battle calculator

[LPhillips]

Hello all.
My opinion is that there is some value to the mystery of a slightly misunderstood, somewhat randomized battle outcome. Experienced players have no trouble estimating quite clearly what outcome to expect from battles where enemy strength has been probed, but there's still the edge-of-your-seat sense of risk and reward involved when you run the game on a tight line, aggressively pushing the probabilities with risky strategies and riskier battles. This may seem a bit ambiguous, but there's a certain desirable quality to a mysterious opponent. So let us remember the cost to that aspect of the game if you reveal battle outcomes without the need to fight them.

KGB,
I think that a more entertaining and less perplexing solution for ladder games would be allowing players to see the final strengths of the units. Flat numbers, with no indication of the bonuses and such involved. Why are those Wizards at 65? Is it Leadership? Morale? What do I counter it with?
This could add immeasurable value to allowing heroes access to varied bonuses, if the game goes by that route. For example: it's downright entertaining to slam into a hero with 35-strength units and watch him Ambush the heck out of your stack without being able to predict it (short of hitting it with 8 crows, of course).

[KGB]

LPhillips,

As you noted, showing the final numbers only makes sense if you allow heroes to have more than 1 bonus to upgrade. Right now the game is so static it needs to hide all the numbers to keep players from getting the exact values an opponent has in terms of heroes, items, blessings etc.

KGB

[hungrytales]

Damn, how did I miss this thread?

Strach,

I'm totally with you on this. I think your points were clear-cut and rational, and those of KBG and LPhillips kinda vague and mystical. I don't buy all this talk about player experience. How exactly can you calculate battle odds with your experience? You can't. Actually what you can achieve with the current system is a bad estimate based on educated hunch (and then a lot of frustration when your educated hunch-based estimate fails). To add insult to injury now, when the game is still in beta state, you find yourself in the dark every time units get stats overhaul and you see your 'experience' tossed overboard.

I read, what KGB has done here eristically, is he changed the subject from "do we need a battle calculator?" to "do we need the game to tell us too much about an opponent". Quite clever . These aren't the same questions though. I, as strach, see clearly the need for the calculator (as it happens I even wrote one for myself - more on that later), and I also agree with KGB and LPhillips that the game shouldn't necessarily reveal too much. The answer to the problem KGB so cleverly put up here to play an impostor role is the game should have additional mechanics for revealing or not information about an opponent. In short words - it would be nice if you actually are able to know almost everything about your opponent (and then use it for your sinister calculations), but ONLY at a certain (strategical) COST.

Now, heading back to the proper subject - do we need a battle calculator? Well, strach needs one, I need one and KGB obviusly also needed one at some point as he openly admitted he made one for himself. So much for calculating odds with your experience . As a matter of fact what propped me to do the same (apart from too frequently missing my educated hunches in games ) was this impression I got reading the forum between the lines that there are people who already done it (and keeping silent about it ). So strach and KGB needs it, I need it, and what about the rest? Well...yhm...well...

Isn't it like that? Come on, in another thread we just settled on the fact that people willingly cut A LOT of their fun for the edge in winning. I tell you - I'm too old not to know that there are many of those who would as willingly cut a lot of their honesty for the same reason.

So yes, strach. I think you're right - battle calculator in this or other form is inevitable and for all of us to be fair and square it should be available for everyone. I'm not a programmer, only a hobbyst, and my 'development' stalled recently out of total lack of free time, but yes, I dusted off my rusty java and thought it through and right now have a version where anybody able to add a few java code lines in one class can run a proper simulation and read a nice summary. I'm planning on releasing the thing just after I scramble together some gui (which means more or less: not that soon).

[Moonknight]

I really doubt KGB, or anyone else for that matter, are checking a battle calculator for every battle.

I also think that anyone who has the capability (and the spare time) to build a battle calculator, not to mention alter it each time the units change or a new unit is added, deserves to have an advantage.

