A discussion how to improve GMR, a reward system, more info, and other things

[EEE_BOY]

Do I get an immortal hat, an emperor's crown, or a king's robe?

Semi-seriously, in addition to named levels for your donors, you'd be surprised how much silly little graphical rewards will motivate some people. It's bells and whistles, though, you're probably focusing on more important stuff. Im not sure if it would open anyone's wallet more or quicker...I just know that I've played some games I was pretty bored with and over, just so I could get the next fancy helmet, or whatever it was....

I just had an idea about a GMR point system:

Player gets +1 point for taking a turn;
-1 point for spending two days in vacation mode;
-2 points for spending one day without timer or vacation mode;
-3 points for skipping a turn with a timer;

Player starts with 10 points (settler), and earns points to advance (50 points for chieftain, 250 points for prince, 1250 points for King, and so on...)

In order to earn more points quickly, player needs to upgrade account and join more games. Also player can see if other player is consistently taking turns based on their points collected. Even more seriously, if points dropped below 0, that player should automatically withdraw from all games.



An important summary from GMR admin is post #37 on page 2

 

[Atlas627]

That's so cool that you thought of this! We've actually been planning a system like this for a while now, slated to be implemented with our new community and user profile features coming next month.

We hadn't considered situations that a player might loose points though, only gain them. And the primary way of gaining points would be taking a lot of turns and taking them quickly. We don't have a solid grasp yet on what would be good numbers for it but something like taking your turn within an hour = 40pts, 4-hours = 30pts, 8-hours = 20pts, 24-hours = 10pts, and if it takes you longer than 24hours or you get skipped then you gain 0 points. Also while on Vacation Mode you would gain no points. We'd probably also include bonus points into the equation if the game has more players in it. Since it would be super easy for a couple people to jump in a 2-player game and just spam turns, we'd want to keep it appealing to participate in larger games as well, where turns come less frequently.

And the point system would have many levels of prestige, probably starting at "Settler" and increase on some curve up to "Diety" or beyond. Anyway it's all just theory right now, but we'd love to hear any ideas or thoughts you guys have!

Remember, not everyone intends to play once per day either. So it would be unfair to not reward people for playing how they wish too, and it certainly wouldn't be fair to punish (reduce points) for playing how they wish to!

Simply give 1 point per turn taken. Multiply that by the number of players in the game (or that number squared) and give that many points to the player. In this way they are rewarded for playing, not for playing quickly or slowly, but just for playing. Since your one turn lets 7 other players take theirs sooner, it is worth more points to enable 7 people's fun than it is to enable only 1 other player. And the squared part is to compensate for the fact that you simply cannot control all 7 other players to play quickly to enable you, whereas only 1 person can easily get the turn back to you without trying particularly hard.

Of course, prizes will have to use inflated numbers with this scheme.

Edit: And of course if you get skipped because the host (and players) agreed on a time limit, then you get 0 points. You didn't play.

 

[m4stermel]

Remember, not everyone intends to play once per day either. So it would be unfair to not reward people for playing how they wish too, and it certainly wouldn't be fair to punish (reduce points) for playing how they wish to!

Simply give 1 point per turn taken. Multiply that by the number of players in the game (or that number squared) and give that many points to the player. In this way they are rewarded for playing, not for playing quickly or slowly, but just for playing. Since your one turn lets 7 other players take theirs sooner, it is worth more points to enable 7 people's fun than it is to enable only 1 other player. And the squared part is to compensate for the fact that you simply cannot control all 7 other players to play quickly to enable you, whereas only 1 person can easily get the turn back to you without trying particularly hard.

Of course, prizes will have to use inflated numbers with this scheme.

Edit: And of course if you get skipped because the host (and players) agreed on a time limit, then you get 0 points. You didn't play.

Great feedback! It does make sense to not create a built-in expectation for how long it should take someone to play a turn (that's what the turn-timer is for). And I do agree that completing a turn in a game with more players should reward more points. The scale and math for it all and when prizes are unlocked can be worked out in the details, but the core idea seems great!

 

[EEE_BOY]

That's so cool that you thought of this! We've actually been planning a system like this for a while now, slated to be implemented with our new community and user profile features coming next month.

