Proposed 2-tier GOTM System

[da_Vinci]

Moderator Action: I've moved these posts (no's 1-12 of this thread) out of the GOTM Brainstorming thread to allow for a dedicated discussion of the 2-tier system that a few people have proposed in that thread - DS

Same here. I can't imagine a better system. Let's say, two difficulty levels.

Tier 1 games that rotate emperor, immortal, diety, and Tier 2 games that rotate noble, prince, monarch?

Now, make the WOTM monthly ... (does the larger staff allow?)

GOTM: Emperor/Noble, then Immortal/Prince, then Diety/Monarch

BOTM: Diety/Monarch, then Emperor/Noble, Immortal/Prince

WOTM: Immortal/Prince, then Diety/Monarch, then Emperor/Noble

This gives us a game at each difficulty every month (for those with all three versions).

culture: Something for everyone, a comedy tonight!)

I would probably play the emperor games and the immortal games, and skip the diety games for now. Instead, I would play the tier 2 monarch game, but not be allowed to compete for tier 2 awards. Rather, I play it to post a spoiler and be aware of the game to comment on spoilers for the tier 2 players.

If feasible, a system like this accomplishes a lot: higher difficulty games more often for the elite, lower difficulty games more often for the newcomers who may need or want them, a second tier of awards to incentivize newcomers and begining players. And to the extent that emperor level players might skip the diety in tier 1 and play monarch in tier 2, you have ready made coaches to advise the tier 2 group.

Seems this would require assigning players to tiers 1 and 2 ... as some benefits don't accrue if tier 1 players are allowed to COMPETE in tier 2 (they can play if they want ... ). For established players, perhaps a winning record on emperor in submitted games puts you in tier 1, or any victory on immortal or higher. One can always self-assign to tier 1. Tier 1 only competes in tier 1, can play tier 2 but not eligible for tier 2 awards, and tier 2 games are not factored into global rankings for tier 1 players (or not into tier 1 rank if we split the rankings ... see below)

Other players can be tier 2. They can play in tier 1 and be eligible for awards there. They can compete in tier 2. If they achieve winning record on emperor (over 5 submitted emperor games), any win on immortal or higher, any three medals or fastest awards in tier 2, they promote out of tier 2 into tier 1 (solves the problem of a tier 2 eptahlon, doesn't it?). The time one can spend dominating tier 2 can be adjusted based on our experience with the system.

This would also require segretating global rankings into Tier 1 and Tier 2, and has implications for eptathalon (only in Tier 1? or now have it in both Tiers? How to handle those partway there?).

This approach seems to synthesize several of the ideas presented in this thread: incentives for elite and for new, tiers of play, a ranking system (although just a simple two-cagegory system). The ideas of gameplay challenges, spoiler incentives or awards, etc. can always be added to such a system.

Feasibility will depend on how much of the added tracking work can be done by the system on auto-pilot, once the programming is set up. Six maps a month sounds large, but we already have 6 to 9 starting saves a month ... (oh, I think implicit in this is to eliminate adventurer and challenger saves)

dV

 

[Frederiksberg]

I like this tier system. If something like this was implemented I wouldn't mind if the adventurer and challenger saves disappeared. It would work well in combination with a rating system replacing the global rankings. Things would be much simpler. Sometimes less is more...

 

[Ronnie1]

I think the 2 tier system sounds like a great idea if it is feasible for the staff. That is the ONLY thing I would be worried about. The staff here already does such a great job, I would hate to make their job more difficult.

Adrianj's idea is very interesting also, kinda like the WOTM from a few months back IIRC, which was a fun game!

 

[da_Vinci]

I like this tier system. If something like this was implemented I wouldn't mind if the adventurer and challenger saves disappeared. It would work well in combination with a rating system replacing the global rankings. Things would be much simpler. Sometimes less is more...

The "two difficulties" idea would help a lot for this. If you get to play a level similar to your skill level you'll be more interested in discussing the game post-mortem as well--whereas cakewalk victories aren't that inspiring.

I would play regularly if this was implemented.

I think the 2 tier system sounds like a great idea if it is feasible for the staff. That is the ONLY thing I would be worried about. The staff here already does such a great job, I would hate to make their job more difficult.

Oooh! ... is there a groundswell brewing? If Rusten plays regularly, there's another spot on global ranking I would lose ... This tier system WOULD put me back where I belong ...

Absolutely depends on feasibility. As I said ...