I think a battle calculator that would provide 1 vs. 1 probabilities would be suitable, but anything beyond that is taking away from the game.

And I do think with experience, you can guesstimate the probabilities for about 90% of your battles, at least in terms of i'm likely to always win this battle, i will usually win this battle, this battle is a 50/50, i will usually lose this battle, and i'd never win this battle (90%/70%/50%/30%/10%).

The only time i would use a battle calculator if it were available, would be for end-game type of stacks (super hero stack versus another super hero stack).

[KGB]

HungryTales,

The calculator I created was originally designed to check that Piranha's code did not have a bug. I simply would use mine to verify that his numbers were accurate for lots of different battles and bonus's etc. This helped debug all the battle results. It wasn't created to be used as a way to accurately model every battle which is why it never had a GUI and why I still have to manually work out all the bonus's for all the units when I want to run it with anything but trivial battles.

My argument was, and is, that players do not need exact percentages for every battle in order to play. No prior incarnation of the game provided anything like that. I also rightly pointed out that having such a calculator TOTALLY INVALIDATES fight ordering. In other words, if such a calculator becomes available, Fight Ordering will have to be taken out of the game. Why? The answer is that a calculator can figure out exactly what order the attacker will get his best odds of winning which is GROSSLY unfair to the defender and totally defeats the idea of fighter order.

Since you are determined to write a GUI calculator and make it available I would suggest that Piranha makes the following change:

Instead of every battle using dice of 100, it selects a random number between 80-200 (my simulator shows that range varies things by a few percent but doesn't wildly change outcomes). Then the battle is played out with that as the maximum dice roll. This will 100% invalidate any calculator since it will be impossible to know what number the max dice roll will be for any battle and thus the actual odds of winning and losing any battle will vary by 2-10% depending on the number of units involved. For post battle checks for bugs, the game can show what number was selected as the maximum.

It's a simple and fair change for everyone without access to a calculator. It also works with the idea that no general ever knows the exact true odds of any battle.


It may also be necessary to tinker with how fight ordering is done as well . But that can wait at least one more Beta until it proves absolutely necessary if players are spending too much time working out optimum fight ordering.

KGB

[Fordus]

As long as detailed and up to date description of battle mechanics is available, there also has to be in-game battle calculator.

For the simple reason that someone can and almost certainly will make good calculator for himself or for limited distribution. That will give advantage to whoever uses it. Advantage might be minor for most cases, but quite annoying. Imagine finding out after few ladder seasons that one player/team has used calculator all the time...

[hungrytales]

Fordus,

My point exactly. The thing is, 'this detailed and up to date description of battle mechanics' is so far not available, at least to (the limited extent of) my knowledge. My 2 attempts at learning the exact phases of battle calculations (what comes first, when caps are applied) on the forum went answerless. Now I have a much better understanding why

But all of this is wrong, isn't it? By any known definitions rules are what constitutes a game and by definition they should be clear and available to players. And the true reason I felt like I needed a calculator was I wanted to understand the game better. That's the core of this and I sincerely don't think there's any harm in wanting to understand the game better (it's probably one of the main reasons for having this forum in the first place :>).

I see the greatest reservation one can have for having a battle calculator in the current state of things (game in beta, calculator as a separate program) is that it drags gameplay towards hardcore and afar from casual. It's something I imagine piranha and Snottling would see as not too desirable as they probably want to keep the game as approachable to as many people as it's only possible considering the times and the genre .
From their perspective, if the community developes itself a calculator and it's use becomes common, then new players learn that if you want to stay competitive playing our game you have to download this additional thingy, learn to use it, do this and this and so on. Probably doesn't sound like a lot of fun to most

So I understand the reservations but I see the approach of withholding some crucial info about how the game works to hinder possible development as... kinda lame . I hope a better formula is worked out. Preferably involving sth in the shape of what we discussed earlier - and additional game mechanic of veiling/revealing info about an opponent and his armies, bonuses and such (e.g. with the introduction of some new scouting units, or just some new behaviour of current scouts, or maybe buildings?). So the advantage of knowing odds better than your opponent comes not only from: a) having better experience, b) having a calculator and sacrificing time to use it, but also and foremost from c) making strategical decisions in-game.

