Tactical Map Ratings

[Kraellin]

ok, i think it was magy who suggested it might be a good idea to rate maps according to their tactical advantages and disadvantages. his use for this, i believe, was for use in his online campaign stuff.

well, i got to tinkering with this idea a bit and at first glance this doesnt seem like a very difficult thing to do. i mean, you've got attackers and you've got defenders and one side may or may not have a tactical advantage due to terrain over the other side. pretty simple. just stick some numbers on those advantages and disadvantages and there ya go....HA!

it aint quite so easy. ok, you start with ironing board. that's easy enough. 0 for attackers and 0 for defenders, so 0/0. piece of cake. even totomi isnt that tough. no trees of any significance. no big hills or valleys and is still fairly close for both sides, with maybe a tiny advantage to defenders because of some of those hillocks and small valleys; so maybe 0/1, attacker to defender. or, if you'd rather use a smaller scale, maybe 0/.1. i prefer integers, so i was using a larger scale, say 0 to 100 or even 0 to 1000 and trying to keep away from negative numbers.

but, this starts to get more complex as you add more variables. stick some trees on ironing board, and depending on where you put them and how many, the odds start changing radically. and what if the attacker takes cav or missile or mostly infantry. what if it's a mongol game? will the weather change a map's rating? is that hill within a deployment zone or just outside of it? what if it's a 2v2 or 3v3 or 4v4 game, or even worse, a 1v3 or 2v4 or something like that? i know of that one map where it at first seems like a big advantage for the defender because he can deploy and sit on a nice big hill at the edge of the map, but when you look closer, there is a ravine that divides the armies in a 3v3 or 4v4 and this can play hell if the attackers all choose to attack one side or the other. and what about that oddball situation with the 4th army in a 4v4 where he is deployed in front of the other allied armies and off to one side? it gets messy on some maps. and what about the impassable zones and water and bridges and forts and reinforcement maps. some maps force reinforcements and that must be considered.

it's a tricky proposition. i could see endless debates on some of those maps. yamato is a good example. lots of trees, lots of hills, bowl shaped around the edges. a 4v4 puts the defender in a fairly bad position, but a 3v3 the defender is in better shape. it's a great ninja map. it's a good missile map. infantry can hide in trees and ambush cav, but cav have enough room to move around a bit and nail other units.

and all of this doesnt even take into account that some defenders attack and dont defend. i think we can just assume, for the sake of rating, that the defender is defending, but still, it's a question mark in some ways.

at the moment this is just a mental exercise i'm toying with. i'm not rating 4 or 5 hundred maps, but i thought i'd throw it out to the community and see what other's thoughts on the matter might be.

this is originally magy's idea, i believe; all i'm doing is exploring it a bit.

and as an aside, whatever happened to that site that was rating maps, not on the tactical aspects so much, but on the playability and interest and aesthetics and map making ability of the creators and so on?

K.


------------------
The only absolute is that there are no absolutes.

[MagyarKhans Cham]

in the online campaign we had small bonusses a clan could obtain to overcome a 5% or even 10% wronglyrated map.

u could use the battles database for recalculating the maprating. so lets say totomi starts with 100% and after each battle the defender wins 1% is substraracted from the next-battle-defenders-kokuallowance.

so after many many battles on totomi we all play it will probably swing around 95%.

Ofcourse this might be slightly wrong for other amount of players in multis but at least its better then nothing.
--------------------------------------------
an other possibility i like a lot is that u use the koku that aint being used from the allowance. so lets say its a 6000 vs 6000 koku 1 vs 1 battle then if one side uses 5950 koku and the other 5900 koku then the side with the lesser used koku is the defender.

lets brainstorm about this idea and the possibilities it might have. it means that u add a little gambling depending on the map which imo comes close to reality.

u have a defending army and reduces teh koku u spent a lot, unfortunately teh enemy uses even less koku and u have to attack with your defending army....

at least its fun.

perhaps the idea of this kokugambling can be used in a different way...

what u think krae?

[TosaInu]

Konnichiwa,

It's fine if one can set different koku amounts for attackers and defenders, I don't want any kind of fixed maprating or fixed kokuamounts. Like Kraellin said, I might be an offensive 'defender'.

------------------
Ja mata
Toda MizuTosaInu
Daimyo Takiyama Shi

http://www.takiyama.cjb.net

[Celtiberos Shimazu]

I am agree with you. A forum with the map name, and all users could post the account of defeats/victories. I think is better calculate all games, without take account of the number of player.
Simply,should be a informative matter. More later, with the stadistic of the maps, we could to talk about their use.

