I'm thinking of and RPG, FPS or a turn based game (does not really matter), where the player does not know his character(s) exact stats. Of course they exist as random variables of normal distribution inside the game engine. So no visible HP bar, Level, ATK or DEF. You can only determine it's status by it's signs of life (good mood, fatigued, dying, dead, etc...). With regard to casting spells, when your mana is running low, your chance of failing to cast a spell increases (so there is no "zero" point when you need to find the next gas station for refill). To measure your skill you can battle with monsters which has always have fixed (but yet unknown) stats.

The stats would improve as you use them, but each character would have a random talent set so some skills can be improved easier than others.

So my question for the game devs and gamers, is making the true self of your character a mystery a good idea?


From the answers it seems it depends on heavily the genre of the game.

For RPGs, I thought finding out what's good for your character is fun. From the responses, it seems it's not so fun because the players usually have an established play style and will try force it from their character for all costs.

The actual game idea I dreamed about is manager like game somewhat similar to the Pokemon game series. The main character has several champions and wanders around the virtual world and battles with others in an arena in a HOMM like fashion. And those champions would have the hidden stats. I didn't mentioned this in the original question because I don't wanted to bias the answers around this particular game concept.

For the feedback to the player (in both cases): There would be ratings similar to the elo ratings that a chess player has. One for the melee hitting, melee blocking, ranged hitting, ranged blocking abilities, and one for the magic abilities. Hitting the opponent with a sword will increase the melee hitting rating, if the enemy blocks the hit melee hitting decreases. If the opponent hit us with a sword, our melee blocking rating decrease, for successful block it increases, same for the ranged attacks. For the magic skill it's a bit tricky. High magic ability means: easier to cast spells successfully, harder to be affected by cursing spells, and easier to be affected by enchanting spells.

The standard abilities like hit points, toughness, strength, magic power, are estimated from the performance of the champion. These determine the actual damage dealt on successful hit. There would be "standard champions" that has a fixed and known stats, and the other champions' stats can be measured against them to bootstrap the system.

So if the player tries out all aspects of his champions he can find out the skillset his champions have, if the player don't like the skillset he can trade his champions to get the ones with skillset he likes.


So to get straight to point: Would it be fun if I center my game around the concept of getting to know your characters/champions, instead of serving all stats readily to the player?

It would be similar to the football players. The coach will need to see how their players perform in various positions to assign them to position that's the best for them. So he will need to try each one as a goalkeeper, defender or forward. After several matches it will be clear which position is good for them.

  • 5
    \$\begingroup\$ It's a good idea if it's how you want to make your game. If you don't want to make your game that way, it's a bad idea. I think it's an interesting concept, but I don't think there's a answer to it. \$\endgroup\$
    – House
    Apr 18, 2013 at 21:27
  • \$\begingroup\$ I asked my friends about this idea too. Most of them say they like seeing all the exact numbers, and having full control over the skillset they build for their character. So I'm a bit concerned about this idea, and decided to ask a larger community about this. \$\endgroup\$
    – Calmarius
    Apr 18, 2013 at 21:35
  • \$\begingroup\$ About the answerability. I think there is someone here that knows what are the things the players like. I often fail to think with brains of the players, and come up with crazy overcomplicated ideas that's are not interesting or enjoyable for anyone except me... \$\endgroup\$
    – Calmarius
    Apr 18, 2013 at 21:37
  • \$\begingroup\$ It depends on the gameplay and how well you make that clear to the player. If I don't know I have hidden stats that improve with practice, I won't bother training them up, I will just assume the result that appears most often is expected and only rely on what works regularly which hinders player progress/stat improvement. Personally I am not a fan because it just turns things into a guessing game. \$\endgroup\$ Apr 18, 2013 at 23:45
  • 1
    \$\begingroup\$ 4 votes up and 3 close votes? Talk about crossed opinions. \$\endgroup\$
    – Tim Holt
    Apr 20, 2013 at 4:36

6 Answers 6


This technique is called a black box. This is much more a style instead of a "right or wrong" decision that makes a game better or worse. As implied in the other answers, it can be done well or poorly.

There are two concerns of note.

1: First, you should be asking "why" you should blackbox, instead of whether it's a good idea or not.

Some games benefit from having the various statistics readily available. Players can more immediately and accurately determine if various components are desirable. The game can easily become a min-maxing fest if you aren't careful, but players generally feel more certain in their choices.

Also, any poor design choices related to statistics, such as a weapon you can only acquire late in the game with low stats, become more apparent and are easier to be found by testers and players after release.

Some games benefit more from blackboxing the math. If your focus is less on optimization, less on the statistics, and more on accessibility or some non-statistics-based mechanic that is central to the experience, hiding the stats would make more sense.

