FTA and 2(3) actions on first turn

[wonderpug]

(09-10-2012 10:09 PM)Xpander Wrote:  What I was saying there is that if both players pass, then the game is reset. All that happened was nobody did anything. There is no bonus 5 wits. Both players get an equal number of wits to work with. I was referring to something Joggies said in his post. I understand why people use that as an example, since p1 essentially gets to respond to p2 with 5 more wits if he saves up. My point was that by saying that, you're essentially ignoring all kinds of other things like board position, build capacity per turn and various other things specific to the game...essentially oversimplifying a complicated issue.

I understood your both players pass scenario, but you're changing the situation. The "why doesn't P1 always pass his first turn?" question is most certainly an oversimplification, but the "why" aspect helps one ferret out all the more complicated reasons. He doesn't give himself 5 free wits because he wants those bonus wit spaces claimed, or he wants to have a chance at nabbing an enemy medic or sniper, or he wants to get some map vision, or he doesn't want to lose that turn's chance to medic boost.

(09-10-2012 10:09 PM)Xpander Wrote:  It's entirely possible that there is FTA. I'm not saying there isn't like Garcia seems to think. What I'm saying is that it's not proven. There's a big difference there. As I mentioned earlier, the only proven case has been addressed.

Hard proof would definitely be great, but from what we've heard from the devs the match result data pile is so massive it's hard for them to do large scale analyses. The reason they ended up releasing data on just the top 100 players was because they said that was a more manageable data set. It will definitely be very interesting to see the results from the World Cup.

But just because we can't get hard info on win/loss for going first vs going second doesn't mean we can't talk about things.

Take the Glitch map, for example. Start a Pass & Play match and imagine that you're the second player to move. What do you have to do to make sure you don't lose your medic to a runner P1 may have deployed? He's not safe where he's standing, moving the heavy to the wit space won't give you visibility, and any place you move the medic to boost the heavy will be under threat.

So ok, you decide to scout. You spawn a runner. You can instantly rule out a few spots a medic-killer runner might be, but not all. You still don't know if your medic is safer where he is, or safer if you move him over to your spawn point.

So now you claim your bonus wit space with your runner. Great visibility! And oh crap, there's a runner in the hex 3 spaces from the enemy spawn, the one that touches one hex of the double barrier, the single hex barrier, and the edge of the double-hex hole. What do you do now with your remaining 3 wits? What can you do to save your medic?

There are only two things you can do. One, you can retreat your medic down the back door alley of your base. Well that's a waste.

Option 2 is to move him into the nook between your heavy and your base. Well great, he's safe, but now he's in a dumb position. He can boost the heavy from where he is, but you'll have to move him again to get him in position to boost newly created troops. Another wit wasted.

And what did P1 have to do to get P2 to have to go through all that? Spend 2 wits to spawn and move a runner. In his turn P1 could have also moved his medic to the good spot between the heavy & spawn spot, and boosted the heavy.

Is this not a decidedly better first turn for P1 than it is for P2?

[Necrocat219]

Quite simply first turn advantage means that P1 has some sort of initial advantage. Even if you ignore wits completely as some people don't understand how player 1 doesn't have an advantage through wit count, player 1 always has positional advantage to start off with, he moves unite into important positions before player 2. Place a runner in a spot so they automatically threaten player 2. Move a heavy faster towards player 2. Start generating a soldier wall earlier. position wise, until player 2 outplays 1, player 1 has the advantage

[Alvendor]

(09-10-2012 01:33 PM)Xpander Wrote:  [quote='Joggies' pid='7021' dateline='1347225932']
If P1 for example does absolutely nothing (just passes), then P2 is in an identical position to as what P1 would have been except that the opposition now has 10 witpoints in the bank. If as P1 I can effectively choose to be P2 but with twice as many points, why would I ever choose to be P2?

The fact that most people don't pass their first turn as P1 just indicates there is an advantage in being player one IN ADDITION TO the bonus five witpoints.

(09-10-2012 01:33 PM)Xpander Wrote:  I saw a few people saying this in various threads as proof of FTA...I'm not sure what this proves if anything.

In this post I use the term "perfect" as in the best you can play given the vision you have. Not perfect as in "I use a map hack and can now play perfectly".

Joggies example it is fairly close to a full-blown proof. Granted, to completely prove if player 1 has an advantage if both players play perfectly you probably need to solve the complete game and that is pretty unlikely (harder to do than chess?).

Lets just play with the thought of being able to solve the game completely. We would end up with one of three results:

1. The game continues indefinitely, a draw that is.
2. Player 1 always wins. (The example above proves that player 2 cannot always win, if so player 1 would just skip turn).
3. Player 1 wins X% of the time. As this is a game with fog of war certain positions are of the rock-paper-scissors type and the players have to guess the exact positions. In these cases it could for example be that player 1 wins 70% of the time with perfect play.

