all maps. For example the maps where we have regions with a negative bonus or overlapping regions will have some strange stuff going on. Sadly, because some regions are used in a bit of a hacky way, we don't really have a good definition of what is actually a region.

If we are to change the game setup algorithm, it has to work for all maps, not just world map

Cireon, could you provide a more in depth explanation with a map and regions as an example? That might help to understand the issue and propose an additional solution.

As in, what map has a negative bonus for holding a region? What do you mean by overlapping regions?

Black Sea, Eld World, and East Indies are just three.. Oregon Cities, Battle of the Elements are two more.
Not so on Black Sea I do not think though.

Thorpe is right. Negative regions are often used in cases where you get bonus for holding X out of Y territories. Battle of the Elements is a good example regarding the air temples.

This is a screenshot of the region setup for that map:



I realized that we have a "special region" flag for those region which makes them show up differently in the log I think, so maybe we could use that to exclude those regions from the generation as well.

To be honest, I am highly skeptical about making changes to the generation algorithm for two players. I am a bit of a purist I suppose when it comes to keeping things as close to true pseudo-random as possible. However, I am willing to give this particular thing a try. We could generate like 10 or 100 maps with both the original and this algorithm and see if we indeed like the new algorithm better.

@Matty what do you think?

We could try it, but I'd rather not.

You see, in 1v1, every troop that you lose killing a neutral, is a troop you could've used to kill the other guy. So it's usually a really good thing to not kill any neutral at all. So the side effect would be that in 1v1 games (or 2v2, whatever) it will (almost) never be worth it to try to go for a region, and therefore kill off a large part of the game.
Of course, sometimes you are desperate, so it could be ok to pass a very few neutrals. Sometimes you can take a region that your opponent really can't get to (because other neutrals blocking the way?) so it may be worth taking it, as you can't use those troops anyways and the region is safe (and even if you opponent breaks it, he too will have to waist troops on 'something that's not you' ).

Of course, these situations are pretty much the situations that the topic starter was trying to avoid in the first place...

So yeah, I'd rather not.

Could you not compare current algorithm vs new algorithm? Maybe make it an option like "balanced dice". Do you want "normal" start positions or "balanced".

@dough_boy: No. I think adding even more options to the create game menu is actively harmful for the user friendliness of this website.

@Matty: Yes, you make a good point, and I realised the same thing. However, it is the best solution I have heard so far I think. I guess we'll have to keep searching for a good solution...

The other solution proposed is not to add less randomization (what Sethrab proposed), but to add less neutrals (I think this was dough_boy's suggestion).
Right now 1/3 of the territories are neutral (rounded up). How would it play if they are, say, 1/5 (rounded up)?
That's something I'd like to experiment with.

As for the option, we can use an extra setting in the create a game page that is NOT visible by default, but only toggled on in some hidden setting somewhere, so that, say, dough_boy and Hoodlum can experiment with it (set up a tournament without and one with it) and then see which one is more fun or something.
I'd be willing to try this at some point.

Works for me...happy to help out as needed.

Obviously not a major/pressing issue.

I think there is a focus on two different issues here:

1. Too many neutrals or neutrals influencing/blocking gameplay (to me this isn't an issue.. as Matty states, every troop used to kill a neutral is a troop that could have been used to kill the enemy, so.. uh.. strategy.)

2. Neutral generation/randomization leading to some players starting with a continent, therefore having a HUGE advantage over another player, especially if they also start the round. (To me this is a real issue - random, sure, but it happens to also be incredibly unfair and very difficult to come back from.. essentially, this type of game is more often than not decided before any player puts troops on the board because of the rather unbalanced advantage.)

I'll add more thoughts later, I want to take some time and examine what Cireon posted in more detail.

So, thinking about the less neutrals suggestion (Issue 1 in post above) - would this not then exacerbate potentially the no region from drop (Issue 2 in post above)?

In other words, by fixing one issue, you could potentially make the other worse.. i.e. now a player might start with two continents rather than just 1 from the drop and have an even larger already unfair advantage over another player.. I get that it's random.. however, random isn't always best.. I really do think random with some semblance of order is better.

So, obviously not knowing how the algorithm is constructed, there's some guess work here, but I think the most fair way to start any map would be to have defined regions first, and then the drop (algorithm) provides each territory with one player of each color within that region until it is filled, alternating the first player to get a region as the last player to receive.

As an example Region 1 in Battle of Elements map defined as Eastern Air Temple with 3 territories. Territory 1 goes to neutral, 2 goes to player 1, 3 goes to player 2. Then region 2 defined as Fire Nation with 7 territories territory 1 goes to player 2(the last player to get a territory), 2 goes to neutral, 3 goes to player 1, 4 goes to player 2, 5 goes to neutral, 6 goes to player 1, 7 goes to player 2. Etc, etc.

The other switch to this might be just considering that the first territory in a region ALWAYS goes to a neutral, thereby being certain that a player never gets a continent or unfair bonus. This should solve the negative territory issues I think, so those "specials" are always occupied first by a neutral.


My thought is why not do a set of test games (maybe 10-20?) of each proposed viable solution. So test set 1 takes Matty / dough_boys suggestion of 1/5 neutrals. Test set 2 takes my suggestion of organized randomization. Just for kicks, maybe a good idea is a combination of both in test set 3 that gives 1/5 the neutrals AND organized randomization.

I'd be willing to jump in and help test if needed.