Alternate battle system

[LPhillips]

I'm quite tired of seeing battles of full stacks with 90% or better chance of victory get ridiculously slaughtered. I don't think it adds anything to the game at all; I just find it to be frustrating when on the receiving end of this RNG ***kicking, and disconcerting when on the lucky end. I hate to have my glory stolen by some crappy roll of the dice that gives me a superhero where my opponent's superior play should have gotten me skull****ed. The thing pushing me over the edge is that I should be on the last few turns of my Middle Earth game, which is an unashamed 2v1 (they're practically using each others' cities), BUT I've lost 3 very powerful hero stacks over the past several turns with 75%, 64%, and 90% chances of success, respectively. So now it's going to turn into a prolonged near-stalemate unless I perhaps just burn all their crap to the ground.



Is there some possibility of a "tournament rules" type setting, where I won't see this happen multiple times a game? It would be nice to have a fairly predictable outcome. I'm not any more pleased when I get an unfair advantage because someone attacks me with superior odds and turns my hero into a superhero due to absolutely ridiculous rolls.

The great tournament was played with a battle engine that gave 100% predictable results. It calculated the outcome exactly as it should be, and produced precise results. It left huge opportunity for strategy. Rearranging your battle order could change the outcome dramatically, and all of the same elements for strategic play were present that we see now. I'm not saying that is necessarily the way Warlords/Warbarons ought to be played, nor that all elements of dice should be removed from Warbarons, only that I'd like to see the option to turn on results that are never highly unstable and unreasonable.

[KGB]

LPhillips,

Whats the great tournament? Never heard of it.

Also starting in Beta3, the battle system was already changed to give more predictable results. All outcomes of <10% were dropped to prevent crazy 1% battles from occurring. So once you exceed 90% chance of victory you can't lose. The only variable is how many men you'll lose. So if you are seriously over 90% success chance and lose a battle that's a bug that should be reported (I saw one once early that Piranha fixed a long time ago but have not seen it since in dozens of games).

Incidentally, 64% and 75% isn't very high chances of success. You'll lose 1/3 and 1/4 such battles respectively. In a long game repeatedly testing those odds like that is going to cause you to experience those losses. I seldom commit my heroes unless I am over 90% (and thus guaranteed to win) and cross my fingers anytime I'm under 80%. Even at 90% chance of success, fighting 8 such battles during the game gives you only a 43% (.9*.9*.9...*.9=.43) chance of winning all 8. Which is why the longer the game the more likely the odds catch up to you.

I've personally found the battle system in Beta3 to be vastly superior to Beta2 in terms of producing predictable results (for example in Beta2 you could fight 8 Lt Inf vs 8 Lt Inf and win with 7 men left. That can't happen now so the result will always now be 1-3 men left for the victor).

KGB

[LPhillips]

I'm glad to hear that. Maybe that information should be made more readily available.

Yes, 64% was unintentional. The last battle was 89.3%. Average outcome would have left me with 4-5 units remaining, judging by my experience and the particular circumstances. It's important to remember that the computers' odds are often very far off from actual mean outcomes. I don't mind seeing 75% battles where the victor has only a unit or two left; in fact I expect that. It's when the cost is a full, well-balanced stack including a very good hero, and the reward for your opponent (having nothing to do with his planning or skill) is a bucketload of experience or simply a crucial victory where he should never have one, that I get irritated. Hell, I don't see why anyone should ever win a stunning victory when they have a 75% chance of a total ***kicking. Inflict high casualties perhaps, but never stomp a hero, pile of heavy cavalry, dragon, archon, and +2 command -1 stack hero with some heavy infantry and an elephant.

The best reference I can find to the tournament is in the Warlords 2 Wikipedia article. Even a few years ago, a great deal of information was still available online, including the exact workings of the "Warbot" system and several strategies for battling with it. There were two resource-rich personal fansites including full documentation of the tournament and its results, and I know for a fact the website once housing that archived information still exists (although all links to tournament info are dead as it was over a decade ago and no one cares about the results anymore).

[piranha]

I'm open to hear different battle systems.

One thing that could be done that I thought about adding for beta 4 is increasing the 90% rule when the armies get bigger. If the combined STR value go over a certain level it could be 85%, and then 80%.
I'm not sure about going further than 80%. Might turn out that you can attack and be certain to win too often.

[kenc80]

youre probably talking about that big tournament in like 98 that heeter put on?

im with KGB on this one, the battle system is vastly improved and im ok with it. At the end of the day LP we should want a little randomness. If its 100% predictable doesnt that get a little boring? There needs to be a little randomness every once in a while. Maybe an upset here or there. Wouldnt the NCAA tournament be awful boring if every year all the 1 seeds advanced because they were expected to win?

[LPhillips]

Ah, I like the excitement. I do. I just hate to see the outcomes be entirely bull****. A whole game can hinge on your very well-leveled hero and carefully arranged stack being wiped out by something that should at max have inflicted 6 casualties on a very, very bad day. I'd like to see some restrictions on casualties in battles from 75-90%. Maybe you always win, but your casualties can range up to 7 out of 8.