KGB,

as to your last argument, well, aren't you now arguing against your own points made in the discussion about the 90% rule? (I don't have time to check it, but I'm fairly sure it's the case ) . Also, I don't see how adding more randomness to mechanics invalidates figuring out a better fight order with a calculator but maybe it's only my meager maths skills. I thought choosing a battle dice doesn't matter much in the general scheme of things. Warlords used (if I remember it right) d10 and Warbarons use d100, which mainly enables better fine-tuning the balance (like with the recent Eagle's strength change from 10 to 13 and anti-air to 17 - impossible with the Warlords system). It probably elongates the process too (giving the authors more options to tinker with) but that's another story... . The only thing I can come up with that could have some slight influence on counting the odds is rounding the fractions when changing dice, but it should be marginal.

[KGB]

HungryTales,

My point exactly. The thing is, 'this detailed and up to date description of battle mechanics' is so far not available, at least to (the limited extent of) my knowledge. My 2 attempts at learning the exact phases of battle calculations (what comes first, when caps are applied) on the forum went answerless. Now I have a much better understanding why

Actually it's been fully answered already in the forum. I recall your post asking about it but I assumed you eventually did a search and found the thread where Piranha laid out the how bonus's worked and the exact order they are calculated, the min/max values etc. It was in a very long thread where LickKing and I were discussing bonus's and how they worked. It shouldn't take long to find.

There are several other threads that describe the basic mechanics of simple combat between units where no bonus's are involved. So nothing is 'hidden' at all although it does require a search to find exact details if one is looking for them.

By any known definitions rules are what constitutes a game and by definition they should be clear and available to players. And the true reason I felt like I needed a calculator was I wanted to understand the game better. That's the core of this and I sincerely don't think there's any harm in wanting to understand the game better (it's probably one of the main reasons for having this forum in the first place :>).

I agree 100% that all the rules should be clearly available to players. Eventually someone needs to write a detailed description of the combat mechanics with some example. Piranha/Snotling are far too busy to do it. I suspect a few of the long time players like myself, LichKing and a couple of others could do it but believe me, it takes a lot of time (1-2 hours) to write it all up and fact check it. Usually asking questions about a specific topic of interest in the battle mechanics will get someone to reply but asking for the overall rules is going to be too much for anyone to do in one post.

I see the greatest reservation one can have for having a battle calculator in the current state of things (game in beta, calculator as a separate program) is that it drags gameplay towards hardcore and afar from casual. It's something I imagine piranha and Snottling would see as not too desirable as they probably want to keep the game as approachable to as many people as it's only possible considering the times and the genre .
From their perspective, if the community developes itself a calculator and it's use becomes common, then new players learn that if you want to stay competitive playing our game you have to download this additional thingy, learn to use it, do this and this and so on. Probably doesn't sound like a lot of fun to most

Actually I think it's the other way around. The Battle Calculator takes a lot *away* from the game. For example, a long time ago I used to play tournament chess. In tournament games it was obviously strictly forbidden to consult books or other players during games for help with your moves. These days I like to go to my local Casino and play hold-em Poker. If someone showed up at a Hold-em Table with a poker book that contained all the odds for the hands or had a computer program that listed those odds and advised you of the bets to make you'd be asked to leave. That is in essence what the Battle Calculator does. It takes away all the learning of the rules of the game and instead says, 'here you go, here is the exact odds of winning and what the best fight order is to use and your losses of units'. So I think it's absurd to allow a battle calculator in ladder games and even in casual games it really has no place.

