Why do we have two different factory tiers?

Question: I understood that no units were supposed to be strict upgrades of any other units. So if the goal of the game is to have each unit performing a different function or role, why do we even need two different sets of vehicle, bot, and air factories? Couldn't we just combine all the tanks into the same factory and then balance the units so they each have those different roles?

Partially because there does need to be some form of progression, if you had everything right away there wouldn't be as much depth. Also one can still have 2 tiers without them being strict upgrades, to use my favorite example, AA weapons.

Basic AA unit would be your average Direct fire weapon, average range, low damage high rate of fire.

Advanced AA unit would be something like Flak, Shorter range, low damage but with a large AOE.

A Second Advanced unit w0uld be missile based, long range, high damage, but a slow Rate of Fire.

Those two advanced units don't replace the basic unit overall, in some situations yes the Flak or Missile unit will be a better pick, but not ALL the time.

Considering we only have what, 20-30% of the planned final unit count I think it'd be pretty cluttered if we tried to jam all units of the same type together into a single factory, especially down the road as more units are added and such.

Mike

It's important to remember that you're commenting on Completely unfinished balance and design. Just so we're clear.

As for putting them all into the same factories which you get at the start; there's a subtle pacing that having distinct levels of factories gives to the player. It's important since the player having full access to every unit from the word go would be overwhelming, not to mention lure them into 'traps' from which they'd never recover.

I would hope that, down the line, you would want to build 'basic' (T1) units before 'advanced' (T2) units anyways, not just because of their cost, but because basic units are more versatile, which is useful at the start of a match. Basic units should be the bread and butter of your army. Advanced units should be there to enhance the basic units, not replace them.

In space sims, fleets are not made up of just battleships and dreadnaughts (advanced units). Fighters and frigates (basic units) are needed to protect and assist them, and are generally present in much greater numbers. Hopefully, PA will work in a similar way.

Good points. I have to say I basically concur with everything said, especially the progression factor mentioned by Mike. I'm also glad to see that we are all striving for (eventually) T2 units that are not replacing T1 units.

I think economic growth can provide all the progression which is needed. A small economy promotes generalist units because an army purely made up of specialists requires many units in order to cover many roles and not have any fatal weaknesses.

Yeah, it just adds a layer of basic progression and strategy. Gives the player choices like 'do I spam endless t1 or try and get early t2 to get an edge'. When to move up to t2 is an important strategic choice. Levellers are crazy powerful when used well (i recently won a game whereby a group of about seven levellers coupled with advanced radar coverage absolutely decimated an incoming attack and managed to move in on the enemy's base and make significant progress).

They out-range t1 laser turrets too, which can be critical for penetrating a base.

Anyway, I see your point, but I think the extra strategy involved in having separate t1 and t2 factories is important and makes for good gameplay. You should probably be thankful that there isn't t3!

Please stop using that word. It becomes a calculation as to when you build the 'Upgrade'. There is no choice whatsoever. It's a mandate, and a binary one at that. Get it right and you live. Get it wrong and you lose.

Thats not true, you can't just calculate when you do it. It depends on so many factors including preferences for different strategy that there is no correct time.

Additionally if t2 isn't directly superiour its also not a I win/you die situation. I can try to win through t1 and catch you early or try to prolong it and go another path. Thats still a choice.

No, it isn't. You can choose to prepare for an prolonged game, but that already was all you could choose.

Well, actually you could not "choose", because you are forced to be prepared for a prolonged game either way in case your rush attempt fails. The goal is to win in every possible scenario, not just with a certain percentage, so you can't say "I play that strategy and when it fails, I give up".

When you fail, you are forced to go T2. Well, you are forced to go T2 Levelers anyway because there is no alternative. And there is a perfect point to go T2 which can be calculated, independent from your "strategy", that point can only be delayed by wasting resources on T1 rush which is risky and therefor inefficient.

And that is the very point at which T1 can no longer break through enemy defenses without massive losses. At that point you are playing the very same game again, but only with the next tier of units.