Feasibility will depend on how much of the added tracking work can be done by the system on auto-pilot, once the programming is set up. Six maps a month sounds large, but we already have 6 to 9 starting saves a month ... (oh, I think implicit in this is to eliminate adventurer and challenger saves)

dV

Currently in WOTM months we have 3 maps and nine saves (map modified by unit or tech additions or subtractions, generally). Tiers without challenger or adventurer saves would be three maps, six saves (modifed by difficulty level). Is that less, same or more work than currently?

Some added work in having to validate the entries for 0.5 more games a month ... and if successful, more entries per game. 0.5 more results threads to post ...

The real question is how much added work a two category ranking system is to maintain. My hope is that after initial programing, it kind of runs on autopilot. But I have no expertise to know for sure.

Wonder if these ideas export to the Civ III series?

dV

 

[ngraner42]

The real question is how much added work a two category ranking system is to maintain. My hope is that after initial programing, it kind of runs on autopilot. But I have no expertise to know for sure.

There of course is a chunk of start up work, but if the same map is used for both tiers, it should be less work to maintain than the current 3 save system. I like the idea and know several people who are put off by the higher level games. It seems like the more winable games generate more submissions, but the high level games generate more discussion. By offering both it seems like we could increase participation.

I personally like seeing interesting variations thrown in, such as we have seen in the SGOTM series. Jesuin's barbarian city map was great. On the dislike side, I hate always war and am not too fond of raging barbarians; these games seem tedious.

 

[Munro]

It seems that we might be limiting game variety for tier-one players. Must they always play only Emperor, Immortal and Deity games or can they drop down and also play Monarch, Prince or, perhaps, an occasional Noble game? Should they do so, what happens to their ranking?

I'd agree. I can't really see any reason to restrict it (and it is all for fun after all ).

For the awards and rankings, I'd have thought we could keep things pretty simple.

• Each month, just let people compete in whichever tier they want.
• Have two sets of awards.
• Have a sensible tier 1 multiplier (per-game) for the (combined) global rankings.

This provides a built-in incentive for people to move up when ready (better global rankings and shinier awards ). While also allowing people the flexibilty to play whichever tier is most suited to their tastes each month. (Deity is not for everyone).

The only thing to be wary of, I think, is any kind of tier 2 heptathlete award. While the awards are nice to have (so tier 2 awards in general are good) artificial incentives to continue playing in the lower level (just to sandbag the remaining awards after you are already ready to move on) are probably not a good thing...

Obviously, the tier 1 ranking modifier will help with that too. Just need to make sure that it is high enough so that, for example, a low tier 1 win (or narrow loss) will always rank above a high 2 tier win (it is three difficultly levels harder after all). Could always just look at some past GOTM scores at different difficultlies for an idea of what this might need to be set to... (and / or tweak as appropriate after a month or two).

(All that said, I'm not really that bothered about the rankings. I'd just like some harder games each month! Fingers crossed that this two-tier idea makes it to fruition... )

 

[da_Vinci]

Two quick thoughts ...

Tier 1 players being eligible for awards in tier 2 seems to defeat one of the reasons for tier 2 ... awards that the tier 2 players can actually win.

Also, the scoring system makes such a huge point difference between a win and a loss ... someting like tenfold (just how does the score recognize a narrow loss?), that I don't see how one can make a tier one loss end up better than a tier 2 win ...

Separate tier 1 and 2 ranks might be unavoidable. Unless a ratings system proves workable.

dV

 

[AlanH]

I don't see a single ranking system as practical in a 2-tier system. How would we decide the score multiplier for each game? What would we do about finish dates for speed rankings? Many players consider the speed awards and rankings as the more important measures of skill than scores. (That's speed as in game date, not playing time!)

 

[azzaman333]

One thing I don't like about a 2 tier system is that it won't get me to play games above my comfort zone. If I have a Prince/Monarch game every month or so, I doubt I'd play any of the games that would appear too hard for me. One of the big plusses of the current system is that I will have games that I will struggle in, but that just means I can see where my weaknesses lie in trying to move up levels.

 

[Frederiksberg]

Some questions:
It seems that we might be limiting game variety for tier-one players. Must they always play only Emperor, Immortal and Deity games or can they drop down and also play Monarch, Prince or, perhaps, an occasional Noble game? Should they do so, what happens to their ranking?

Would we need to have two ranking systems or can we maintain a single, combined, ranking system? I have been reading up on ranking systems and, using some version of ELO, we may be able to use one, combined, system. Any suggestions?

