Rewards System

[Kimball]

Although I can't seem to locate the original link, a few months ago someone posted an article discussing how game developers are using psychological evidence to get gamers hooked. While I am not suggesting that we aim to get people addicted to 0 A.D., I propose that we not rule out implementing certain features based on inspiration drawn from the same features discussed in that article. This primarily consists of complex reward systems that keep players coming back to achieve that next level that they're just "so close to."

For Christmas, I got Gran Turismo 5. It has been in development for 6+ years now, and I think the reason is a lot of this theory has been thoroughly discussed and eventually implemented into the final project. Though their GUI is a mess, they've essentially built the game around the community and their rewards system. In playing this game, a few ideas came to mind for 0 A.D. that I'd like to share with the team. I've written a little comparative outline describing some of the ways we can mimic this system to develop a powerful method of training and retaining our fan base to play the game.

0AD_derivinginspiration.GT5.rtf

I beg of you, please take this with a grain of salt. 100% of what is written here is purely hypothetical - I want the team to review these ideas, discuss the pros and cons, and eventually decide on what we want to take from this. I am not suggesting that we implement all of these ideas into our DD immediately, nor am I suggesting that our game will suck without these features. Perhaps the game would be better off without any of the ideas outlined - so be it. I'd rather have evaluated the possibility of a rewards system thoroughly and come to a logical decision collectively than to have produced a final product that not everyone is content with - including the fans.

I have also not taken into account the amount of work necessary to implement these features. As the game remains in the alpha stage, I suggest that we not eliminate the possibility of developing additional features. Rather a game that we can be proud of that the community waited for than disappoint everyone with a rushed product. Release dates should not be of much concern to us at this stage, even though we have made significant headway in recent months.

Anyway, time to hear your feedback.

[feneur]

I don't have time to read it just yet, but wanted to post a few general comments (which of course may or may not be at all relevant, so only take them seriously if they're relevant, otherwise move on and forget about them )

In general I think rewards systems can be "an easy way out", when the game itself isn't interesting enough it's easy to throw on some arbitrary rewards, like: get a ribbon if you win 25 games, another if you win 25 with a certain civilization or a third if you win a game only using citizen soldiers. While that can be nice to have as a minor addition, it should never be put in as a major feature/instead of something else. I think having more game modes is more relevant and better. From things like nomadic, herocide, etc through (custom) scenarios and random maps to things like the proposed idea of having a "conquer the world" mode where you have to conquer several maps in sequence to win the overall game.

Any kind of rewards system is probably more funny/useful if it's in some way "social" (i.e. either multiplayer or submitted scores/statistics from singleplayer games), not necessarily in the way of comparing yourself to others in an unhealthy way, but more like giving people a small badge they can show in their forum profile if they've achieved a certain goal etc. So it's at least a bit more than just having a title light up in an in-game dialog or so. I don't think we should do things like having special content that only players who have achieved a certain goal can access/can access before others. That would go against the idea of 0 A.D. in my humble opinion.

So, now I'll go cook and eat and other things and hopefully I have some time to read through the document and enter some more related comments later (Otherwise I'll do it tomorrow, so please remind me if I haven't posted by Saturday )

[Kimball]

I'll wait until you've got time to read it then. It's not necessarily a cheap "easy way out", rewarding players for arbitrary things like winning a certain number of games. You'll see.

[Ykkrosh]

Interesting thoughts

I think an important factor is whether the meta-game (of collecting achievements and unlocking features and beating high scores etc) reinforces the primary gameplay (which we design to be fun by itself), or undermines it, or is indifferent to it.

With racing games, the primary gameplay involves winning races and getting around tracks as quickly as possible, and that involves components like accurate steering and accelerating and handling particular corners of particular tracks. It sounds like the GT5 license tests reinforce those components, and the unlocks reinforce the main goal of winning races, so they're encouraging you to play the game more and better, which is a good thing.

