Aucune réponse à ce sujet

[Alerus2]

One thing that is kind of unfortunate in the ME games is that there is no persuade option if you are neutral in morality. This is unfortunate because the game pushes you to play as either a strong paragon, or a strong renegade since there is no advantage for playing the middle. Since this in an RPG, the player should be able to play as they see fit. To combat this, I think there should be neutral persuade options, which perhaps relates to a diplomat score. There are few ways to handle this with varying complexity, advantages, and disadvantages, which I detail below. Ultimately, I think my final solution is the best, but I'll explore all the ways since I think some version should be implemented. 

The first and most basic way is to simply award diplomat points for choosing neutral options. However, this is not sufficient because it doesn't factor in that neutral person will sometimes err on the paragon side and sometimes on the renegade side; a neutral person is a mixture and this is not captured here.

The second way is that your diplomat score is (Paragon + Renegade + Neutral) / 3. It's fairly simply and would mean that someone who plays in the middle would be able to use neutral persuasion in the same amount of time as a paragon in renegade for a comparable score. The major downside to this, is discussion choices are zero sum, which means a full paragon would always be able to make both paragon and neutral decisions. This clearly is not correct.

The third method is to simply require two requirement checks for a neutral persuasion: one for a certain amount of paragon and another for a certain amount of renegade. If the comparative paragon/renegade persuasion choice requires a score (in the respective category) of s, then the neutral persuade would require both a paragon score of s/2 and a renegade score of s/2. The problem with this approach is three fold. First, there is no easy score someone can look at to see how they're doing as a "diplomat." Second, someone who has played the game a long time is likely to have some small amount of renegade such that they could perform easy neutral options even though in the grand scheme of the experience, they don't represent a diplomat. This isn't violated as badly as it is in the second approach, but it still exists. Third, this doesn't incorporate neutral choice options that people make. One might suggest further altering this to be three checks: s/3 for paragon, renegade and neutral, where neutral is defined as the number of neutral decision points in discussions, but this gets even more messy and still doesn't respect that a neutral player make all neutral decisions or some neutral, and some renegade and some paragon, for which the latter two are balanced.

This leads me to my final suggested approach for handling diplomat scores. If we can define a function that represents the paragon and renegade score balance, then we can create the diplomat score to be a mixture of that balance and the relative amount of neutral choices/points made/received. In order to do this, we will utilize the notion of entropy as it is used in information theory. Information entropy takes a probability distribution and returns a value between zero and one. Specifically it is defined as:

-1 * summation [ p(xi) * log(pxi) ],

where the log function base is in terms of the number of possible events to which a variable can be set. If a probability distribution is uniform (that is, all events equally likely), then the entropy returns a value of 1.0. If a distribution is completely biased (that is, only one event ever occurs over the possible events that could occur), then the entropy is 0.0. Therefore we can analyze the balance of paragon renegade decisions by looking at the relative entropy of them. Here we create a total number of renegade and paragon points which we call tpr = paragonScore + renegadeScore. Then we compute the ratio of paragon score to renegade score pr = paragonScore / tpr; rr = renegadeScore / tpr. Finally, this means that the entropy of paragon to renegade is computed as:

-1 * [(pr * log(pr)) + (rr * log(rr))].

When the number of renegade points is equal to the number of paragon points, the entropy will be 1.0. When it is completely lopsided, the score is 0.0 and mixtures in between move between these values appropriately. From now on, we refer to the function that computes this entropy as PREnt. We can now represent a final diplomat score by working in the neutral points received. First we will compute the sum of all possible points tp = tpr + neutralScore. Then the ratio of neutral points becomes nr = neutralScore / tp. Similarly, we can compute the combined paragon-renegade point ratio as prr = tpr / tp. Using these values we can compute the final diplomat score as the weighted average of the neutral score and how balanced the paragon renegade score is:

DiplomatScore = [(prr * PREnt) + nr] * tp

To finish it off, you can then of course give the final point normalization so that progress can be tracked through the game (that is, when only a few decisions have been made, the diplomat score is low). This gives the player a raw diplomat score that incorporates neutral decisions and the balance of renegade to paragon decisions as well as the portion for which a player makes those non-neutral decisions.


Given how soon ME3 is coming out, I fear that they wouldn't have time to incorporate this, but maybe in future games they will. Thoughts?

Modifié par Alerus2, 12 décembre 2010 - 07:30 .