We need not think in terms of tiers. Most of the ideas revolve around having one game but with two saves of different difficulty. If they are always paired Deity/Monarch, Immortal/Prince and Emperor/Noble (As suggested by dV) it should be possible for most people to select a game that is interesting for them. I would suggest that any player can play any save he wants (obviously not both). So you can select Prince one month and Emperor the next if you like.

On the ranking issue I would suggest to start using a rating system. There's no doubt that it can be implemented (I have already outlined an option) and it would give us an opportunity to create a system that doesn't favor participation as much as the present system. It is in our interest to reward participation, but not to an extent where the rankings start looking heavily biased. A rating system shouldn't have any problems with the dual difficulty saves because they can be rated independently. Only problematic case would actually be if there is a strict separation between the players so that the "tier1" players never compete against "tier2" players. This is one more argument in favor of letting people chose the difficulty freely.

 

[da_Vinci]

We need not think in terms of tiers. Most of the ideas revolve around having one game but with two saves of different difficulty. If they are always paired Deity/Monarch, Immortal/Prince and Emperor/Noble (As suggested by dV) it should be possible for most people to select a game that is interesting for them. I would suggest that any player can play any save he wants (obviously not both). So you can select Prince one month and Emperor the next if you like.

On the ranking issue I would suggest to start using a rating system. There's no doubt that it can be implemented (I have already outlined an option) and it would give us an opportunity to create a system that doesn't favor participation as much as the present system. It is in our interest to reward participation, but not to an extent where the rankings start looking heavily biased. A rating system shouldn't have any problems with the dual difficulty saves because they can be rated independently. Only problematic case would actually be if there is a strict separation between the players so that the "tier1" players never compete against "tier2" players. This is one more argument in favor of letting people chose the difficulty freely.

Certainly, the two difficulties per map can be separated from tiers, as the two ideas seek to achieve different goals: one goal being more games at both high and low difficulty; other goal being awards and incentives for the newcomers/beginners.

If tiers have the "drawback" of limiting tier 1 players to high difficulty games, and that is a problem, then we could expand the game range in the tiers to also include a month of Monarch/Warlord into the mix. Would the eptathletes really want to play lower than monarch difficulty?

Note that if a game has a major variant, difficulty might not match stated level (like always war) and some alteration of the schedule might be warranted

Perhaps formal tiers are not necessary, but an implicit tiering of awards would exist based on game selected. In the absence of tiers, who gets the award for fastest finish diplo, between 1700 in the emperor game, and 1500 in the noble game? Maybe the answer is a high diff speed award and a low dif speed award for each VC? Also, a set of medals for high dif and low diff? This without assigning players to tiers.

The one drawback of this is that if enough top players play down often enough, we lose the incentive of reachable awards for the developing players.

Now, in chess tournaments (at least when I used to play), everyone is in same tourney, but awards are given within brackets of ratings. So might there be room for an "honorable mention" of highest 3 scores and fastest finishes for Recruits, and for Veterans (in a system where we separate folks into Recruits, Veterans, and Elites based on ratings) ... keeping one set of awards for the game?

Which still does not solve the "1700 in the emperor game, and 1500 in the noble game" issue. If elite players playing down get the speed awards, how does that dilute the value of the eptathlon? Might be moot if the elite players do better with higher difficulty for speed (need advanced AI to get useful trades ... )

Might not be able to satisfy all of the competing interests in one manageable system ...

dV

 

[Niklas]

Since this thread seems to change shape from brainstorming to actual discussion, I suppose some critique is now allowed.

I'm not so keen on an actual two-tier system, for various reasons. Perhaps the strongest of those is that I think it's a sledgehammer approach for solving a problem for which there are other, better and less intrusive, solutions.

Looking from my own perspective, I consider myself a solid Immortal-level player, and thus a "tier-1" player. I've never lost on Emperor, I mostly win on Immortal, I have yet to win on Deity (but then I haven't tried it much). And yet I definitely still enjoy playing the odd leisurely game on Prince, knowing that it's not the AI I'm competing with but the other players playing the same game. Who can best exploit the way the game plays on Prince is a different - but no less challenging - competition than who can beat the Deity AI. So I would not appreciate a system that locked the "tier-1" games into Emperor and above.

