Ban list is great.

I dont want to play with quitters.

K

+1.

This idea is nuts.

And scheduled gaming is not "the" solution to the problem of having people play others who are at a similar skill level. A better solution is having a ladder mechanic of some kind. Have people play who have similar win ratios or something like that.

+1 to this to.

I am very, very surprised if the majority of players, like me, don't find "scheduling" games to be very feasible or even a pleasant prospect when feasible.

the fact they are working on this for iPhones is abit worrisome to myself, seriously what percentage of demigod players even have an iPhone? 1%? less? Seriously resources (time, money, ideas) are being spent on this?

focus on adding new in game content or adding some story to the single player aspect.  this is my first negative critical assessment of something stardock has proposed, I'm a bit shocked I'm writing this...

true... clan support is needed, but this iphone stuff just seems like a gimmicky addon.

Jaradakar where were you in beta when I argued for True skill

sounds like alot of ppl are looking at this and expecting matchmatching.  Look at it this way, competitive players want to play competitive players, not pubs (that is if they are serious and want to get better).  this feature makes it easier for them to set up a scrim, (which is ALOT harder than it seems), therefore allowing them to scrim more frequently and taking them out of the public player pool, if not completely then for a good portion of the time.  There just needs to be something that accommodates the beginner level, grouping them together so they dont become feeders in a middle tier game and ruin it.

Assuming demigod hits critical mass...

-------------------------------------------------------------

Scenario #1 (Teams & Auto matching w/Trueskill)

You log on, check to see which friends you have online and which ones that are not in a game already. 

Then you group up into a team, queue up to find people to play against.

Search for opponnets feature then matches you up via your teams True Skill to another team that is also online looking for a match.  After a bit of searching, you're in a game.  Depending on who's online will determine how long the search takes and how accurate of a match up (skill wise).

--------------------------------------------------------------

-------------------------------------------------------------

Scenario #2 (Clans & eVite of Death)

Prior to when you wanted to play you send out an evite to your friends.  Some accept others can't make that time and suggest a new time.  Some can't make the new time so you pick and choose who wins out.  You then send an evite out to more people and finally get... what 5 people?  Or seeing as there is not automatching for teams, 5 people is not good enough you have to find and invite 10 people. 

Okay lets say there is also Clan/community forming tools.  You've joined a clan and have 10+ people in your clan.  Lets also say the tools support forming a team with clan members.  So you have your 5 friends and have set up a time that you can all play. 

Now lets also assume that the Clan system has a method that lets you challenge another clan to a game.  So you send out a 5v5 challenge to another clan at say 8:00pm PST.  They can either respond by accepting, decline or suggesting a new time/day.

Lets assume some back and forth happens and after much negotiation (multiple time changes, multiple day changes, etc) a final match day/time is setup.

Outcome A

Day of the match -- hopefully everyone shows up and it's a good time.

Outcome B

Day of the match -- Someone forgets, something comes up, etc and not everyone shows.  Back to the drawing board for trying to sync up 10 peoples lives.

--------------------------------------------------------------

Scenario #1

Pros

• Would match you based on your teams skill (via a method similar to MS True Skill).
• Easy/quick!
  • Log in.
  • Invite friends to form a team.
  • Search for opponnets.
• Does not require any pre-planning
  • You can easily plan if you wanted to via tools everyone already has (email, phone, etc) you could plan outside the game.

Cons

• Fails if the game does not reach critical mass and there are not enough people online playing.
• Can only draw from the pool of players online at the moment.
• Dependent on a robust friends list.
  • Need easy ways to add people from random matches to your friends list.
  • Need easy ways to add previously played people to your friends list.
  • Need ways to to invite friends to jump right into team lobbies.

Scenario #2

Pros

• If the community is small (Demigod commercialy unsuccessful?) and it never hits critical mass, Scenario #2 might have some advantages in that if you can't just log on and find people to play against it might still let you get a game going.
• It can prevent the scenario where you logg on and all you friends are already in a game and if you have really limited time it can help avoid this case (of course my argument is that's when you should just play single player skirmish or pantheon).
• More accurate skill match ups.
  
  • Assuming (and this is a BIG assumption) that this schedualing feature has a way to try to schedule against unkown random opponnets.
• If you're in a low population time zone (or low amount of Demigod players) it could help you find more games.

Cons

• Have to spend lots of time negotiating a play time.
• Requires you to plan.
• Requires you to join a clan.
• Might never be able to find acceptable play times with people of the correct skill.
• Possible that the match making of this system does not take skill into effect at all.
• Higher change that match might not actually happen.

--------------------------------------------------------------

Seeing how every developers goal should be striving to reach critical mass (Demigod is still in the top 10 selling list) and I can only hope that once connectivity is rock solid and advertising starts up again this will only increase...

A developers goal is to defeat the major cons to Scenario #1.  Scenario #1 also has far less cons than Scenario #2 and imo far outweighs implimenting Scenario #2.