In our game, the basic goal is to win matches by defeating your enemies, and that involves various low-level micromanagement tactics and some higher-level strategies. We need to be careful that anything we add should reinforce that gameplay. In that case, I think it could be good to follow the license tests approach and have lots of short missions that focus on particular aspects of the gameplay (optimising your economy, choosing troops to counter your enemy, deploying your troops for battles, etc) that can function as an introductory tutorial and also as ways for skilled players to practice and improve. We'll have to have tutorials anyway, and if they can be reused as advanced challenges (by tightening the time/resource constraints etc) then that's a cheap way to provide more value to players.

But our game isn't about winning matches as quickly as possible, or using as few or as many resources as possible, or training as many units as possible. If we rewarded players for those things, outside of specific training missions where those things are directly relevant, I think it would undermine the gameplay: players will always use rush strategies, or always train huge numbers of expendable troops, because it gives them more experience points. Players who want to take a different approach, e.g. slowly building a heavily defended city then doing a sneak attack on the enemy's vital buildings, would be penalised in the XP system even though they're doing something good and fun within the main gameplay. The result would be harmful to the game and to the players' enjoyment - the meta-game would be distracting them from the game itself.

So I think we shouldn't reward the player for artificial achievements; we should just reward them for playing the game properly and well, which pretty much means rewarding them for winning matches. Single-player campaigns do that - the reward for winning on one scenario is access to the next. A conquer-the-world mode would do the same - the reward for winning is expansion into new areas and new maps, and perhaps new civs and new units. It doesn't matter how you win, so players are free to take whatever approach they fancy without worrying that they're missing out on anything.

Victory is a very binary thing, though - it can't tell the difference between an hour-long struggle in which the player nearly wins several times and is unlucky in the end, and a match where they fell asleep after five minutes and eventually got slaughtered, so it doesn't seem an ideal measure. Are there other ways we could measure progress and reward players, that wouldn't have the danger of introducing artificial goals that distort the fundamental gameplay?

[Kimball]

Yes, the "license test" model would be specifically to teach aspects of the game, not mindless goals to reach. We would also of course set the time limitations would be reasonable enough to be attained and encourage players to optimize their use of the game. While they may still advance through the ranks at a normal pace, to earn the higher rewards they'll have to make use of hotkeys, build orders and other techniques to optimize micromanagement and speed. This doesn't always mean an early rush strategy, this just means that players are using all aspects of the game to run whatever approach they take quicker. You can still "turtle" at a fast pace, so long as that means that you're actively managing your economy. That's what the rank promotions will facilitate. People will become better accustomed to our game environment and learn some helpful tricks along the way.

As for rewarding those who make stupid decisions to gain experience, this is where we would need to fine tune the system so that the players who do only train large quantities of useless units don't gain more experience for it. For example, where a player would repeat a scenario over and over again, the experience earned would be a diminishing value. A player would be better off diversifying his player style, training different types of units and using different strategies to accomplish goals, as this will earn him more experience.

That being said, winning a match will inevitably earn you more experience than simply gathering a bunch of resources and training a bunch of units. Also, experience could only be rewarded to those who do win the match. In the effort of providing incentive to the ultimate goal of winning, perhaps you only earn 50% of the experience points you've accumulated if you lose. Also, you wouldn't lose experience for losing a match. Players that fall asleep at the wheel will simply be unlucky and only earn that 50%. If they're falling asleep mid game, we've got work to do in other areas anyway.

The experience system wouldn't be the fundamental method of rewarding the player. Winning matches would be what would ultimately reward the player with additional content, not doing the same thing over and over until they earn something. For example, let's say initially, you can play with the choice of two factions on 5 maps against any AI faction. For winning using one faction on three different maps, you could perhaps earn a new set of maps (e.g. the desert biome set). Winning with both factions or winning a certain number of games against some other faction would enable you to play as them. The amount of experience points earned is certainly not our primary motivator here. In this sense, we are in no way diminishing the single player experience, but rather providing the player with incentive to try the game out, use all the factions and try out all the different maps. When they've earned a new faction or map, we can almost guarantee that the first thing they'll want to do is go try all of the new stuff that's available to them.