I also think azzaman's comment is spot on for those still struggling on the mid-range difficulty levels. The current system is good in that it gives everyone a chance to play slightly above their level, to get a feel for how the game plays at that level, and the chance to compare with those who might be way better and those who might be just enough better to manage a win. I think there's lots of learning to be had from that.

I would very much prefer a system closer to what we already have, but where we use difficulty levels as equalizers as has been proposed several times in this thread. If we cycle the "Contender" difficulty as (W-)N-P-M-E-I(-D), like we do now, we could use something like (P-)M-E-E-I-D(-D) or (P-)M-E-I-I-D(-D) for the "Challenger", and (W-)W-N-P-P-M(-E) or (W-)W-N-N-P-M(-E) for the "Adventurers". If we go with the latter suggestion, that would mean that out of every 12 games (for a given xOTM series) there would be:

* Challenger: 3 Deity, 4 Immortal, 2 Emperor, 2 Monarch and 1 Prince.
* Contender: 1 Deity, 2 Immortal, 2 Emperor, 2 Monarch, 2 Prince, 2 Noble and 1 Warlord.
* Adventurer: 1 Emperor, 2 Monarch, 2 Prince, 4 Noble and 3 Warlord.

Looks pretty darn perfect to me, at least for the two upper levels, the lowest might need some adjustment.

 

[DynamicSpirit]

Just to give some context, the GOTM staff are likely to experiment soon with providing varying difficulty levels for adventurer/challenger - a slightly different idea that also emerged from the brainstorming thread. In that idea, instead of adding bonuses/handicaps to the gamefor the adventurer/challenger saves, the game would be released with adventurer on a slightly easier level, challenger on a slightly harder, level than contender. We are still discussing the idea so I can't yet promise what/when/if anything will happen, as there's a fair bit of backend work on our systems that Alan will have to do to implement it. Most of our staff discussion has been in the context of a 1-level difference, eg. if the main game is on prince then adventurer would be on an identical map but on noble, and contender would be on monarch; though I see that Niklas in the previous post is suggesting a larger difficulty difference on some games.

OTOH if we do go with the two-tier system there's a good chance that that would mean adventurer and challenger saves would be scrapped as otherwise it'll just get too complicated. So to that extent, it might be wise to regard the two systems on offer as mutually exclusive alternatives

 

[da_Vinci]

I'm not so keen on an actual two-tier system, for various reasons. Perhaps the strongest of those is that I think it's a sledgehammer approach for solving a problem for which there are other, better and less intrusive, solutions ...

And yet I definitely still enjoy playing the odd leisurely game on Prince, knowing that it's not the AI I'm competing with but the other players playing the same game. Who can best exploit the way the game plays on Prince is a different - but no less challenging - competition than who can beat the Deity AI. So I would not appreciate a system that locked the "tier-1" games into Emperor and above ...

I would very much prefer a system closer to what we already have, but where we use difficulty levels as equalizers as has been proposed several times in this thread.

If the immortal level players enjoy playing the lower levels, then tiers make less sense. I wonder what the distribution is between those who enjoy the lower difficulty games (Niklas, maybe Frederiksberg? ...) and those who seem to want more higher difficulty games (Munro, Rusten, ...)

So you are still suggesting a one map, two or three difficulty level solution, right? In which case the game has tiers but the players do not ... an important distinction.

For one map, would you envision difficulties that are adjacent (Monarch, Emperor, Immortal for one map, for example)? That is a different idea than my Diety/Monarch split, which gives a wider range between the extremes of difficulty, and thus a bigger problem if awards are not tier specific. Your narrower range may alleviate some of the award issues.

Your point of exploiting the lower AI differently is well taken ... as I am finding that as I play lower difficulty than where I now usually win, while the game is always well in hand, and I have big leads in scores and techs, I don't seem to actually get to that diplo win any faster ...

dV

 

[Frederiksberg]

Perhaps formal tiers are not necessary, but an implicit tiering of awards would exist based on game selected. In the absence of tiers, who gets the award for fastest finish diplo, between 1700 in the emperor game, and 1500 in the noble game? Maybe the answer is a high diff speed award and a low dif speed award for each VC? Also, a set of medals for high dif and low diff? This without assigning players to tiers.

I think you are right that awards have to follow the game difficulty. Conquest and domination is obviously faster to achieve in a Noble game compared to an Emperor game while it's the other way around for Space Race.

Now, in chess tournaments (at least when I used to play), everyone is in same tourney, but awards are given within brackets of ratings. So might there be room for an "honorable mention" of highest 3 scores and fastest finishes for Recruits, and for Veterans (in a system where we separate folks into Recruits, Veterans, and Elites based on ratings) ... keeping one set of awards for the game?