We hadn't considered situations that a player might loose points though, only gain them. And the primary way of gaining points would be taking a lot of turns and taking them quickly. We don't have a solid grasp yet on what would be good numbers for it but something like taking your turn within an hour = 40pts, 4-hours = 30pts, 8-hours = 20pts, 24-hours = 10pts, and if it takes you longer than 24hours or you get skipped then you gain 0 points. Also while on Vacation Mode you would gain no points. We'd probably also include bonus points into the equation if the game has more players in it. Since it would be super easy for a couple people to jump in a 2-player game and just spam turns, we'd want to keep it appealing to participate in larger games as well, where turns come less frequently.

And the point system would have many levels of prestige, probably starting at "Settler" and increase on some curve up to "Diety" or beyond. Anyway it's all just theory right now, but we'd love to hear any ideas or thoughts you guys have!

Right, I haven't considered the case for larger games. Maybe 1 point for game with 2-4 players, 2 points for 5-7, 3 points for 8-10, 4 points for 11 or more. AI should be excluded, so that game starts with 8 player will drop to 2 points if one player quits or defeated.

I don't think pushing players to take turns quickly within a hour or 2 will be necessary. 8-hour is acceptable, up to 24 hour is a bit long but ok if real life calls, beyond 48 hour is too long so some penalty points are reasonable. When more games are in progress, one does not wish to hurry all of them, but when only 2 games are tried, it felt the turn is very slow. So the fine balance should be adjusted. Edit: trial player should be limited to game less than 6 players, why? Because they have only 2 games to play, and if turns are longer the experience is not good at all, once they got more games or points, larger game then starts to provide better experience. but the reason should be explained to new player clearly.

I think penalty points are necessary to get a whole view of playing status. Right now we had several cases when a player signed up to multiple games and then left them unplayed. In that case, the system will help us deduct points, so the more games left unplayed, the quicker points dropped, to a point maybe a week or two the player should automatically withdraw all games (right now the host has to bear the moral risk of kicking or not kicking a player). Surrendering should also attract some penalty but it may be more complicated.

Penalty points can also scale according to game size, because more players are stalled in larger games. Gifted points can be given according to account status, so trial account can receive points enough to skip 2 large games, say for a week (assuming deducting points every 2 days, no timer). Premium account can receive points enough points for skipping, say 15 large games for a week.

Also retrospective points can be calculated for turns played so far if possible. Right now we tend to avoid new players because we don't know if we should bear the risk of losing him/her in the future, which is not good for increasing the GMR player base. And we ended up playing most of games with players that we are familiar with. If an unfamiliar player have earned some points in other games, it at least gives some kind of warrant that the risk of leaving is relatively low.

 

[EEE_BOY]

That reminds me to report a potential bug: is the turn timer working on players set as vacation mode right now?

We had players appearing as AI in the games, so we assume he is in vacation mode, but the turn timer has to expire before the turn skipped. Shouldn't the player skip turns directly if on vacation mode when turn timer is set? Also is it possible to make the next player host if the host left the game unplayed in vacation mode and no one can modify the settings? ( that's another reason points should be deducted to allow host to withdraw after certain limit)

 

[Atlas627]

Right, I haven't considered the case for larger games. Maybe 1 point for game with 2-4 players, 2 points for 5-7, 3 points for 8-10, 4 points for 11 or more. AI should be excluded, so that game starts with 8 player will drop to 2 points if one player quits or defeated.

I don't think pushing players to take turns quickly within a hour or 2 will be necessary. 8-hour is acceptable, up to 24 hour is a bit long but ok if real life calls, beyond 48 hour is too long so some penalty points are reasonable. When more games are in progress, one does not wish to hurry all of them, but when only 2 games are tried, it felt the turn is very slow. So the fine balance should be adjusted. Edit: trial player should be limited to game less than 6 players, why? Because they have only 2 games to play, and if turns are longer the experience is not good at all, once they got more games or points, larger game then starts to provide better experience. but the reason should be explained to new player clearly.