2: The second concern is that, if you decide to black box, you need to ensure that the player has some other form of feedback. Statistics are a (very basic) form of feedback. If you hide them, you must have some other form of feedback such that players understand the ramification of their decisions. As mentioned in other answers, visual feedback works (wounds, blood, red edges on the screen). Other options include audio, text, use of color, a means of experimenting with a frame of reference (a magic room in the menu ala Fable 3 with a "training room"), and more.

Creativity is a plus when creating other types of feedback, but whatever the form, it must suit the basic need of providing players with a means of weighing decisions. If the players can't see the "attack power" of two weapons, they need some way to decide and to feel certain the decision was right for them. If a player cannot see "Hit Points", he/she needs some means of feeling certain of the chance of risk for an encounter (not rushing in and finding out they were very close to death after they die).

Hope this helps

  • 4
    \$\begingroup\$ +1 for feedback. When the player fails at something, you must tell them if it's because they played bad, because their character just isn't suitable for the situation or if the scenario is hopeless anyway even with a more suitable character. Otherwise you will likely frustrate the players. \$\endgroup\$
    – Philipp
    Apr 21, 2013 at 2:09
  • \$\begingroup\$ Definitely. The player must feel as though all risks are obvious (after careful consideration or experimentation, at least), and that any decision made is an informed decision. Failing without knowing the reason (because of the lack of feedback) feels as though the game "cheated", even if it was theoretically fair. Definitely frustrating when it happens. \$\endgroup\$
    – Attackfarm
    Apr 21, 2013 at 3:05

Hidden stats can be fun if done correctly. It sounds like you are going in an interesting and unique direction which can land you in a very good place.

The problem I foresee is having random stats. While it sounds good on paper, it can yield very different results. Gamers have different play styles and if you force them out of what they are comfortable with, they can get frustrated. They may even resort to restarting the game multiple times when they discover their character isn't how they want it to be.

Maybe instead of random stats, you have them answer a quiz or perform actions on a vanilla character and base stats on the answers the player chooses or the actions they perform.

That way you can still have hidden stats but the player still has some control over how the character grows.


I think there is one very important aspect you have to deliver on and if you can then this is a perfectly good idea. And that is communication. The problem with abstracting over the numbers is that the difference may be very difficult or observe due to the manifesting over long periods of time or being very subtle. The number are an obvious if very crude way to communicative to the user exactly what is happening.

To touch on a few examples that you mentioned.

With regard to casting spells, when your mana is running low, your chance of failing to cast a spell increases (so there is no "zero" point when you need to find the next gas station for refill).

If the player just stops working that is going to be very frustrating to the player. You need to communicate the mana levels somehow. You can do this by the character visibly changing or complaining more vehemently as the mana levels decrease.

The stats would improve as you use them, but each character would have a random talent set so some skills can be improved easier than others.

This too has to be communicated or it is going to be really hard for the player to notice. Say if a character has a propensity for archery that should be built into their back story or they should volunteer or show enthusiasm for tasks they are talented at. So I think just a random talent set is a bad idea unless you can communicate the talents in some other way than observation or the player's trial and error.

  • 1
    \$\begingroup\$ You can make the feedback even more explicit. E.g. instead of showing a mana bar, show a "casting progress bar". Highly capable wizards would fill their casting bar quicker. For archers, you can do the reverse. Show an aiming circle, which shrinks over time. For better archers, the dispersion shrinks faster. (This is in fact how World of Tanks communicates gun dispersion in-game) \$\endgroup\$
    – MSalters
    Apr 19, 2013 at 11:11
  • \$\begingroup\$ I thought of communicating the abilities via ratings. Melee hitting, melee blocking, ranged hitting and ranged blocking would have elo rating. Which is adjusted after each interaction. Weapon damage is a known stat (weapon manufacturer knows the strength of the weapon), these elo ratings are known for the player. The strength and toughness (that affects the damage given and taken) are difficult to give rating to. \$\endgroup\$
    – Calmarius
    Apr 19, 2013 at 16:46

The point of having variable stats in the first place is first and foremost that the player can increase those stats. It is a simple but effective skinner box technique. Second it gives the player some customization options, choosing what skill to level up feels like making an important decision, and that is good.

The downside of variable stats is that you need to design the game to be enjoyable no matter how the stats have developed. This can be quite hard as tuning a fight to the desirable difficulty for a player with one stat set will often make it either too hard or too easy for a player with a different stat set.

If a player can neither see nor influence a stat then what is the point of making it variable? As I see it you get the full effect of the downside, but none of the upsides.


There would be ratings similar to the elo ratings that a chess player has.

A concern I would have with this system is this: ELO is based on number of games played. Which means that, initially, the player has no accurate information about their character(s). All they have are a bunch of people with random, hidden stats.

Indeed, ELO ratings are very misleading, because ELO is relative. If you're fighting a bunch of guys with low defensive stats, your attack ELO is going to be disproportionately high. In fact, if you have a lucky streak with the RNG, this can make a lower attack character look like a higher one.