Awards to players who perform much better than their rating predicts (exceed the expected score) is also seen frequently in chess tournaments. The use of a rating system obviously facilitates this.

Which still does not solve the "1700 in the emperor game, and 1500 in the noble game" issue. If elite players playing down get the speed awards, how does that dilute the value of the eptathlon? Might be moot if the elite players do better with higher difficulty for speed (need advanced AI to get useful trades ... )

In my view Eptathlon should always require that your awards are from the highest difficulty save. That way we preserve the Eptathlon as something out of the ordinary .

I would very much prefer a system closer to what we already have, but where we use difficulty levels as equalizers as has been proposed several times in this thread. If we cycle the "Contender" difficulty as (W-)N-P-M-E-I(-D), like we do now, we could use something like (P-)M-E-E-I-D(-D) or (P-)M-E-I-I-D(-D) for the "Challenger", and (W-)W-N-P-P-M(-E) or (W-)W-N-N-P-M(-E) for the "Adventurers". If we go with the latter suggestion, that would mean that out of every 12 games (for a given xOTM series) there would be:

If I understand you correctly your suggestion is actually very close to what has already been outlined by various people including myself and da_Vinci. The main differences are:

1. Saves of 3 difficulty levels (instead of 2).
2. And only one common set of awards.

I rather like to have saves with only 2 difficulty levels for greater comparability. More people will be playing the exact same game with only 2 saves. I'm not sure I can see any compelling argument for having the 3rd save. If we want to include a warlord game then we could add Monarch/Warlord to the 2 game cycle as dV suggests.

If the awards are common there is a risk that people going for Eptathlon will select the low difficulty save. The logical remedy would then be to allow only one save to be used in pursuit of awards. Now, we are already moving away from the common awards principle. And we are getting close to the old system where the challenger save was almost never chosen - for two reasons, I believe: One is that people going for eptathlon preferred the easier Contender save and the other is that most people like to play a game that is comparable to other peoples games and hence they chose the Contender save which is by far the most popular.

 

[Niklas]

I rather like to have saves with only 2 difficulty levels for greater comparability. More people will be playing the exact same game with only 2 saves. I'm not sure I can see any compelling argument for having the 3rd save.

I agree that comparability is the main reason for not adopting this system. But the compelling argument I see for having 3 saves is that it lets us cater to more tastes with the same game. It would allow us to run more Immortal+ games over a year, while at the same time still having lower-level games available for those who prefer them.

It also makes the choice to "take the leap" less dangerous for those that want to move up. Chosing Contender (say Monarch) over Adventurer (say Prince) is a much smaller step than to step from Prince to Immortal as in da Vinci's suggestion.

 

[Frederiksberg]

I agree that comparability is the main reason for not adopting this system.

With the current system there is already quite a difference between Challenger, Contender and Adventurer making comparison difficult. What saves the day right now is that noone plays the Challenger save anyway.

But the compelling argument I see for having 3 saves is that it lets us cater to more tastes with the same game. It would allow us to run more Immortal+ games over a year, while at the same time still having lower-level games available for those who prefer them.

This is obviously a matter of taste, but I don't find it particularly compelling that you can add a few extra high level games at the expense of dilluting the player pool for each game and reducing comparability. Not to mention the award issues. I think you owe us to explain how awards and Eptathlon should work with 3 saves and common awards...

It also makes the choice to "take the leap" less dangerous for those that want to move up. Chosing Contender (say Monarch) over Adventurer (say Prince) is a much smaller step than to step from Prince to Immortal as in da Vinci's suggestion.

With 3 games every month you can choose any level within the range Noble-Deity inside one month. So what's stopping people from challenging themselves with hard games every month - nothing!

 

[azzaman333]

Why play games that I'll struggle in when I know I can win by taking an easier save?

 

[Ronnie1]

Why play games that I'll struggle in when I know I can win by taking an easier save?

Why play games when the end is not in doubt?

I sometimes play the Challenger save now when the difficulty is below Monarch, just because Monarch is almost always a win for me now. I'd rather have a bit of a challenge!

 

[DynamicSpirit]

Why play games when the end is not in doubt?

Because even if winning is not in doubt, the date that you eventually will win by is. The challenge for many people who like playing easier games sometimes is getting a better/earlier win than the other players playing the same game.