Sponsored By

This article offers interesting considerations about the practice of "savescumming" in videogames, by discussing the random number generation mechanic used in XCOM Enemy Unknown.

claudio scolastici, Blogger

November 9, 2013

9 Min Read

How I fought the Aliens

Exploitation of the XCOM Project

xcom art

Few days ago I didn't know I am a savescummer, nor that there was a name for the practice of saving before entering a hard challenge in games (in my opinion, I simply have a good hearth and get attached to my game characters!). Good to know.

Secondly, because these days I spent almost all my gaming-time playing XCOM Enemy Unknown on my PS3 and I actually faced the problem, as now I understand it is considered, of savescumming.

Let's begin by stating that I am the kind of player who reloads previous saves when he doesn't like the outcome of a game situation, no matter the cost. Example: I burned like 8 hours of gameplay in Dragon Age when I realized that it was more convenient for me to approach a series of quests in a different order. So what?

I totally agree that if all outcomes of game challenges, even the negative ones, provided an equivalent perception of having achieved something important, there was no need to reload a sequence. But how many games really do that?

When I played Mass Effect for the first time, I lost Wrex. It's been quite a shock when it happened, but it still happened and I had to bear with it. I considered that turn in the plot an excellent one, because I was really sorry I had to shot down the character I was using most during missions. I liked Wrex.

When someone told me that it was possible to save him, I was so happy and curious to see how the final fight would have been if Wrex was there, that I started an entirely new campaign and picked all the decisions that would have helped me saving Wrex towards the end of the game. So I saved Wrex, I quite easily defeated Saren and saved the galaxy and all.

My point here is that even if both “Wrex alive\Wrex dead” outcomes provided me with equivalent, though different, emotional reactions, as both outcomes are epic enough, still I took advantage of my distinctive faculties (memory for game events and the possibility to save and restart the game whenever I want to) to get to the point in the plot where I wanted to get to.

Should I feel sorry for having saved my favorite character by exploiting my knowledge of the game during the second run, or guilty when I met Wrex in Mass Effect 2 as the king of his people? I don't think so, even if, from the point of view expressed in the article, I cheated because I didn't want to bear with the consequences of my actions during the game.

I believe that the player has any right to make use of his head to play any game as he likes it. Designers don't do games for themselves, they do games for people.

My brother liked to play Starcraft and Heroes of Might and Magic. It happened one day that I thought him a few cheat codes and, from that time on, he never played those games again without cheating. To me and other people he was ruining his gameplay experience, ok...but what about his opinion?

My brother wakes up at 5 am to go to work. When he comes back home he has a young daughter and a wife to take care of. If he gets half an hour a week to play a videogame, should he be blamed for playing as the hero who can't fail? If he couldn't play like that, he would not probably play at all. Would it be better like that? We game developers only want the most competitive players to play our games?

I want to be clear on thing, though: I am not a player who searches for easy challenges. I play games at Normal difficulty only when I don't really like them and thus play them as “part of my job”. To pick an example that has been mentioned along this thread, I beat Demon's Souls single player's campaign and I am now on my second walkthrough. I know what it means to swallow shit while playing videogames. But how many players enjoy such games?

So let's get to the point.

While playing XCOM I realized that the game implements a distinctive system to manage random numbers.In XCOM, each offensive action performed by the game characters (read: shooting) has a chance of success which depends on many factors: cover, aim, altitude, armor etc.I realized that the outcome of a shot didn't change with repetitions, no matter how many times you do it.

Let's make an example. One of your soldier has a 80% chance of hitting the enemy and thus save everyone. Instead he fails the shot and, due to a chain of unfortunate events, your entire squad gets killed: Mission failed! (if you ever played XCOM at Classic difficulty, you know that such things actually happen). Now assume you have a savegame right before taking that shot. You reload that save, repeat the 80% shot and this time you think he can't miss, because, with such a high chance, he cannot fail twice in a row. But the soldiers fails the shot, again.

The reason is that in XCOM, as it seems to me, random numbers are not generated the very moment they are needed, when you call a shot. Instead, random numbers are generated at some point during the game (it could be anytime from the launching the game to the beginning of each game turn, as far as I know) and put in a list. When a check is needed, the next item in the list of random numbers is picked and confronted with the chance of success of that particular game action. Then the outcome is resolved.

This discovery initially pretty much surprised me, because, as an old-school role player, I assumed that using random numbers in games was like throwing dice during role-playing sessions: you do it when it's needed. Instead, in XCOM (and probably other games too) it's like you rolled a hundred dice before starting playing and, whenever you need a probability check, you pick the next roll from the list, instead of rolling dice.

This design choice doesn't make any difference, if we consider it having the Probability Theory in mind: the system uses random numbers anyway, as numbers cast by a computer can be truly random. But it makes a great great difference from a gameplay perspective: from my savescummer point of view, it turned in fact XCOM from a strategy-tactical-turn based game with random chances into a puzzle game where the goal was to find the optimal set and order of actions for each game turn\game battle.

The funny part is that, when I made this discovery about XCOM, I got excited as I thought the designers made this choice specifically for players like me who save frequently and reload to get the outcome they prefer. In fact, by exploiting this mechanic, I could beat XCOM at Classic difficulty, loosing just 2 rookies in the entire campaign! I am pretty sure that if you ask my soldiers who made it, they won't complaint about my game-saving habits! By reading this article, on the other hand, I understand that probably that design choice was taken to fight against behaviors like mine and that, by taking random numbers from a list, I should have been prevented from reloading before any important shot, to get a positive outcome.

This design choice makes completely sense, on one side. I can imagine that it effectively prevents a player from keep reloading a specific save game until that 20% chance to hit turns into a positive outcome. If your soldier is shooting with just 20% chance to hit, it means that you poorly deployed him on the battlefield. Instead of simply re-taking the shot, try to better deploy your soldiers. On the other side, this solution is useless against savescummers like me, and I'll explain you why.

Let's go back to the example of the soldier with that 80% success shot that he must not fail: how can we make sure he doesn't? Simply, let's have another soldier take a shot before the important one. If you are smart enough to create savegames in “mission is going well” conditions, it is possible to exploit this technique. Let another soldier “burn” that roll on a less meaningful game action, and then have the 80% soldier shoot again.You'll find out that, this time, your soldier won't miss! Since rolls are picked in the given order from the list, by trial and error and patience you can even get an idea of the values of each roll and make sure that, if not any shot is a hit, at least you get the most out of the rolls available for each specific game turn (I guess at this point, the designers of XCOM could really hate me!).

Yes, that's actually cheating, but I can assure you that my fun wans't touched by that. The pleasure of saving my dear soldiers and see them progress in their career was infinitely higher than the sense of guilty for taking advantage of a game mechanic. It also didn't make the game easier at all! I mean yes, I didn't have to face problems like not having the right men to send into mission because they were all dead or gravely injured, but still, I had to complete several game missions elaborating attacking and defensive strategies for my team and then testing them to check the outcome they produced, hoping it was good enough. Isn't XCOM a game about that, in the end?

Designers don't do games for themselves, they do games for people!

Read more about:

2013Blogs
Daily news, dev blogs, and stories from Game Developer straight to your inbox

You May Also Like