A great idea!

[Vanya]

GAH! Best rating is a statistical figure exclusively.

Show games won/lost by each stance per game type, where game types are 1v1, 2v2, 3v3, 4v4, 1v2... etc.

Then show losses per victorious/defeated party per stance.

(I do this in my log reader. The 'score' was my lame attempt to judge those numbers...)

So... lets say you looked at Totomi...

1V1s - ATTACKER
300 games played
109 won by attacker
191 won by defender

In victory, the attacker
killed 678 enemy soldiers on average
lost 311 soldiers on average
fielded 878 men on average

In Defeat, the attacker
lost 802 soldiers on average
killed 208 soldiers on average
fielded 960 men on average

2v2
...

Wouldn't this give you all you need to know about a map?

In fact, a basic extension of my logreader database could even give you the historical 'most effective' armies used per koku range! (A feature offered on the current version, which has never been distributed...)

So, you could show for 'Totomi', 6000 koku, 1v1s:

The most effective army in terms of kills/killed ration was:

8 WM, 8 AQ, fielded 21 times by Someplayer Akechi, between the dates of A to B.

The most effective army in terms of wins/losses was:
3 Na, 3 ND, 8 CA, 2 YC, fielded by Someplayer Akechi, Anotherplayer So, Somedumguy Date, and Yourmother Yokohama between the dates of D and C.

You could even include links to view each instance of the fabled armies.

Talk about intelligence! GAH!

(No wonder I haven't released it... http://www.totalwar.org/ubb/eek.gif )

On the score, the method I used tends to settle at 50% if all things are perfectly balanced. But it is flawed in that you cannot examine but your own scores with it, or the figure will be misleading.


[This message has been edited by Vanya (edited 04-22-2002).]

[This message has been edited by Vanya (edited 04-22-2002).]

[Kraellin]

vanya, et al,

i think that's exactly what's needed. i hadnt even thought about the log files. i'll bet each of us has quite a stack of information already accumulated about how various maps play. all we need is a coallation system to bring all these files together in one place and then just post the results per each map. that would immediately tell you what's what about a given map. nice.

there used to be an upload feature on the .org that allowed one to upload files and have them immediately available to anyone else that wanted them. tosa put this up at my request a while back. if it's still there one could simply zip up his log files and make them available to whomever wanted to coallate the information from all submitted log files and then spit out a finished file and post it in the same manner. i think that would be the most expedient method.

K.


------------------
The only absolute is that there are no absolutes.

[Vanya]

Ah so!

If...

The Org can host a local database, then my logreader code (or parts) could be migrated for use over the web. If its ASP we are talking about, the headache is negligible. If not... ouch.

But... I have over 1,500 logs in my personal database... its takes up about 26 MBs on my disk. So... if you don't have the space, it will be very painful.

Naturally, you could use a trimmed down version that cuts out some of the detail.

To start, you'd have to host the database and be able to import the data in, either by submitting logfiles individually or 'uploading' them. Reporting can be added once that is proven to work.

In fact, I could even add a feature to 'transmit' all the contents of your personal database to the central one at the Org. But that would be getting far ahead of myself...

It would be a starting point. Nice thing is, all this code is practically already written (minus the sync stuff, of course...)



[This message has been edited by Vanya (edited 04-22-2002).]

[Vanya]

Also... if you want me to, assemble all the files, and I can produce the reports if you want. I can then send you the HTML outputs.

Or you can do it. Use my logreader and it will make it easier on you. Otherwise, just import them and write your own custom 'report' tool that reads from the database...



[This message has been edited by Vanya (edited 04-22-2002).]

[MagyarKhans Cham]

I still like the old idea of different settings for def and att but somehow that is never supported, perhaps we can do a poll

[Vanya]

Quote Originally posted by MagyarKhans Cham:
I still like the old idea of different settings for def and att but somehow that is never supported, perhaps we can do a poll[/QUOTE]

The difficulty with different koku is what becomes fair... This opens the door to folks hosting a 99999 koku game, taking attacker and letting the defender stand with 1000 koku. Surely thats abuse.

Equal koku is always the fairest. Sure, terrain gives defenders an advantage at times, but as you so often point out, attackers enjoy other advantages.