The Warbot system is actually incredibly interesting. I really would like to know where the info is on that; maybe we should ask the SSG main site?

The NCAA tournament is made interesting because of individual human striving. A more accurate comparison would be if you played your heart out, sank 10 baskets, then had to roll a 10-sided die with the possibility that 1/10 of the time you get 0 of the points you earned.

Making the results of a strategy game reasonably predictable does nothing except reward better play.

I don't get any more enjoyment out of a ridiculous random outcome than I get playing roguelikes where one roll of the RNG wipes out your 50 hours of playtime without any input from your intellect or skill. It's just pointlessly wrong. I understand the need for risk and danger. That's quite obvious. I understand the value of the fact that you can't conquer 10 neutrals with 1 Griffin because of the dice system. I don't understand any desire at all to see the major conflicts of the game decided by sheer luck.

[KGB]

LPhillips,

I don't think SSG knows anything about the WarBot system. The tourney wasn't theirs and I've never heard of the system so I doubt they created that either. It was likely done by someone who created the tourney.

And casualties *are* limited with the Beta3 change to the battle system. That was one of the things I wanted and Piranha did. You should go back through the older posts on it and read what was done (it's a long post so should be easy to find). Basically the end 10% of results are discarded for both sides. So only the middle 80% are used. This eliminates all the extreme outcomes you (me and everyone else) hates so much. This is why now once you reach 90+% success rate you can't lose because that last 10% is no longer allowed. Conversely if your odds of success are <10% you can't win.

Sometimes you will take more casualties than average. Other times less. But you won't get extreme results anymore. So anything you are seeing is within the middle 80% of likelihood of occurrences. If you think you are seeing something outside that middle 80% can you post the exact units involved on both sides so it can be confirmed (Both Piranha and I have combat simulators to check) whether its OK or not.

Piranha,

At the moment I think the system is fine. I would not lower the threshold to 85% as it would make it even easier to guarantee victory in a battle and thus less risky for heroes. 90% is a good cut off point because it's 9/10 chances of success and it can be quite hard to achieve between stronger stacks as it should be.

KGB

[LPhillips]

The cutoff has to be somewhere, so I can't complain too much. 89.3% turning into a total wipeout really seems like crap, either way. As stated: I'm glad there is a cutoff for potential BS. I'd like to see BS not play much/any role in major battles.

The Warbot system was described in length from a link to the SSG site. I believe the link led to a fansite. I read the link years ago, and I know the basics, but as to the origin that's all I remember. Battles could be much more strategically planned, as you would stack your crow behind your Dreadlord when attacking two light infantry, for instance, or match your units such that your heavy cavalry would die and expose your wizards exactly as you penetrated to the enemy demons. However, the system is highly flawed in the sense that you can conquer infinite enemies under the right circumstances with no losses. That flaw only comes into play in specific circumstances (notably capturing neutral cities).

I won't crusade on this any longer. But total losses on a battle you would win with 3-5 units remaining about 50% of trials and 1-3 units remaining 35% (or 39.3% in my case) of trials will always be ridiculous.

Addendum/Edit:
Remember that I asked for an optional alternate battle system, not a change to the current one. Something much more stable and reliable for players who want it, and possibly employing it in competitive ladder games.

Also, you're telling me that I shouldn't see battle outcomes of 6.2%? Because I'll pop some screenshots if that is the case.

[piranha]

Its not this page you mean: http://www.heeter.net/warlords/ ?

You will see results like 6.2%, but you should not see someone win with 6.2% chance to win.

The battle is run 1000 times to find out what the chances are for each side to win. Then the real battle is run and if its less than 10% chance for one side to win, and they do win the real battle will be run again until the side with 10% to win loses.

This removes extreme outcomes, but it will also make you sure to win when you have enough odds on your side which also have a kind of impact on the game.

[KGB]

Piranha stated it correctly.

Think about 8 Lt Inf vs 8 lt Inf. There are 16 possible battle outcomes, 8 for each player. The likelihood of each one is something (just guessing, not at my PC to get exact odds) like:

Player A: (8 left) .0001, .001, .01, .04, .065, .08, .13, .17 (1 left)
Player B: (8 left) .0001, .001, .01, .04, .065, .08, .13, .17 (1 left)

Putting them together looks like a Bell curve with the most likely results (1 left) in the middle/high point).
.0001,.001,.01,.04,.065,.08,.13,.17,.17,.13,.08,.065,.04,.01,.001,.0001

What Warbarons did is drop the low 10% on each end. So the .0001,.001.01,.4 all get dropped on each end. Giving a middle 80% of 1-4 men remaining for each side as the potential outcomes.

Thus when you see your odds to win are 89.3% that means the sum of ALL your winning outcomes (from 1-8 men left assuming you had an 8 stack) is 89.3. So when you see a 6.5% for the battle it just means that particular outcome was 6.5% likely, not .001% likely.

You'll notice that even the MOST likely outcome is still only something on the order of 17% simply because there are so many possible outcomes in this battle.

KGB