Now to completely prove that player 1 has a FTA on a map we would need to prove one of these scenarios which we can't. But if we just assume this sentence we can prove existence of a FTA:

"Having more wits in the starting position increases the chance to win."

Now take player 1's choices. He can either
A) Move as usual
B) Skip turn and become player 2 with 5 bonus wits.

Assuming we agreed on the sentence above we know that choosing choice B) increases your chances in comparison to starting as player 2 without bonus wits. This is all we need to know, we don't care about what the other player does.

(09-10-2012 01:33 PM)Xpander Wrote:  It's not entirely clear if going first is a quantifiable advantage, for the reasons above and others, that can be represented by a number of wits to balance the game.

Again, you are correct we can't strict mathematically quantify the FTA in wits. But from a practical point of view it would be fairly straightforward to test various approaches and making sure both players win close to an equal amount of time. Either measuring all the gaming population, or measuring high ranked players. Player 1 could get fewer starting wits or as I have suggested earlier, give player 2 bonus wits.

[JonB]

Perhaps this has been mentioned -- but the first turn advantage exists in large part because P1 has more knwledge than P2 when making the first move. P1 knows exactly the location of all of P2's pieces, thus can move with impunity, forcing P2 to defend from multiple angles.

A way around this would be to allow users some choice in initial placement before the game starts. Maybe a choice between A, B, and C arrangement. Perhaps with different initial wit costs. That would probably be really hard to balance, but that is the fix for the FTA -- equalize the initial knowledge gap somewhat.

[Kamikaze28]

(09-11-2012 05:38 AM)Alvendor Wrote:  In this post I use the term "perfect" as in the best you can play given the vision you have. Not perfect as in "I use a map hack and can now play perfectly".

Joggies example it is fairly close to a full-blown proof. Granted, to completely prove if player 1 has an advantage if both players play perfectly you probably need to solve the complete game and that is pretty unlikely (harder to do than chess?).

Somehow related to your post is a thread I started way back when about the difficulty of creating an AI for Outwitters which demonstrates the difficulty of establishing 'prefect play' in Outwitters.

After skimming through the last page, I would like to summarize the different factors that go into FTA:

• Perfect knowledge of P2's unit placement
• Opportunity to lead in unit count
• Opportunity to move first (initiative)
  

By passing as P1, you effectively give all these advantages to P2 and get 5 Wits in return. I agree with Joggles and stipulate that having 5 additional Wits is not equally important to having these three advantages.

[Alvendor]

(09-12-2012 07:48 PM)Kamikaze28 Wrote:  Somehow related to your post is a thread I started way back when about the difficulty of creating an AI for Outwitters which demonstrates the difficulty of establishing 'prefect play' in Outwitters.

Yeah even assuming a computer with unlimited processing power it would be tricky to solve. The combinatorial complexity from chess together with the unknown information similar to poker would make it a programmer's nightmare to code.

[Xpander]

(09-12-2012 07:48 PM)Kamikaze28 Wrote:  Somehow related to your post is a thread I started way back when about the difficulty of creating an AI for Outwitters which demonstrates the difficulty of establishing 'prefect play' in Outwitters.

After skimming through the last page, I would like to summarize the different factors that go into FTA:

• Perfect knowledge of P2's unit placement
• Opportunity to lead in unit count
• Opportunity to move first (initiative)
  

By passing as P1, you effectively give all these advantages to P2 and get 5 Wits in return. I agree with Joggles and stipulate that having 5 additional Wits is not equally important to having these three advantages.

That's about the best description I have read so far on this. I was kind of hinting at that in my posts if you read between the lines, no amount of wits will balance this. Still, I hate to play devil's advocate, but I'm still not entirely convinced that this is a major issue. People have been saying this about chess for years and the game still marches on without any changes. I agree that the way match pairing is set up right now, it can be abused, but that's probably an easier fix. I admit I haven't played any games as first player in a long time...maybe I've just gotten good at going second, I don't know, lol. I just hate for there to be some sort of high level change that impacts the game in a negative way and then a month from that people will be complaining about Black having the advantage... I just hope that this gets some serious consideration before a decision is made on changes.

[garcia1000]

The nirvana fallacy is the logical error of comparing actual things with unrealistic, idealized alternatives. It can also refer to the tendency to assume that there is a perfect solution to a particular problem.

A person using the nirvana fallacy can attack any opposing idea because it is imperfect. The choice is not between real world solutions and utopia; it is, rather, a choice between one realistic possibility and another which is merely better.

[Xpander]

(10-07-2012 11:49 AM)garcia1000 Wrote:  The nirvana fallacy is the logical error of comparing actual things with unrealistic, idealized alternatives. It can also refer to the tendency to assume that there is a perfect solution to a particular problem.

A person using the nirvana fallacy can attack any opposing idea because it is imperfect. The choice is not between real world solutions and utopia; it is, rather, a choice between one realistic possibility and another which is merely better.

You must enjoy trolling, lol. The next time you quote Wikipedia, use actual quotation marks.