Units preview. Why it should be back...

[Iluvalar]

Hello, I already told that elsewhere but now that I tested more carefully, i'm sure of what I am saying.
----------------------------------------------------------------------------------
This topic in a bottle (edited after because my ideas made more clear):
As soon as the opponent make a choice against you (specially in the lab). An hidden timer start in the game. Your chance to win start to decay. That's why you need to react. That decay is hardcoded in the stats of the game and in the mechanic of it. For the game to be intense and short enough that decay is pretty fast. If you neglect the fact that the opponent is sending borgs vs your cannons, you will definitively lose in around 7 minutes after the production start. You will still be alive the next 7 minutes, but with close to no chances left. The more you react fast, the more your chances stay good. That's what make the game, we will all agree with that...

Lately, we removed the units preview from the intelligence display. The idea (that i mostly agree upon) was to promote scouting and also hiding as part of the game. However, we didnt changed the mechanic to give players the time to scout ! The players chances still start to decay as soon as the opponent production is started. This remove control from the players that can't possibly take an action before having the information. This skew the game to turn it from a mostly skill game to a mostly random game mostly determined by the choices we made before we have any information.
----------------------------------------------------------------------------------

Knowing the enemy design affect deeply what we research in this game. Having a good design; the good propulsion at the good moment is simply as important as tactic on the field. Let's decode the game a little...

The standard weapon modifier is ±30% there is 2 of them one for the weapon mg and the other for your propulsion. Therefore there is a possibility to be 69% stronger or weaker because of bad research lines. That's how the game goes in the research lab, theorically, there is a point where you'll want to exchange your actual line for another one that will suit your opponent tactic. If you do it correclty you will normally change a bad weapon modifier, to a good weapon modifier (whatever it's the propulsion or the weapons). Let's say the mean gain would be 50% of your strength. However to do that, you need to leave your actual research to change for the new one. Thx to the linear progression, the new one will go faster, so even if you research advantage is over 50%, after sometime of transition, you'll get stronger by changing your weapon.

I'd like to explain you the extreme math equations behind that, but i'll show you only the final graph. The result is clear, if you take 9 minute to change your research instead of 6, you will suffer 3 time more because it's a 3rd degree function ((9/6)^3). And when we talk about a player having a clear advantage on another, 3 time more damage is never a good thing .

I'm afraid that when the initial research path you chose is weak against your opponent choice, you will simply lose without any further decision or skill involved. In any case, the unit preview removal was a serious balance change that will triple the importance of the impact provoked by the initial research choices. This will seriously improve the randomness of the game.

[ODDity]

One word. Scout.

[stiv]

I'd like to explain you the extreme math equations behind that, but i'll show you only the final graph

Without the math & data behind it, that graph is just a pretty picture.

[Iluvalar]

Without the math & data behind it, that graph is just a pretty picture.

I didn't pushed the thing up to the real data. But i can describe the logic like this : Each minute where you ignore or neglect that you need to change your research will cause 2 minutes during which you'll be weaker than your opponent. That look linear as well, so not that bad BUT the "height" of your weakness change over time. The quantity of your weakness is a second degree parabola. You end up suffering more during a longer period of time. It end up being a third degree parabola. timeToReact^3 . When you play cannons and you discover that you opponent is concentrating of borgs since the beginning, you really need to react as soon as possible to have a chance to win the game. The game balance is too much fine tuned to add 3 more minutes in delay without causing it to collapse...

One word. Scout.

I assume you are informed about the enemy units each 3 minutes in my graph, I dont believe you could do better. Unless you want to make is the main emphasis of the game. Even i you were cutting that to 2 minutes, you would still double the quantity off initial luck and randomness in the game.

[Reg312]

i disagree.
1) why only math equations should decide what good and what bad? i mean. if something is funny and liked by people then it is good!
Math equations don't cover people's psychology. When you cannot see enemy's last produced unit is more funny! it gives randomness in game, every time you should try to prognose enemy tactics.
2) with units preview you are able to see when enemy preparing rush. its not good
3) you see how fast your enemy in first minutes and you can see if player good or not
4) with units preview you are able to see what progress has your enemy in research
so if enemy doing research rush and started strong units its signal for attack righ now and not later
5) Randomness good because even strong player can lose
when strong pro player lost what we got:
- noob winner have fun
- pro player have fun to make revenge

so my conlcusion: unit preview should not be back

[ODDity]

What if you're trying to outfox the enemy, by maybe building a few flamers and scouting with them, meanwhile you're seriously teching up your rockets or cannons instead?

With unit preview the element of surprise is lost, it's not "realistic" (without scouts) and shouldn't be back, no matter how many pretty pictures and maths you apply to this GAME.

[Emdek]

ODDity has good point, you can always try to trick enemies, especially if they are humans.

[aubergine]

I'm coding my AI to "not cheat", in other words it won't get a list of opponent units it can't see yet. It has to wait until it can see them on the map. (although even then, it still has really no idea what sort of unit they are - still trying to work out how to solve that one).

So I vote for "don't bring it back" - neither humans or AIs should know what the enemy has until they actually see it, otherwise why bother with a fog of war?

[JDW]

One word. Scout.

+1

[Iluvalar]

What if you're trying to outfox the enemy, by maybe building a few flamers and scouting with them, meanwhile you're seriously teching up your rockets or cannons instead?

With unit preview the element of surprise is lost, it's not "realistic" (without scouts) and shouldn't be back, no matter how many pretty pictures and maths you apply to this GAME.

You wont be able to outfox your opponent, because the principle of adaptation will NOT survive. He will see your flamers but it will be simply too late to change his path. Everybody will know that we can send any kind of random units and nobody will care about them anymore. Everybody will have a bunch of random units to intercept your flamer scouts you wont learn nothing from them anyway.

