Lower cost of shop items

Now that tokens are so very precious, it might be an idea to lower the cost of game types one can unlock at the shop.

I always understood the logic here being that we wanted to lock most advanced game types so that new players wouldn't accidentaly stumble across them. They were free to unlock, only costing a free currency which was easy to get.

Now tokens are harder to obtain and much easier to lose, it would be a good idea to (dramatically) lower the cost, so that there isn't an actual cost associated with unlocking new games?

Here's also a tip to earn more tokens. Paste your referral link into facebook, or other type of gaming forum to share out site.
I've earned 318218 battle tokens (present count) and still earning extra tokens from a whole bunch of referrals.

https://dominating12.com/account/referrals


Referral program
When you tell another person about this site and they join, you can earn tokens from them. You will receive 20 tokens for every game they play up to 25 games. That is 500 tokens per person you refer.
In order to gain this benefit the member must put you as the referrer. If you use the referral link it will automatically put you as the referrer when they try to register and they will not be able to change it.
You are not only helping yourself but helping us all when you refer other members. The more members, the easier and quicker we can join and play games. That is what this site is about.

Ultimately, we'll get games going where you can bet tokens or something

I checked, and gametypes are 200 tokens each - that's pretty cheap, so what's there to change? I really like it if new players are discouraged from playing easy-to-ruin games like capitals or same time (live) games.

I used to think 200 tokens was pretty cheap too. And I think it is pretty cheap for you or me. We have thousands of tokens lying around. So what's the problem, really?

I think the problem is people who have joined in the last few weeks. They come and play for as long as they have tokens. Most days 3 games. Every so often they can get in a 4th game. They play until they have no more tokens. And so they never have enough tokens to branch out into other game types.

Sure, these people are new now. But in a few months they won't be. And they will still have no tokens.

200 tokens use to be, if you'd pardon the phrase, a token payment. It was free to unlock games. You unlocked them with a currency you had lots of and you you didn't need for anything other than cosmetic effects like rank. That has changed, There is now a currency barrier to unlocking other game types.

Well, a big part of this site is long term games, which are not limited by tokens the same way.

perhaps to alleviate some of the token stress on players maybe a "free day" would be nice,
or perhaps you pay x amount of tokens and are allowed to play unlimited live games for a day. 

You're entirely right, I'm writing here about live games and people who play these. Not about the long term games part of the website, with which I have very little contact.

And this is exactly where the problem lies. The stress that the server receives from live game players is a lot higher than the stress from long term game players (and I am talking about factors of 10+ here!). The new restrictions have been put in place on live games for a reason (and subsequently, the restrictions on long term games relaxed). If live game players want to branch out in other game modes, they can save up a few tokens. Playing two additional live games a day already costs more than a new game mode, so I consider the "sacrifice" you have to make for a new game mode very low.

We will also hopefully see an increase in premium players who want to play more live games per day. For these players the token costs will become less of an issue, and they will buy the game mode, so I don't think we have to be afraid for a lack of people interested in playing those games.

@Matty, Cireon: thank you for your replies. I have a question regarding server stress levels.

Prior to the upgrade there were more players online (you probably have figures of population decrease since upgrade). How come the servers are more stressed after the upgrade with fewer people?

Is the improved code more inherently more unstable/bulkier than the old code? Or has the site moved to a different server with reduced capacity? Or is the server stress due to bugs in the code that are still being ironed out? Or is it not yet known what is causing the undue server stress?

Hey guys, any chance of some more info on the server stress/lag questions above? I think a fair few in the D12 community would feel better knowing where we are at with this. And do please say if you would like any help to run tests/do anything else!

The problem is that it's hard to say something defenite on this.
Let's just name some semi-random things:

- If players spam click the attack button, then they open up a lot of sql connections - there is a max limit on them, theoretically one player could block the site for everyone for a little while.
Of course, one player alone can't do everything, but if three players are spam clicking at the same time, and the troop counts are high, it can mean that the entire site is pretty much frozen for a few seconds.

- To ensure there are no weird glitches in, say, same time games, there is a block on the database per game. So only one player can conqueror a territory at the time. If two players try to attack a territory (in the same game) at the same time, then one of them will have to wait x time (not sure what that x is), and then try again (if then someone is attacking, he ...)). This should not change anything for normal games, but same timers (like you) will probably notice this when the server is stressed already.

- Some often executed queries are cached (like the game state - a spectator should get the game state without firing a large query). Of course, this does mean there is less ram available. What happends when a game state is invalidated (changed) all the time?

- The whole idea of how a game works is fundamentally different. Right now all 'moves' are stored in an event log. From there a game can be reconstructed. For example: "kill 3 lose 2, conquering", or place 20 troops at territory y. This way a game state can never be corrupt. Of course, we don't build up every game every query, so a game state is cached (that's how glitches are possible again), and also the turn code is not inside the event log model (because that would be way too much - and it can be reconstructed for the most part, in case of something going really wrong).
In 'the old way' there was one game state that was overwritten everytime sometihng happened (like an attack). This means that there is a lot more chance of getting bugged games (which happened quite frequently) and no way of figuring out what happened (well, there was an incomplete game log....).
This could be quite a bit slower, it could also be quite a bit faster, it's hard to say.


Some of these are things we will look at soon-ish (like the spam attack button thing). Some of them are reallly things we should monitor. Of course, we haven't optimized every query yet, like the old site did. Of course, the old site was also getting close to the limit of what it could have (browse a few news posts and you'll see that we had a server upgrade at some point (multiple points? Not sure)). So by making things a little bit slower means we sudddenly can hit a limit at peek time, which is what's happening now.

I'm curious what would happen if we disable all locks and transactions (apart from weird glitches that is). But really, we can't tell unless we go monitoring specific things, or just try and see if it solves something.
The good news is that all major bugs are fixed (there is one that we aren't sure of, but we haven't seen it in a while, so it probably is fixed). So we have time to look at speed again.

Thanks for the detailed reply Matty, that really helps (or at least it helps me understand what's happening on the back end a bit better).

Do please let me know if you'd like any help with testing at some point.

@TheMachiavelli I know that you are an advocate of political ethics that hold effectiveness as more important than morality and believe that the ends justify the means. Therefore, I'm sure you'll understand the need for tokens to have a value and for that value to be important.