[LOW] Initial placement of Kings

[Natmus]

In my opinoin there has been very few good 2 player games lately. Especially against Ringthane I have encountered him again and again during the first two or three days. The constant pattern is that one start in the NE corner, the other in the NW corner. I truly have problems recalling a single start in either of the S quadrants in a 2 player game. Something is clearly wrong here.

As a temporary solution I would suggest that opponents start in opposing quadrants in 2 player games (Yes, I know that automatically will give away the starting position of your opponent), as the current method is intolerable. So if the first player randomly is placed in the NW quadrant, the opponent should be placed somewhere in the SE area.

One immediate effect on new games would be to allow players to recruit some sizeable armies before bumbing into each other, allowing for wargames instead of recruiting games.

BTW, I have chosen to place this under Bugs rather than Suggestions, as I firmly believe the current method or 'random' placements is flawed somewhere.

[sparrowhawk]

It stems from the dodgy shuffle function in PHP, which I use to randomly shuffle the quadrants used to start games. It has a known bug in it which renders it less than optimally random!

I meant to implement a free 3rd party variant for 0.1g, bit did not get around to it, though it should be very easy to do (test that new function is more random, then swap one function call for another).

I'll see to it asap.

[Matija]

I quite agree. It is a bug. It spoils the game.

Not only should players start in opposing quadrants, but also the same distance on both x and y coordinates from the center of the map.

In my last game with Fernando, I started near the centre and he somewhere near the W edge, the result being that I control mor than 3/4 of the map.

[sparrowhawk]

Placing players in opposing quadrants and equidistant from the centre would let you know where your adversary was. I think that that would spoil the fear-factor and lead to a rush to the centre every time.

However, with better random quadranting, and maybe making the quadrant ranges smaller so that they are not flush to the frozen wastes would help to resolve the problem a little more I hope.

I also want to implement that comepletely random option (ie non-quadrant based). But as an option of course. Just to make things really tense!

[celebaglar]

Matija said:

I quite agree. It is a bug. It spoils the game.

Yes it is and it does. At least we now know why.

Not only should players start in opposing quadrants, but also the same distance on both x and y coordinates from the center of the map.

I totally disagree. This is not chess. I for one feel this would be disastrous. Knowing exactly where your opponent starts to the square means you can work out what his recruitment strategy is likely to be. You might as well play the game with a top-down view and all armies visible.

In my last game with Fernando, I started near the centre and he somewhere near the W edge, the result being that I control mor than 3/4 of the map.

Luck of the draw. Sometimes it works for you, sometimes against. The problem is not the random positions, it's the overpopulation of the land. This is what spoils the games, as a small advantage in starting positions often results in 15-20 extra lords, which is bad for the game after the recruitment stage.

How about some of the lords only appearing in 3 or 4-player games, but not in 2-player ones? I'm convinced this would improve one-on-one contests which are the ones to suffer most under 0.1g.

[sparrowhawk]

Hmm... overpopulation does seem an issue now. Before we had lord-deserts, now we have lord-oases (is that the plural of oasis!?).

Maybe what you suggest would work. An option to drop minor lords (at random, so that you never quite know who's about) could be done easily. How would we determine minor though? Lords without strongholds (but Kathryn for example is fairly potent), size of starting army (but some citadel lords start with smallish armies), or maybe on a lord-by-lord basis. The latter would require a minor DB change, but nothing that would impact current games. However I'm sure that it would also generate intense debate and maybe never get reolved!

Or.... remove some of the new lords altogether?

[Ringthane]

No, not removing. Just increase the map size by 10 squares x and y. Use some random fill along the N, S, W and E lanes.

[sparrowhawk]

Oh yes, nothing controversial then, just expand Midnight itself... ;D

[celebaglar]

Just as a side-effect, removal would also help one-on-one games by reducing the amount of time one needs to move all his lords. One-on-one games don't need so many new lords. They were fine pre-01.g.

[sparrowhawk]

Maybe the system should randomly remove x number (5, 10?) of lords if the option is selected, irrespective of who they are. Even the citadel lords maybe, opeing the possibility of abandoned cities...

A nice side effect is that as you move around the land, when you see in the distance that a stronghold is unoccupied, you won;t know whether the enemy got there first or whether the lord was never in the game at all.

[celebaglar]

TheCreator said:

Maybe the system should randomly remove x number (5, 10?) of lords if the option is selected, irrespective of who they are. Even the citadel lords maybe, opeing the possibility of abandoned cities...

A nice side effect is that as you move around the land, when you see in the distance that a stronghold is unoccupied, you won;t know whether the enemy got there first or whether the lord was never in the game at all.

Yep. I like that suggestion.

The only proviso would be that removing random lords still needs to leave the map roughly balanced, or else the effect would be detrimental rather than helpful. One way to try and achieve this is to always remove lords in pairs, one from each diagonally opposed quadrant.

[sparrowhawk]

I've added the functionality in R7 uploaded just now (ok, I know that I wasn't going to add new features, but I couldn't resist!)

The quadrant-balancing is not implemented yet, so it's just a random selection and their garrisons if any. I will look at the balancing option for future release.

Also: the Game details don't yet show the nature of the game, so no indication of whether 5 or 10 lords have been removed. Will do this next, but in meantime, please indicate this in the Forum thread, eg 123 - Test game (-5 lords)

[Matija]

I don't like this feature at all.

I like to plan ahead, and with this it's impossible.

[Ringthane]

See the Game 52 thread for some insight on field testing this option.

[Natmus]

I'm with Matija on this issue. Randomness will only make luck an even more important factor in the initial placement of king, combined with randomly missing lords I can easily imagine even more lopsided games.