I think penalty points are necessary to get a whole view of playing status. Right now we had several cases when a player signed up to multiple games and then left them unplayed. In that case, the system will help us deduct points, so the more games left unplayed, the quicker points dropped, to a point maybe a week or two the player should automatically withdraw all games (right now the host has to bear the moral risk of kicking or not kicking a player). Surrendering should also attract some penalty but it may be more complicated.

Penalty points can also scale according to game size, because more players are stalled in larger games. Gifted points can be given according to account status, so trial account can receive points enough to skip 2 large games, say for a week (assuming deducting points every 2 days, no timer). Premium account can receive points enough points for skipping, say 15 large games for a week.

Also retrospective points can be calculated for turns played so far if possible. Right now we tend to avoid new players because we don't know if we should bear the risk of losing him/her in the future, which is not good for increasing the GMR player base. And we ended up playing most of games with players that we are familiar with. If an unfamiliar player have earned some points in other games, it at least gives some kind of warrant that the risk of leaving is relatively low.

Is this a system meant to protect the enjoyment of 1/day players from leavers? Because then it shouldn't apply to the GMR app. If that is what the system is for, then that is what NQ and Civplayers are for. They record these things and get player feedback.

If you are going to apply this to GMR, whose only goal is to enable the enjoyment of PBEM Civ5, then at least make sure it does not penalize players who use GMR but are NOT 1/day players.

One thing that my system (no penalties) does not do is differentiate between new players and bad players. But if you record the number of games joined as well, then it is obvious when a player has no points because they have never played, compared to having no points because they joined and dropped several games.

 

[EEE_BOY]

Is this a system meant to protect the enjoyment of 1/day players from leavers? Because then it shouldn't apply to the GMR app. If that is what the system is for, then that is what NQ and Civplayers are for. They record these things and get player feedback.

If you are going to apply this to GMR, whose only goal is to enable the enjoyment of PBEM Civ5, then at least make sure it does not penalize players who use GMR but are NOT 1/day players.

One thing that my system (no penalties) does not do is differentiate between new players and bad players. But if you record the number of games joined as well, then it is obvious when a player has no points because they have never played, compared to having no points because they joined and dropped several games.

I've never intended the 1 day limit. If desired, 3 day or 5 day is fine, as long as there is a lower limit and player takes their turns. You have to admit that there is a need to address some problems that host left the game in vacation mode and spoil the whole game. If left in vacation mode for too long is actually worse than having an AI, because human player in vacation mode will build nothing!

What I meant is, the point system can serve multiple purposes, and it should provide last resort to keep games going and provide rough estimation on whether an unfamiliar player is willing to commit to take the turns consistently. It can relax on the time limit to 3-5 days, but the limit has to be there otherwise many games are just wasting time and providing bad experience for the rest of the players.

Lastly on your final argument, a no-penalty system can't tell if a player took many turns before but suddenly left, points are still there but many games are spoiled! What I really want is to indicate that from point system (maybe point growth statistics in the past week? Negative growth means the player being irresponsible). Anyway, I never intend to push players taking turns faster.

 

[Atlas627]

I've never intended the 1 day limit. If desired, 3 day or 5 day is fine, as long as there is a lower limit and player takes their turns. You have to admit that there is a need to address some problems that host left the game in vacation mode and spoil the whole game. If left in vacation mode for too long is actually worse than having an AI, because human player in vacation mode will build nothing!

What I meant is, the point system can serve multiple purposes, and it should provide last resort to keep games going and provide rough estimation on whether an unfamiliar player is willing to commit to take the turns consistently. It can relax on the time limit to 3-5 days, but the limit has to be there otherwise many games are just wasting time and providing bad experience for the rest of the players.

Lastly on your final argument, a no-penalty system can't tell if a player took many turns before but suddenly left, points are still there but many games are spoiled! What I really want is to indicate that from point system (maybe point growth statistics in the past week? Negative growth means the player being irresponsible). Anyway, I never intend to push players taking turns faster.

I have not had experiences with dishonorable/unreliable players, so I cannot comment on just how infuriating that may be. But I do know that I have fun while playing, even if a game that I have worked on for a long time were suddenly dropped due to one player. Would that be frustrating? Yes. But I still had fun for all the time I spent on it.

If you want to encourage players to stick with games, why not make later turns worth more points?