You might say in a 10000 koku game, that defenders on Totomi deserve only 8000 koku. Others may say 9900. And those that always lose attacking might say just 4000 is enough. This lies at the very heart of the problem with unequal koku distributions. It become LESS fair overall than the status quo...

[Magyar Khan]

but Vanya, u would never join a 99999 vs 1000 koku game.

its like on the fishmarket, u join the game with the best koku for your side in comparison with the opponnent u gonna face. the nice thing is that u can lure players to your game by even giving them more koku then u get yourself.

so bot kokuvalues must be stated on the screen before u join, and the game is fastest filled for those who present the most koku.

that is the solution

[Vanya]

Thats like hosting a 10000 koku game and telling the foyer folks that you will defend and only use 6000 koku and not a penny more.

Sounds good... but who will join? I'd be skeptical as heck. My conspiradar would be telling me its a hoax... that you'd be saying you'll get 6000 so I'd pick in a silly fashion, only to face your 10K army http://www.totalwar.org/ubb/eek.gif NOT THAT *YOU* WOULD DO THAT, I must say... I'm just saying SOMEBODY would.

I agree that full disclosure is the only way that would work. But original shoggy never disclosed the time limit... http://www.totalwar.org/ubb/eek.gif

And my example was an extreme one to describe the dangers inherent in this system you propose. Sure its a gross exageration. Nobody in their right mind would knowingly join a 99999 v 1000 koku game. But the underlying principles are the same for 10000 v 9999 games.

[Cheetah]

Quote Originally posted by Vanya:
GAH! Best rating is a statistical figure exclusively.

Show games won/lost by each stance per game type, where game types are 1v1, 2v2, 3v3, 4v4, 1v2... etc.

[/QUOTE]

GAH! Nice idea, but ... http://www.totalwar.org/ubb/smile.gif This can be misleading esp. in case of 3v3 or 4v4 games where teamwork counts more then terrain. One might say that good players are equally likely to be in the defending as well as in the attacking team, however, I think this is not the case. Experienced players usually attack (cos they think it is a bigger challange) and unexperienced players tend to defend (cos they think it is easier). (Of course both assumption is wrong http://www.totalwar.org/ubb/biggrin.gif )

Take Wakasa for example. In most of the games I played on Wakasa as 4v4 the defenders lost. But not because Wakasa is unfavourable to defend. The usual pattern is that either the front left defender deploys on the front ridge (usually facing the forest!!! http://www.totalwar.org/ubb/eek.gif http://www.totalwar.org/ubb/eek.gif ) and gets rushed in a split second, or the centre defender deploys in the forest and gets flanked from both side. However, if you can gather all the defenders on the small hill on the left, then you have a very favourable defensive position. So, IMO Wakasa is a deceptive map for unexperinced players but in fact it is a quite favourable map to defend.

[Vanya]

GAH! Hungarian! GAH!

All you can hope is to show the stats. Leave interpretation to the people.

What you say is true. Stats don't show tactics. None. But stats show performance on that day. You can break them down by army too.

[Cheetah]

GAH! Vanya! GAH!

How can I leave the interpretation to a headless horseman??? http://www.totalwar.org/ubb/tongue.gif

[Vanya]

[Farts]
[Foliage and animals nearby drop dead within minutes.]
[Soon, all around lays dead...]

Wisdom emmanates from my every... pore.

http://www.totalwar.org/ubb/eek.gif

Here is a tender morsel for the faithfull...
http://www.airpower.maxwell.af.mil/airchronicles/apj/apj89/boyles.html

[Sir Kuma of The Org]

OH god, you smell terrible!!!!

You know in most countries biological weapons are banned....

------------------
Ils sont grands seulement parce que nous sommes à genoux. Alors levons-nous debout!!!

[Cheetah]

*collects dead squirrels, dead birds, dead lizards, dead horse*

*burries all the animals*

*finds an ugly head, kicks in it*

*mourns*

*erects memorial*

"HEREIN LIE THE VICTIMS OF WISDOM"

http://www.totalwar.org/ubb/tongue.gif

BTW, I am very much interested in the stats and I am ready to help you (despite of your mass killing of animals http://www.totalwar.org/ubb/biggrin.gif ).

[Vanya]

Quote Originally posted by Cheetah:
http://www.totalwar.org/ubb/biggrin.gif[/img] ).[/QUOTE]

Goody. Start collecting log files from people. Zip them up and email them to me. Don't send a 100 MB whopper 'cause my yahoo account will reject it (I think)...