On the other hand, a battle calculator that shows odds in solo games against the AI to help players learn the game. I'm all for that. That helps newbie players learn the mechanics of combat, the odds, how fight order matters etc. But once you start playing even casual games against other humans I think there is no place for the calculator because part of the game at that point is to make decisions on your own with your knowledge of the game. Otherwise players will never get any better at the game. In Ladder Games especially, everything should be hidden so that nothing should be available about your opponents armies.

Preferably involving the shape of what we discussed earlier - and additional game mechanic of veiling/revealing info about an opponent and his armies, bonuses and such (e.g. with the introduction of some new scouting units, or just some new behaviour of current scouts, or maybe buildings?). So the advantage of knowing odds better than your opponent comes not only from: a) having better experience, b) having a calculator and sacrificing time to use it, but also and foremost from c) making strategical decisions in-game.

Exactly. I couldn't have said it better. I think most of the numbers in the battle screen should be hidden in human player games. In casual games I can see showing the final bonus numbers for each stack in Morale/Fear/Leadership/Chaos but that's it, no individual unit strengths for your opponent, no ambush chances etc. In ladder games I wouldn't even show the bonus numbers numbers.

This is also why I put in the wish list the ability to buy production with extra strength because with hidden numbers no one would know when you bought better units and when you didn't and so no calculator could work out the odds.

as to your last argument, well, aren't you now arguing against your own points made in the discussion about the 90% rule? (I don't have time to check it, but I'm fairly sure it's the case ).

Not quite. The 90% rule is applied to the final combat result to remove the 10% extreme results. It doesn't matter what the actual odds of winning and losing are because there is *always* 10% extreme results to remove. In other words whether your chances of winning are 30% or 90%, 10% is getting removed at either end leaving only the middle 80% left.

Also, I don't see how adding more randomness to mechanics invalidates figuring out a better fight order with a calculator but maybe it's only my meager maths skills.

It doesn't do much to invalidate fight ordering. It only invalidates calculating the actual outcome of the battle. In a round about way though it does help with fight ordering because it may be that putting your Wizard 5th instead of 4th allows him to face the enemy Gryffon instead of a land unit. With the extra randomness you won't quite be sure if the optimum place to put him was 4th or 5th.

I thought choosing a battle dice doesn't matter much in the general scheme of things. Warlords used (if I remember it right) d10 and Warbarons use d100

Warlords 2/3 actually used D20. Warlords I used D10.

The dice do matter. Higher maximum values favors quantity over quality while lower numbers favors quality over quantity. So when you get 200, lower strength units will fare better against stronger units and when you get 80 the higher strength units will fare better against weaker ones. The range of 80-200 changes the overall outcome of battle by anywhere from 0-10%. So you might be 70% likely to win at D80 and only 60% likely at D200. In other battles that range might be something as low as 1% and matter not at all.

if you want a clearer reason why that makes sense, imagine a 20 strength unit facing a 25 strength unit. If the dice roll were instead D1000000 (1 million) the odds of either unit hitting are so small that they are basically equal units so the 20 strength unit has a better chance to win. Now imagine the dice roll was instead D26. Now the strength 25 unit can almost never miss and so will almost always beat the 20 strength unit.

KGB

[Fordus]

KGB misses the point... As long as it's possible to know battle mechanics and individual battle strenghts, calculators cannot be disallowed. You can try to prevent their use, but it'll probably fail. So if you want to have level playingfield, someone has to provide calculator for everyone. In-game calculator would be best, but in-game link would probably suffice. Publishing calculator in forums would only further disadvantage people who don't read forums.

As for your poker example, online players have programs to calculate exact odds. Some sites try to prevent it, but it's easily circumvented by smart phones, secondary computers or some other means.

Edit: Hiding bonuses won't make calculator obsolete or unusable, it would just require bit more assumptions based on experience.

[Maze]

"Good"(*) news! Battle calculator will become available (probably) this week with the release of Warbarons 0.8!

(*) "Good" news, between "", as it will be for gold members only. /smiley:pokerface/