Which means that, not only have you hidden information from the player (their actual attack rating), you are now lying to them by letting them think that their attack is higher than it really is.

This is purely subjective on my part, so bear with me. I feel that a videogame represents a social contract between the player and the game designer. The game designer provides a game, which displays information to the player. It is the player's job to see this information and interpret it. It is the game designer's job to make this information sufficiently clear for the player to be able to properly interpret it and make reasonable, meaningful decisions from it.

Game design, at its most core, fundamental level, is what any artistic endeavor is about: communication.

If I see that I have a high attack rating, I should expect that to be legitimate information. That is my right as a player, and it is the responsibility of the game designer not to infringe on that right.

I would consider it a mortal sin for a game to deliberately give me false information, especially info which is critical to my success. This is not misdirection or an unintended accident. This is perfidy: a game designer willfully and deliberately not acting in good faith.

And your ELO system can do exactly that, since ELO is an estimate that only gets better with the number of games played and the level of opponents faced.

It's one thing for a game to accidentally mislead you. Or for me to simply misunderstand a clue. But numbers are hard data. And if I have a high numerical value, I have every right to expect that this number is in some way accurate. Even if I don't understand what that number means exactly, I should expect that a character with 50 in some stat is at least somewhat better at that task than a character with 20 in that stat.

Break the social contract between yourself and your players at your own peril.

Furthermore, your ELO system encourages players to engage in more fights than they might want, just to get a more accurate idea of their ELO. I'm not a fan of a system that encourages the player to play in this way. I would prefer that the player want to play naturally, not has to be tricked into it just to get actual information about their character.

The thing you need to think hardest about is this: why do you want to hide these stats? What do you gain by it in terms of design? Or what behavior are you trying to prevent the player from being able to do? In short, what do you want your system to achieve that showing these stats makes harder or impossible?

Giving players (or people in general) information allows them to make more informed decisions. Thus, by denying them information, they must make due with less. This typically leads to more conservative behavior out of the player if losing has any real downsides. However, because the accuracy of their information is based on number of fights, it could lead to far more wild and experimental behavior, just to find out what a character's stats are.

Is this what you want?

  • \$\begingroup\$ "Because the accuracy of their information is based on number of fights, it could lead to far more wild and experimental behavior, just to find out what a character's stats are. Is this what you want?" Somewhat yes. \$\endgroup\$
    – Calmarius
    Apr 21, 2013 at 10:06
  • \$\begingroup\$ Are there any games that deliberately lie to you about the power of equipment and whatnot? Stuff that just intentionally holds back information or gives you false information? \$\endgroup\$
    – Shiester
    Jul 9, 2013 at 14:07

With regards to your "pokemoon-esque" game idea- having your OWN champions' stats be hidden from the player at first seemed a BAD idea, or rather one that I didn't like at first... however, I HAVE always felt that games which were attempting to be as "realistic" as possible would ultimately not even want to use stats, per se. If your pokemon game were "real life", one would not have the convenience of having a numerical value for, say "strength" but rather one would have to judge its strength based upon one's own knowledge and familiarity with the thing. Perhaps someone who is an experienced trainer can end up with a result quite similar to a numerical value, but a novice surely wouldn't. So I thought, what about making the accuracy of your assessment part of the whole game. Like the trainer has stats which determine the accuracy of what you see as your champions stats. At first you might only be able to judge strength, at least perhaps until you train a champion that is, say, highly intelligent so raising it allows you to judge Intelligence as a value. Because you are able to judge it doesn't mean that you are able to judge it well, or even accurately so you would perhaps have a stat which corresponds to the likelihood of the judgement being inaccurate, as well as one for the severity of the inaccuracy.

So, while I think that it could be used in some really cool ways in your poke-champion style RPG, my general view on the matter would be that if you are making an RPG, include the stats. If you feel the numbers detract from the experience and want to make the character's status apparent through how he looks or how the control changes YOU ARE MOVING AWAY FROM THE TRADITIONAL RPG, which is fine however if you are making a real RPG which utilizes a stats system, you should at least give the player a screen where he is given the numerical values for his abilities and attributes. As the guy above me said (paraphrased) how exactly does your game BENEFIT from what you want to do differently- it is not enough to want to hide the stats simply because it would be cool to hide the stats. Also, it was said that giving people more information allows them to make more informed decisions... giving them TOO MUCH information- either by drowning them in too much data or by giving them unintentionally misleading or unclear data, is just as bad. One must strike a narrow balance, which is why the traditional style is so commonly used- it manages to strike that balance and people are already familiar with it.

I enjoy when there is an entirely new way of doing things with regards to the battle system in an RPG; however while I do enjoy the process of learning a well-designed, unfamiliar system I always end up feeling like I've finally got the hang of it right in time for the last battle... so, it's a balance. Hope this helped :)


Not the answer you're looking for? Browse other questions tagged .