Even with unit preview, the surprise is good enough. You know that they are somehow researching machine guns, but you cant assess how well and how much. You need scouts for that. One could mass product or just build some defense units at any point. One could have built just a few flamer units to outfox you (yeah that work well in 2.3).

No, preview IS realistic. Your opponent spend 3 months researching some flammable new gel in some "hidden" lab. He just ordered enough chemical to feed 15 infinite ammo flamers. The is new laboratories that apparently mix flammable material together in the middle of their military base. The weird green gas exiting from the chimney can be seen miles aways...But you'd need as a commander to order the scout by yourself ? Explains to your spies how to do their job each time they go out ?

My point is : Once we'll get the 3.1 real feel, and we'll stop mimicking our habits of 2.3 to really play like there is no preview. We will face the fact that the inital research path choice decide of the whole game 90% of the time. the remaining 25 minutes of the game will just be a fancy movie of how you died/won.

The balance of the games was very carefully fine tuned to give clever player the time to outcome a bad weapon modifier and do a nice come back. By delaying the information to him by the smallest amount of time, make the damage he receive double or triple very fast. Since the game was designed to be challenging and it was already a close call, now it will simply be too much. If after you opened your weapon modifier are bad against you enemy, you lost. Whatever you try to outcome it.

[JDW]

You wont be able to outfox your opponent, because the principle of adaptation will NOT survive.

If you engage your opponent on a frequent basis, you get an idea of what tech he is going with. The sensor turret/tower isn't just to help your artillery target better.

No, preview IS realistic.

It's just a cheat.

[Iluvalar]

If you engage your opponent on a frequent basis, you get an idea of what tech he is going with. The sensor turret/tower isn't just to help your artillery target better.

That's what i try to explain. Even in the BEST possible scenario, your units will need to travel the map.

Wait... I believe I found a way to express it in words and use real data as a base.

[effigy]

I've been 'on the fence' on this subject since it was added to the master (quite a while ago)... at first I didn't like it, but I was getting use to it. Now that more people are playing 3.x, and different strategies are presenting themselvesI can say what Ilu is saying has some merit. Specifically, I'm experiencing issues vs. flamers. By the time I see thermite it's too late for thermal armor, and too late to build sufficient hardpoints.

Maybe tweaking the balance a bit, could help?

[zany]

why did the devs change this without talking about it first?
this is backwards

[Iluvalar]

Please, forgive the bad presentation of my datas. And please believe me when i say that it's a simplification, but still a good representation of the game.

Here you have:
from left to right the number of minutes you take to react.
from up to down, is the value of the weapon modifier your reaction turn upside down
The junction is your odds to win the game in % (assuming both players are of same strength)

As we can see, the more the we can flip an hard weapon modifier to our favor, the more we have a potential of gain (if we can find a way to turn it back completely). I left the 0% and the 10% modifier behind because i was getting weird results. You'd probably be better not to adapt and play with those odds. My premise "when do you adapt" was making little sens.

Please, keep your eyes on the 30% weapon modifier. As far as i know, it is the standard in most game. One of the 2 player will notice he need to change his research (at least a bit).

And let's talk about 3 scenario .
A) "Lucky day" : You were looking at the intelligence screen when the first unit popped out. At 5 minute, and you magically had a reslab that was ready and nothing more urgent to do. You quickly flip the a better weapon. You odds are still at 48%. No hurts from the bad start.
B) "fast guy" : You look at 5:30 in your intelligence screen, but your labs are used. While you build the end of your base you think simultanously to your plan B and you "adapt" at 6:00. You still have 40% chance to win. The randomness of the game is 20%. (10% for you, 10% for your opponent) we have a game...
C) "normal player" : You looked at 6 and you've been a bit slow... relax it's just a game after all... You finaly conclude that you will need to adapt and found a good plan B. You adapt at 7:00. 34% chance to win for 34% of randomness. We love this game...

Let's now suppose a "perfect scouting" behavior : You sent your very first unit at 5 for a suicide mission, he quickly cross the 100 second truck time map, go over the defense, and out of the 6 units the enemy already built you cleverly deduct what exaclty he is researching.
A) Lucky day : With the best odds of all (adapt at 7), you kept 34% chance to witn and the game in 34% over already.
B) Fast guy : at 8:00 (still very good) beside all your best skill you now only have 1/4 chance to win, half of the game is already decided.
C) Normal player: at 9:00 your chance to win even if you adapt is 18%. Actually it's not sure anymore if it would not be better to continu to shoot his borgs with cannons and hope he will do a mistake. 64% of the game was already played.

Now the "rationnal scouting" because face it... on single viper wheel have pretty low chance to pierce the defence and get reasonable information. So you build 3 scouts, and take the long route to avoid the enemy defense. you know that the opponent know that you badly need to scout, he will start by producing some random units. So you need to wait another minute to have good result. You took 4 more minutes to scout.
A) Lucky day : 9:00 hey 18% chances. Your gaming skill can still count ! Lucky you.
B) Fast guy : 10:00 You saw those borgs, but your labs are already reasearching heat cannon mk3. you still keep one chance out of 9 to win . 75% of the game is now randomness
C) Normal players (with normal scouting.) Note that this is not a "noob" scenario, just a decent player, that made a decent scouting... but still because both player mess with the intelligence of the others by producing random crap as long as they do not need to reveal their true plan, You know finaly discover what your enemy have in mind. One of you discover that he only have 7% of skill to show, 86% of the game is pure randomness... What ? but i sent my scouts !! where is the game ?... gone in the garbage.