But whatever you do, you cannot unilaterally penalize players for surrendering games. Many times you surrender because the game is over, or because the game became unplayable due to bugs, or because you accidentally made the game incorrectly.

And if 2 friends want to play a game together where they take turns once every week, they should not be penalized for that. What if one of them decides he wants to play with others in a 1/day format, and his rating is already tanked?

 

[EEE_BOY]

I have not had experiences with dishonorable/unreliable players, so I cannot comment on just how infuriating that may be. But I do know that I have fun while playing, even if a game that I have worked on for a long time were suddenly dropped due to one player. Would that be frustrating? Yes. But I still had fun for all the time I spent on it.

If you want to encourage players to stick with games, why not make later turns worth more points?

But whatever you do, you cannot unilaterally penalize players for surrendering games. Many times you surrender because the game is over, or because the game became unplayable due to bugs, or because you accidentally made the game incorrectly.

And if 2 friends want to play a game together where they take turns once every week, they should not be penalized for that. What if one of them decides he wants to play with others in a 1/day format, and his rating is already tanked?

Alright alright, guess we covered the two extreme cases of playing GMR turns.

If points are not to be deducted, I think point growth over the past 7 days serves the similar purpose. In other words, two stats: total points earned, recent points earned. This way one can easily tell whether new player or dropped player. And it's much simpler to implement.

 

[Atlas627]

Alright alright, guess we covered the two extreme cases of playing GMR turns.

If points are not to be deducted, I think point growth over the past 7 days serves the similar purpose. In other words, two stats: total points earned, recent points earned. This way one can easily tell whether new player or dropped player. And it's much simpler to implement.

How can you tell if someone used to play a lot, but simply doesn't have time? And then wants to rejoin?

This seems like a fine system if you fix the above problem, which could be done by adding a "date of most recent game". Not turn played, but game they are "registered for".

Of course, someone could be an idiot and never surrender out of their last game in the scenario I described, but that seems like an acceptable compromise to ensure the enjoyment of everyone else.

 

[EEE_BOY]

How can you tell if someone used to play a lot, but simply doesn't have time? And then wants to rejoin?

Well, the two stats method can still tell: a lot of total points indicate a player used to play a lot; but 0 recent points collected over the past 7 days also indicate the player does not have time to play now (or skipping turns a lot lately depends how you interpret it). Nothing prevents the player to rejoin.

I don't understand your second paragraph, which sounds more complicated info than adding points

 

[PrimEval]

For this discussion of points, the way I see it there are two goals here:
1) Help people who want to play "fast" turn timer games find reliable opponents
2) Hand out fun titles as rewards

For #1, the important information for each player is:
a) average time to play turn across ALL turn timer games
b) average number of skips across your turns in ALL turn timer games
c) total number of turns played
I don't think it makes sense to penalize going into vacation mode, these games will go a long time and sometimes you need to take a vacation And while I have never used vacation mode, I assume it will make for even faster turns

For #2, you don't need a universal score, you could have a monthly hall of fame for things like:
Most turns played this month, and
Fastest average time to play turn (you could break this into categories eg: players in 2 or less games and players in more than 2 games)

 

[EEE_BOY]

For this discussion of points, the way I see it there are two goals here:
1) Help people who want to play "fast" turn timer games find reliable opponents
2) Hand out fun titles as rewards

For #1, the important information for each player is:
a) average time to play turn across ALL turn timer games
b) average number of skips across your turns in ALL turn timer games
c) total number of turns played
I don't think it makes sense to penalize going into vacation mode, these games will go a long time and sometimes you need to take a vacation And while I have never used vacation mode, I assume it will make for even faster turns

For #2, you don't need a universal score, you could have a monthly hall of fame for things like:
Most turns played this month, and
Fastest average time to play turn (you could break this into categories eg: players in 2 or less games and players in more than 2 games)

1. Not fast turns. But consistently taking turns.
(a)(b) we have those info in local game level, so repeating is unnecessary.

2. Disagree on hall of fame. We want to see profiles of the player base, not a few GMR fanatics.

Staying in vacation mode for a few days is fine, but leaving it in vacation mode for weeks is another issue worse than surrendering.