[Shield Bearer]

Players could get experience based on the diversity of their army and how well they use units to counter enemy units and also on how well they use their civ's special abilities. Then experience could be gained on the longevity of units. Flanking and other strategies could also add to that and how many attacks on their base they survived.

One way how the faction unlocking would work could be like this. When a player creates a new profile, he gets to choose one faction and accompanying that faction would be a set of maps reflecting the faction's nationality. So in order to play as a different faction the player would have to first play against them on his map and then play against them on their map. For example, if your Persian and you want to play as the Celts then you would have to first play against them on Persian soil, in other words, on a Persian map and then on a Celtic map after which the Celtic civ would be made available to you. (i cut and paste this from what i wrote before so what i said could be redundant)

[Ykkrosh]

As for rewarding those who make stupid decisions to gain experience, this is where we would need to fine tune the system so that the players who do only train large quantities of useless units don't gain more experience for it. For example, where a player would repeat a scenario over and over again, the experience earned would be a diminishing value. A player would be better off diversifying his player style, training different types of units and using different strategies to accomplish goals, as this will earn him more experience.

If we reward players for training different types of units (e.g. they get 10XP for the first unit of type X they ever train, 9XP for the second X, 8XP for the third, 10XP for the first of type Y, etc), then a player who's nearly won a match will think "I should drag this out for as long as possible and use my spare resources to train a bunch of those rubbish ships that I never otherwise bother using in order to get the XP". Training units needlessly isn't a fun gameplay mechanic, so our reward system would be encouraging players to be bored, and I'm not quite sure how to avoid that.

(I think there's plenty of evidence that players will willingly do boring things if you reward them for it (even with essentially meaningless tokens) - see e.g. grinding for experience or loot in MMORPGs, which is sometimes so boring that players will pay other people to do it for them, or see people who buy terrible Xbox games to get achievements and boost their gamerscore. The reward structure is addictive and makes people play the games for longer, which is good business sense when you charge for subscriptions, but if the player's not actually enjoying the game then it seems exploitative and dishonest and should be discouraged. And it's not just a problem in games - people will aim to improve measurements regardless of the real value of what's being measured.)

((I'm definitely vulnerable to this exploitation myself. When Team Fortress 2 had a week-long Halloween event where a present would randomly spawn in the level every few minutes, and whoever (out of ~24 players per server) picks it up first wins a wearable paper bag with a face painted on the front (from a set of 9, where if you collect them all you can convert them into a rare new paper bag with a slightly different face on it), I'd set a stopwatch and make sure I was ready to run around the level searching for the present as soon as it was about to spawn, instead of actually playing the game. It wasn't fun and it kind of ruined the game, and I recognised that but I kept on doing it. And I only got two of the hats.))

If we can fine tune the system so it only gives players incentives to do things that are inherently fun, then I think it'd be great. I just don't think I've seen any suggestions yet for accumulating experience points that achieve that, without also giving incentives to exploit the system by playing in non-fun ways.

Do people agree with this as a general principle? If so, that probably makes it much easier to evaluate specific ideas - imagine a devious player who'll do anything to maximise all their scores and unlock everything as quickly as possible, and then see if they'll be doing non-fun things to achieve their goal, and if so then it's a bad idea. If the principle's wrong or confused and I'm just crazy then I'd like to understand why I'm mistaken

That being said, winning a match will inevitably earn you more experience than simply gathering a bunch of resources and training a bunch of units. Also, experience could only be rewarded to those who do win the match. In the effort of providing incentive to the ultimate goal of winning, perhaps you only earn 50% of the experience points you've accumulated if you lose.