Just because something has too many variables for a human brain to process, doesn't mean it's not a calculation that has a definitive answer,

If you follow that line of argument all games are solvable. Just because chess is theoretically solvable doesn't mean you don't have choices in your moves.

Theoretically, Chess is calculable. On a human level it is not.
Deep Blue demonstrates this to a significant extent. A more powerful computer would do better. A sufficiently powerful computer could optimise play to the extent that no human could beat it.

There is a finite number of possible positions for chess pieces on the chessboard. A finite number of variables.
With a finite number of variables a computer can calculate the optimum solution.

Edit: a hilarious side note.
A mobile phone won a category 6 tournament with a performance rating 2898: chess engine Hiarcs 13 running inside Pocket Fritz 4 on the mobile phone HTC Touch HD won the Copa Mercosur tournament in Buenos Aires, Argentina with 9 wins and 1 draw on August 4–14, 2009.
-Wikipedia

Chess actually is actually considered unsolvable, at least there is no proof or even evidence that there could ever be a perfect solution which guarantees you to win.
That's because the game is highly dynamic and there are scenarios in which you can only protect yourself against one of two or more threats. You have to make decisions based on statistics in that case, but it still won't protect you from every possible move.

It's not until the last 5-10 moves that one player is actually guaranteed to win, there is an infinite number of possible moves (not necessarily distinct moves!) before one player can only loose. So while it is calculable, it isn't really solvable, you can only achieve a certain chance to win which ends up somewhere around 50:50 with a slight bias.

The game Go on the other hand IS solveable, even though not by machines yet. You can't win by points gained, but the starting player gets a bias and it is possible to carry the bias over which results in a 100% win chance for the first player when playing perfect. This is an obvious flaw in the rules (the bias).


Actually, almost every single game you have played in the last years is calculable, thats something I can guarantee you, at best you are presented a paper-scissor-rock system on top of the equation to create the illusion of a real choice, that's actually just playing with chances.

But that is what RTS is all about. It's calculable paper-scissor-rock on the strategic level. Unless of course someone changes the rules to "rock beats everything", in that case the game degenerates and that's happening with T2. T2 beats them all. You can go for scissor (rush) and hope for paper (weak defense), but rock (defensive tech rush) beats them all.

And thats when it gets bad. It's suddenly not only calculable (you know your chances), but solved.

There is not a guaranteed win, true. There is significant evidence that proves that, at the highest level, you can do no better than draw. But I contend, there is most definitely not an infinite number of moves.

There are only 32 pieces on the board, each with unique but ultimately predictable movement patterns, 64 spaces that only one piece can occupy at any one time and only one movement per turn.

Not distinct moves, but a game can consist of an unlimited number of moves, if it wasn't for the "game ends if no figure has been removed for 50 moves" rule.

Thank goodness for that rule then

You are really wrong. Just... really wrong.. on the internet.

Chess is solvable. This means that we know that either white wins, black wins or the game is drawn given perfect play. Either one player is guaranteed to win from the start of the game OR either player is able to force the game to not end or draw.

Go is solvable (and considered harder than Chess) but there is no result proving that the first player wins.

You should not think in terms of moves. Think in terms of board states. Clearly there are a finite number of board states for both these games. Some states are endgame states, one of the players has won. Other states have moves branching off them to other states. These games are effectively directed graphs in which players alternate in choosing an arc to follow. You can always win a game if you force your opponent to send you towards one of your win states.

Bottom line, chess hasn't been solved yet, in either of the two senses of the word. It might well be possible, we just don't know yet - http://en.wikipedia.org/wiki/Solving_chess

I know what you're getting at Nanolathe, but i'm fairly sure on a small planet, t2 might not be necessary. You might be better off producing as much t1 as possible.

In all probability though, there probably is an optimal time to go for t2. But generally speaking, it is a choice when the player does it. We're not robots!

I'd argue that it's not a choice at all. You're forced to due to the opponent and due to Basic units having a very poor efficiency rating compared to Advanced ones.

The only thing you have control over is when you try to 'upgrade'. And that's not a choice, it's a calculated risk.

Also, the link you gave very boldly states that it is entirely possible to solve chess.