Vitenka: 1520
I made this / Cat Games - Less lag, more 'thats bad'
[vitenkas.morat.net]
/ MORAT / freespace / Catnews / Editorial
There's nothing to see here except for shadows of the past - and these ones won't be returning.

I'd point you to my next project here - but I'm not that organised. My style is to act and then sort out the consequences, rather than the other way around. Oh, and lying. I do that a lot too. (i.e. if you look closely, you may have seen some links appearing roughly once a week)

Vitenka.com is registered to me for the forseeable future, so you might find something there.

Edited by Vitenka at 2003-04-09 08:22:54

 
Vitenka : Mon 2 12:56:31 2001  
... wherin the properties of a team are discussed, including their preference towards forward defence and also the ways in which they might be manipulated to advantage...

Team Flow

But first a confession. I have started playing on BW2 again, despite everything that was done to it, and the destroyed community. Every other server was empty, llama, or packed too full to get into.

The initial conditions

Assume that only a small percentage of the players on the server are good (this isn't ~A~)

A huge midfield mess inevitably starts up. It can be safely said that the only people who make it through that relatively intact are the good players.

The lone defender

Consider an engineer in this situation. He can build a sentry gun up in the ramp room, where it will do noticeable amounts of work, but get destroyed a lot, or he can build it down in the basement.

If he builds it in the basement, the ONLY people who will see it are the good players, who not only make it across the midfield, but also don't stop to spam the enemy respawn.

Those players will destroy it without a second thought, and get the flag anyway.

So, in that case, the engineer sees long periods of inactivity, followed by it blowing up, and the flag going walkies. Hardly a good place for a SG then.

So they move it to the ramp room, where instead it does lots of good, killing people and stuff, and sometimes gets overwhelmed, and the flag goes walkies.

Conclusion

From the defenders pov the rr seems a better sentry gun place, from an attackers pov, useless and avoidable.

Oh... dear

This is just one example of a myriad of ways in which the wrong lessons are taught. A local performance maxima is reached (the engy will eventually learn the best rr position) but the global maxima of 'the enemy not getting our flag' is never found.

So - how can the better behaviour be taught? There are several ways people learn, let's try each of them.

Imitative

Monkey see, monkey do.

This can actually be a HUGE problem when it's applied mid game. It leads to sentry guns built right next to each other, entire teams playing defence, or as a single class...

But if someone could see what someone else is doing, and that what they are doing WORKS without being able to mimic immediately, then they might retain that knowledge for future games.

One actual positive idea here. The new halflife demo technology brings the potential of instant replays. If during the score displays the players could be shown scenes of what the enemy did right (assuming they did anything right) then they could pick up tricks, while being too late to mimic immediately. You need the new tech to do this, because often it is the combination of players that works. The problem would be automagically deciding what is 'right' to replay.

Book larnin

Read a website, follow its advice. This is preaching to the choir, the people who read tactics websites are already smart enough to learn.

Still, making sure that l33t websites are shut down, and only sensible working tactics propogate is important, otherwise people will learn the wrong thing.

Shout at 'em in game

This can actually work. But you have to be NICE. People don't like being ordered, phrase everything as questions, ask as though you are trying to learn from them - request help. Thank people for their assists, so that they assist more.

Don't expect miracles though, many habits are too far ingrained to change in a single half hour.

Watch for overflow

The monkey see, monkey do effect is VERY powerful. It can be used to great advantage, but also has great peril.

A positive of it, is when it mimics morale. You finally bust the enemy defence, and after the second time, you find that half of your team has moved onto offence to support you and that the enemy offence has slackened.

The negative? Sudenly you realise that EVERY member of your team has followed you onto offence, leaving nothing in your base, which the enemy has happily occupied, and will chain cap several times before you can get it rebuilt even if you COULD persuade your team to get back into position.

Watch out for this. I'm afraid that it happens even in a clan situation (where you have the 'hit them with a stick until they learn better' option, at least) this happens. You will have to learn, very often, to do the opposite of your team.

This leads to you acting as a regulator, and also means you rarely have time to fully evaluate the enemies style before you switch roles and meet a totally different threat.

This can be good, it teaches rapid adaption skills.

When learned behaviour goes too far

I'm sure you remember the days of non deathmatch midfields. Seems everyone pines for them, but the gentlemens agreement is gone.

Well, look on ~A~ they really DO enforce it. If a player is offence, they won't open fire until they are in the enemy base. This includes if they see you in their ramp room!

There is one exception to this, which is if you open fire first, they WILL retaliate.

Now, I'm not certain this applies to ALL ~A~ players, but it does to all the ones I saw last night:

  1. If you wish to DM/chase pick on someone you can beat - the first surprise shot won't win a battle.
  2. You can take potshots as you run, but don't do it too often against the wrong person, or they'll chase.
  3. If you do midfield, they WILL chase. All the way into their base.
  4. I spy with my little eye predicatble manipulable enemy behaviour. This can be used to cause the enemy offence to slack at predesignated times (maybe when you're moving SGs)
  5. Do NOT fall into the trap of predictablility yourself.

Commitment

Go, play, demonstrate by example what works. But I thought I'd point out that they aren't all neccesarily idiots for defending in the wrong places. They literally don't know better, have been taught that their way is better, and are sincerely mistaken.

If you REALLY want to teach them better, then defeat them continually until they get disheartened. Then, when they finally start moving it about, give them some slack (the surprise value will help) And when they move it to the 'right' place? Drop dead. Don't learn to defeat it, don't compensate, don't play to the fullest, let it tear you to shreds. Again and again and again. Want to place a bet that they'll use that position more in the future?

Plus it makes a damn good excuse for why you're playing lousy that day...

Edited by Vitenka at 2001-07-02 13:01:25

Older News  
Main News>>>
[M]
[Haiku]

[Rifle]
[MORAT]
[Fa-Teen]

 
[Froody Comics]



[Nifty Links]
[Editorials]
[Guilt Box]
You owe:
00:03
 
['Tenkas Tips]

This HTML design by Vitenka
I'm aware it sucks, but am also too lazy.
Note that now you've changed all the colours, it could be about anything!
BTW - this site looks fine in IE5 and netscape4. If you have a problem, if no one else can help, and if you can find it, maybe you can use - Netscape 2.