That'll cause the devious player to play easy maps and wait until victory is assured, and then keep playing and rack up experience points by doing whatever grants them. So it'd probably be even less fun than granting the same points regardless of victory, because the player will waste extra time defeating easy opponents.

The experience system wouldn't be the fundamental method of rewarding the player. Winning matches would be what would ultimately reward the player with additional content, not doing the same thing over and over until they earn something.

In that case, why have the experience system at all? For single-player we can reward players for winning (which is always a fun thing for them to do, unless our gameplay is fundamentally broken) with a traditional campaign structure, or with some kind of non-linear conquer-the-world mode where there's multiple paths leading to regions with new maps/enemies/units, or with the more abstract thing you're suggesting where we count number of victories vs each faction. What value would an experience points system add in this case, without distracting the player into doing non-fun things?

[Kimball]

All good points. Given all that, we could just as easily only reward players for victories, at diminishing return for consecutive wins vs. same faction on the same map. The only thing I want to avoid here is only rewarding a player for something once and not telling them that they're not earning anything for the second time around. If we only reward them with new maps/factions the first time they try something new, they have no indication of whether or not they'll earn another if they keep playing. They also have no indication of how well they've done in the grand scheme of things - did they do better than last time? The experience system functions as a visual indicator of their progress. While we inevitably can't prevent players from performing mundane tasks to "level up", we can create failsafes to prevent them from wanting to. Ultimately, we will always have players who grind for something, even without rewards - they'll be working to optimize their speed/strategy/whatever.

I guess my point is, we can never prevent the player from doing things that we don't find inherently fun. Like you said, you ran around the map looking for the present - perhaps the hunt was enough "fun" at the time. Who are we to say that the player grinding for something isn't fun?

[Shield Bearer]

True, good points.

Instead of giving XP we could just give points, say 4 points for playing against a faction on a new map (it'll be 3 for random map) and 2 or 1 for playing the same map multiple times, 10 for playing against a 'new' faction on a 'new' map. Would the amount of units killed upon time played work? say you've killed or destroyed 50 units and buildings in 5 mins, then you would get 10 mini points which if you can get to hundred you get a whole point.

And what about longevity of units? or maybe longevity armies...no that wouldn't work.

[Kimball]

Yeah, maybe destroying the enemy's units instead of creating more of your own would earn experience. Since the AI should be producing units at a relatively constant rate regardless of the player's actions, that seems like a better idea. The only way that could potentially go wrong is if the player intentionally sends just enough units to kill off most of their units/buildings and waits for them to rebuild. Of course, we can't stop the player from torturing the AI - we want them to have fun with it like that.

Now the longevity idea is interesting. How about we reward players for the number of units promoted? Only units who last in combat are promoted to advanced/elite, and I think that could be a great way to ensure that players are motivated to maintain their army. That certainly pushes the player in the right direction.

[Shield Bearer]

Now this just struck me, it seems that killing more units and at the same time trying to keep yours alive will be challenging and will prevent player from doing what u just stated, i.e. ' The only way that could potentially go wrong is if the player intentionally sends just enough units to kill off most of their units/buildings and waits for them to rebuild. Of course, we can't stop the player from torturing the AI - we want them to have fun with it like that.'

[Kimball]

That will depend on the player's abilities and the difficulty of the AI (which I assume could also factor into all this).

[feneur]

I don't have much to add, but I promised to say something after I'd read the document so I'd better do that

To some extent, now that I'v seen your proposal, and read the other comments, I'm almost leaning towards having some kind of titles/ranks merely for their own sake rather than having experience effect unlockables. It seems having won games (perhaps in a sequence, i.e. win map one, then two etc, you can't choose at random) unlock new content is better, and grinding to get titles is more up to the player whether or not they want to spend their time on. That way they don't "have to grind", they just have to win a map (or more) to move on to unlock more content.