 

[EEE_BOY]

Randomly showing Steam ID in game setup window

I also got another idea, given that sub is relatively hard to find and current method is inefficient.

Is it possible to pull player Steam ID from GMR database, who have free game slots left to fill. And randomly displayed in the game setup window for new game waiting to start or started game having player dropped. It can display 3-5 players at a time and randomly changing every 20-30 seconds.

This way the host can click an unfamiliar player's profile and send invitation. To prevent spam invitation, the invitation limit can be twice the game player setting for a given period.

I don't know if there is privacy issue to display Steam ID, but player should have the option to opt out displaying Steam ID in the global settings.

 

[Atlas627]

Well, the two stats method can still tell: a lot of total points indicate a player used to play a lot; but 0 recent points collected over the past 7 days also indicate the player does not have time to play now (or skipping turns a lot lately depends how you interpret it). Nothing prevents the player to rejoin.

I don't understand your second paragraph, which sounds more complicated info than adding points

Does he have no recent points because he is irresponsible or because he simply cannot play? The "depends on how you interpret it" is exactly my point. It should not be ambiguous.

If you see the end date of the most recently played game, in addition to the above information, then you can differentiate between "hasn't been playing" and "drops games".

 

[EEE_BOY]

differentiate between "hasn't been playing" and "drops games".

Basically the same effect, shouldn't gone too much into causes, and it's impossible to enquire further.

 

[Atlas627]

Basically the same effect, shouldn't gone too much into causes, and it's impossible to enquire further.

The problem is not that people don't have time to play, the problem is that they tell you that they do in the first place. So the system should not keep record of whether or not they have time to play in general, but rather it should keep track ONLY of those times where the game was dropped because the player "lied" (or was mistaken) in the first place.

Just because the player did not have time in the past and now they say they do does NOT mean they are lying, or are irresponsible/mistaken. A past history if being mistaken can tell you that, but not a past history of not playing.

 

[EEE_BOY]

The problem is not that people don't have time to play, the problem is that they tell you that they do in the first place. So the system should not keep record of whether or not they have time to play in general, but rather it should keep track ONLY of those times where the game was dropped because the player "lied" (or was mistaken) in the first place.

Just because the player did not have time in the past and now they say they do does NOT mean they are lying, or are irresponsible/mistaken. A past history if being mistaken can tell you that, but not a past history of not playing.

I think the developer is keen to increase the player base, improve existing players' experience and hopefully help their cost of providing the service. If you look at from that perspective, a 'detective' system will not help that much. If you don't want to punish slow player, it shouldn't punish player who might lose interest after a while or whatever other reasons might be. Let's wait and see how developer will implement that. Currently I'm very happy about the existing implementations we've got

 

[Onan]

I agree that a reputation system would be helpful for some GMR players, not all. And that it should be rewards based, without any punitive side.

The best reputation system we can have for consistency and/or quick turn taking is going to be what we have already: word of mouth and reputations made in others players minds via games and forums.

The developers are balancing the interests of all their users, not catering to some few 'hardcore' GMR players (us). So I'd expect to see something very general and fun, which is a good thing.

 

[PrimEval]

What I envision is having an option to say that you are looking for games to join. This would be off by default, but if you select it then your name and stats are added to a list of players looking for a game. Hosts can then invite people from this list(in addition to what can be done now).
Your 'stats' help convey what type of GMR play you are, this is important if the host is looking for a certain kind of game (eg fast turns), but the qualities which are considered good are totally up to the players involved. I think stats, rather than some score would be better at conveying this information.

As I posted earlier, I think the important information for each player is:
a) average time to play turn across ALL turn timer games
b) average number of skips across your turns in ALL turn timer games
c) total number of turns played

EEE_BOY suggest that vacation mode is also a factor some people might be concerned about.
so: d) average number of vacation mode turns across ALL turn timer games

People who play without the turn timer should get a N/A in the turn timer categories (a, b, d) but still build credibility with c [in the arbitrary eyes of our imaginary host ]

@EEE_BOY, we have info for turn time and skips at a per game level, what I want to see on the 'player profile' is the average for this across all a players turn timer games.

It would also be super cool to see a heat map of the time of day when we take our turns, this would make setting an optimal turn order easier!