To me, Scenario #2 feels cluncky and antiquated.  It's trying really hard to make more use of Custom games, which IMO are old school and antiquated to begin with.  Play DoW2 or any Xbox MP game.  Match making should be automated and based on a skill factor (True Skill for example).  The negotiation phase for Scenario #2 is time consuming and due to peoples lives, different time zones I can see it being a real nightmare.  It's also more work on the users part.  Users should not have to work, things should be easy, quick and intuitive and automated for the user.

Scenario #1 takes away most of the negotiation factor due to the assumption that people who are online right now, want to play the game and most likely have time to do so.

-Jara

PS: Custom games has it's place:

• Playing games with all friends (friends Vs friends), I.E. you know everyone in the match. 
  
• Or playing at a LAN party is another great example where Custom games is needed.
• Practice matches for clans/guilds/groups.

But for internet play, custom games should not be relied on much, if at all.  The only reason people are using it now is due to lack of any other way to group up and play team games with your friends.

PPS: To me, team based auto match making based on team skill should come first.  After we have that, if the developers wanted to add additional support that allowed/support for scheduled match making (I.E. my team composed of 5 friends Vs random opponnets of equal skill at a future date/time), I'd be ALL for it.

Hate to say it, but this sounds rather unworkable.  I like the idea of a host putting up a game scheduled for a certain time, so that people can plan to attend it, but making it such that people will have to schedule to join it as well is rather hard to maintain. 

I think what would make significantly more sense, is for you to add a filter in the custom games screen: Show Scheduled Games

The host chooses game options as normal when they schedule a game, including a time.  When you are at the custom games screen and see the scheduled games, they would have a value in the 'time' column, as well as a number of people who have RSVP'd.

To RSVP means that you have shown an intent to join the game at that time.  This sets up a reminder of some kind for the player via impulse, but does not restrict the number of other users who can also RSVP.  That means you can over book a game, and then take people as they come.

Note that you would still be able to password protect a game, if you wanted to schedule a scrim.

The main differences in my approach:

1) You are not at the mercy of players who do not show up

2) It is better integrated into the existing system

3) Short term scheduling easily allows for bigger games.

That is to say, if you are perusing custom games, and notice a game with 9 RSVP's is starting in 10 minutes, you can weigh that option against playing a random game.

Lastly, I'm going to say that I do not believe that this system will solve the skill mismatch problem.  That problem looks roughly like:

• Players have no feedback about the quality of a player, or of the players in a game when they join it.
• A player doing poorly in a game, and dying a lot, will actually increase the strength of the opponent, instead of just causing them to contribute less.
• Players are also often unable to accurately judge their own skills.
• A rather hidden thing, but since players often enjoy winning for its own sake, they will often prefer playing against weaker opponents, either by stacking the odds against the opponent (pre-selected teams vs randoms) or by 'slumming' any system defined delimiters (that is, if you have a 'noobness' qualifier to a game, people will not join it).

These are all the problems related to the skill mismatch meta-problem.  I do not think this feature will solve any of these.  Some random thoughts on fixing these issues:

1. Have join requirements attacheable to a game.  Win%, minimum games, total play time, total Favor earned
2. Some sort of visible rating.  When you join a random game, if a players win % or something was shown in an obvious place, teams could be balanced, people of inappropriate skill could be kicked, etc.  Having to go to html to see it is not good enough
3. Tie favor gains or pantheon gains or something to the requirements of a game.  If everyone in a game gets more favor when they play in a "Greater than 50%" game, they will perhaps seek out these higher stake games for the rewards they bring.  Something like this for pantheon would let high quality players have a greater effect on the overall tournament, for instance.
4. Allow for 'noob friendly' settings to a game: Players get less gold for killing other players, or player kill gold somehow only gets spent towards citadel upgrades, etc etc.

Sorry for the long post, but there you have it.

This isn't going to work very well unless AI is removes and the ladder system is fixed. Right now the majority of the "top" rated players according to your system play vs bots all day which if you know what you're doing you can win games in under 3 minutes making it pretty easy to mass game points.

Putting a 'record' of every game not only makes people ragequit more, but it promotes types of abuse such as people intentionally kicking me from panthion games so that they dont have to get rickrolled by me.

yah simple automatch with arranged teams is what we really need, I never heard of scheduled games before, maybe its common im mmorpgs (never played) but it sounds tedious compared to AT automatch. Instead of picking all those parameters and choosing strangers I just want to play with guys I know vs other guys who know each other. For playing with randoms there is skirmish aka random team automatch.

I don't understand why he posts provocative messages and then doesn't come back in to discuss the issues or at least acknowledge the points made and provide agree-to-disagree closure. Instead we're left feeling like it's a monologue rather than a dialog and wondering if we were heard. I know he's busy as hell, and I suppose it's better that he's posting and not participating further rather than not posting at all, but it's a shame that he can't find the time to at least acknowledge points made and offer his counter-rationales.

I agree with the first half of your agrument, but I believe banlist is a good move.