Using experience/ranks to help users find good opponents to play with might be a good thing, but being good at accumulating experience points, or even winning games against the AI might not be a good indicator of multiplayer skill. I could be wrong, but I'm guessing having a special set of ranks (might be a bit confusing, unless it's very clear that they're different, perhaps even being referred to as something other than "ranks") or simply displaying number of won/lost matches (probably not too accurate either since you might have won a 100 matches against unskilled people, and might in either case push people towards playing against people of a lower skill than themselves and thus risk scaring away new people as everyone would just think about winning over them as much as possible to boost their stats).

In general I'm not sure we should limit the way users experience the game by locking things at first, but both since it helps with getting them accustomed to the features one or a few at a time (and thus have an easier time learning them), and since it can help make people more likely to play multiplayer games, I think it might be worth it.

I think calculating experience based on things like: number of enemy units killed - own units lost / time spent playing, could help with lowering the risk of players killing of enemy units, then waiting for the enemy to gain more units and continue until the timer runs out. Having more complex ways to count experience points will be more difficult to explain to the player. On the other hand, simple statistics are already viewable via the end game summary screen, so rewarding players for playing the game the way we want them to rather than for accomplishing things we spell out might be a good thing in any case. It would perhaps need quite a bit of testing to make sure we're measuring the correct things/using the correct numbers. Perhaps the above equation would favor rushers, so the "own units lost" might have to be multiplied with some number to make it more worthwhile with a more cautious playstyle etc. Not sure exactly what we should measure to encourage people to do fun things rather than to grind just to get as much points as possible, but I think we need to look into more complex things than just number of units trained or enemy units killed.

Edited January 8, 2011 by feneur

[feneur]

Perhaps this should be posted in either the Development & Technical Discussion forum or the 0 A.D. General and Ideas forum? That way we'd get more discussion, and perhaps gather some useful ideas. (As well as be told if people really doesn't like the idea )

[Kimball]

Agreed, should be good to get some community input on this. Moved.

[Ykkrosh]

The only thing I want to avoid here is only rewarding a player for something once and not telling them that they're not earning anything for the second time around. If we only reward them with new maps/factions the first time they try something new, they have no indication of whether or not they'll earn another if they keep playing.

I think the danger is of user interface challenges caused by using abstract indicators of progress (like experience points (especially when based on complex equations) that accumulate to unlock new features or ranks or badges), so players won't always understand what's going on - but we could avoid those problems entirely by not using those abstract concepts at all, and instead using something like a linear campaign structure or non-linear conquer-the-world mode (or some innovative new game mode) where there's a simple obvious relationship between winning a game and making progress towards a higher goal and unlocking access to new things. I'm still not sure the experience concept has significant benefits to balance the difficulties it introduces.

you ran around the map looking for the present - perhaps the hunt was enough "fun" at the time.

It was okay for the first half an hour, since it was something new. But for the next half a dozen hours it was stupid and pointless and tedious, but the frequent tiny chance of a worthless reward compelled me to continue instead of doing something fun. People aren't rational, and they get addicted to things they know are bad for them, and game designers know enough psychology to trigger that behaviour. It's the designer's responsibility to use their power only for good, by enhancing things players want to do anyway rather than exploiting players to waste time on something unfulfilling.

Who are we to say that the player grinding for something isn't fun?

We're the game designers - saying what's fun is a critical part of our job . We design the gameplay mechanics based on a prediction of what players will find fun or not, and everything else we do is to encourage them to have the fun experience we designed: graphics, UI design, marketing, user manuals, psychological tricks like reward systems, etc, all directing the players towards enjoying the gameplay we built.

Some people will naturally enjoy endless repetitive grinding, but as game designers we can decide that most of our target players won't. (Ideally we'd base that decision on research and testing, rather than just personal opinions). We needn't do anything to discourage the first group from doing what they want, but if we used our tools to direct players towards grinding (e.g. by actively rewarding it in any way) then we'd also be directing all the other players towards the same thing (since our tools don't have the subtlety to distinguish these groups), so we'd be making our typical players have less fun. So I think we should be careful to avoid that