Hello! I wrote patch, which disable Random Number Generator

[Rya]

...I think such a mechanic makes the game boring. A unit with a 10-2 attack is suddenly exactly the same as a unit with a 5-4 attack.

Not exactly the same if a unit dies.

[multilis]

"I think such a mechanic makes the game boring"

I have played many many hours of single player game c-evo, a civilization variant with *no* luck (and no cheating help to ai players, except more resources at impossible level). Enjoyed not having some of the games ruined by extreme good or bad luck, not at all boring.

[Velensk]

Like I mentioned in the other thread: the effects of having no luck would be very diffrent in civ than they would be in Wesnoth.

In Wesnoth handleing all the possibilities is a major part of the game and its stratagy. In Civ not quite so much. Instead Civ focuses on empire management. If combat was deterministic in Civ I would think more during the game to try to evaluate the combat correctly for ease of making decisions. On the other hand, if Wesnoths combat became deterministic then I would merely have to calculate for one set of combat results. I would know exactly how much damage they could do to me and vice versa. This would make it so that it would never be a question of "Is it worth risking this?" or "Can I hold that point/Is it worth trying to hold that point" and otherwise greatly reducing the variety of the gameplay.

[Dave]

+1 Velensk.

An empire building game with deterministic combat is a completely different proposition to a tactical combat game with deterministic combat.

David

[Kalajel]

My suggestion for something interesting to try would be to make it so units have guaranteed hits but variable damage. Make it so a 5-4 unit does 4d5 damage, and a 10-2 unit does 2d10 damage. This will keep a good deal of randomness in the game, but mitigate it significantly.

And then we'll get threads full of people [censored] about damage luck in Wesnoth…

There's just no pleasing some people…

[Rya]

The only random generator modification I would like to see would be one that manages to balance extremely bad luck or extremely good luck. Like if some has dealt -50% expected damage and received +50% expected damage in total he should be a bit more lucky in the following turns. The random damage and chances are good because of their gameplay aspects, but it's not really 'good' if someone can beat someone playing far better than him just due to luck.
Ideally at the end of each match the statistics should show numbers close to "0%".

Also on a sidenote, doesn't Civilization also have random factors? Like how fast the development goes and so on?

[Velensk]

No. The development in civilization goes at a deterministic rate, population growth is determined by food and construction is determined by minerals.

Now there are many random factors involved with the creation of the map and in the basic game combat is run with dice. Random events can also occur if you don't turn them off but the basics of how your cities grow/develop is entirely calculable.

[Caphriel]

The only random generator modification I would like to see would be one that manages to balance extremely bad luck or extremely good luck. Like if some has dealt -50% expected damage and received +50% expected damage in total he should be a bit more lucky in the following turns. The random damage and chances are good because of their gameplay aspects, but it's not really 'good' if someone can beat someone playing far better than him just due to luck.
Ideally at the end of each match the statistics should show numbers close to "0%".

That doesn't really work out, because if you get good luck early and kill units that would have survived or wound units badly enough to force them to retreat, it reduces your opponent's potential expected damage, so even if they then get the same amount over their expected damage next turn, it's less damage than it would have been. So even though the RNG swing may balance out over the course of the game, its effects won't. For the effects to balance, the RNG swing has to approximately reverse itself. This is one of the reasons early aggression is so dangerous. With few units on the field, an RNG swing resulting in an unexpected death can have a large effect on the outcome of the game. Then, of course, there's the detail of "luck in critical moments." Things like missing/hitting poison or slow attacks aren't reflected in the stats, nor is whiffing a surprising number of times and failing to kill a unit that is about to level, or getting lucky and killing one. The stats don't reflect the effects of breaking/failing to break the enemy line on your attack. So in all, you can't really say that just because the overall RNG swing for a game was 0%, that the effects of the RNG on the game did not provide one player an advantage.

(Please note that this isn't a criticism of the RNG, it's just a comment on why the stats summary doesn't accurately reflect the impact of the RNG )

[Sauron]

Yes, always succeed. I think now, how to avoid that correct without RNG.

Simple solution, keep using the RNG for only these two weapon specials. IMO, its a fair compromise between using and not using the RNG, and doesn't require rebalancing.

That is, keep doing both calculations. Calculate damage like done in your patch, but keep the old RNG code around and use it to calculate whether poison/slow hits or misses.

It will probably require a bit more of code tweaking, since poison/slow are currently triggered directly by hits.

This issue was adressed and SOLVED long ago in my mod, refer to my sourcecode (see my signature).
Instead to monkeywrenching zounds of mods - mabye time to organize, discuss and focus on one acceptable solution?

[Gambit]

Well it's good to see that at least people are putting their own code up now, and not acting like all the hordes before them who seemed to think that if they just kept asking, they'd win eventually.

[Sangel]

Perhaps we can add a link to Sauron's patch in the Wiki and FAQs regarding luck? While some people complain about luck without doing any research on the issue, it would be good if those who showed a serious interest could find out about practical ways to try Wesnoth with a lower luck threshold.

[booner]

Realizing that there are different styles of play; this is my opinion only and not meant to flame those thinking differently.

For me the random factor is one of the lures to wesnoth.

Welcome to Casino Wesnoth! Please place your bets!

More fun that roulette! Cheaper too!

I play relying on luck. To be able to summon my luck. Believing in the power to encourage the units to greatness; something not scientifically nor mathematically prove-able. Its like having the pretty lady at the table blow on the dice.

There are those that play straight by the stats. That bores me. There is the gambler's thrill in playing with luck in mind. All great generals took chances. Those who did not have luck are not considered great by history. Those that did and won have been immortalized.

To remove the randomness from Wesnoth is like removing critical rolls from rpg in my opinion and turning the thrill of play into mathematical research.

[Sapient]

If you want "Casino Wesnoth", try Royal Rumble maps or some of the other FFA maps.

There is still a fair amount of strategy involved, though, so it may disappoint you if you are looking for a pure dice toss.