On last minute changes & media

It’s pretty hot in the office for the moment, but they say a thunderstorm is coming. I’m looking forward to that. Some rain would be welcome.

I slept about 4 hours in the last couple of days. Better put, I slept 4 hours this morning. And no, it’s not because of Dragon Commander crunch even though there’s quite a bit of that going on.

No, these particular sleepless nights have everything to do with a bad mix of wisdom teeth , inflamed sinuses and nerves that insist on being where they shouldn’t be. It’s a problem I’ve been carrying around for some time and which has the annoying tendency of popping up at the most inopportune of times. Hurts like hell and reduces me to my most animalistic self when it occurs. I really need to get this fixed. It’s better now, courtesy of the terror unleashed by several grams of narrow-spectrum antibiotics on the cell wall of the causative agent. But it was quite a painful trip and it prevented me from functioning normally these last couple of days. Not that I had anything meaningful to do anyway.

Opinions like this and this keep us awake and make us make changes, but that’s not the only thing keeping me awake.

Given my lack of sleep, I offered myself some recovery sleep this morning and so I only arrived in the office around noon. As I entered the office, I couldn’t help but notice that something changed.

Stacked neatly in a corner where normally producer & video boy jack in, stood twenty thousand carton boxes filled with German retail editions of Divinity: Dragon Commander. It took my dulled senses a few moments to understand what these boxes were doing in my office and why David was indulging himself in a bit of hydraulic trolley play.

You see, normally, boxes don’t come to developer offices, even self-publishing ones. They usually travel from printer to distributor warehouse to retailer. But normal of course doesn’t apply when bureaucracies are involved. Insert here a long and complicated story about the absurdities bestowed upon us by our beloved bureaucracies which makes us send out thousands of boxes, printed in the destination country, outside of the destination country, only to ship them back to the destination country the day after. That too is part of self-publishing, at least in the European Union it is 😉

Anyway, once I remembered again the absurdity of why these boxes were being unloaded in our office, my next stop was our technical director.  Obviously I was very eager to hear what had happened between time A when I left the office and overdosed on every single over the counter painkiller I could get my hands on and time B which saw me returning to an office where confused boxes in the wrong language blocked our back-entrance.

A lot it seemed.

All of the changes that we agreed on were implemented or being implemented, and a number of new crashes had been found and solved. If you’re wondering why we are still changing things, well, you needn’t look further than our forums, youtube, the steam community hub, facebook , twitter and what have you. There’s a continuous feedback loop going on there and we’re seeing some very well formed opinions appearing.

Our code of conduct is that whenever somebody posts a bright idea it gets on our list, and then put it in, as long as it remains feasible for us.

I can guarantee you that there’s a lot of people in the development trade (that actually includes guys in my office ;)) that will tell you that this is the way to ruin, but my experience has been such that you’re better of with a game that’s fun and maybe not polished than you are with a game that’s polished but not fun. Fixing the polish is an easy enough thing and almost always a matter of money. Fixing the fun otoh is still somewhat of an arcane art coveted by many but mastered by few and money will not necessarily make the difference. Because my interest and joy in making games comes from stumbling upon ways of making them fun, I tend to sin more than often against the rules of maintaining the outwards production values that are so important these days, preferring a message box if need be over not putting something in that clearly improves the game mechanics 🙂

I joke, but I really do believe that whenever you realize a certain change will make your game more fun, you should do it, no matter how late in the development process you are. You should actually count yourself lucky that you had the insight prior to release. The only reason I can see why you shouldn’t embark on making the change is when you can’t implement the change properly for whatever reason. But you shouldn’t let that be an excuse for not making the change. 😉

We’ve had quite a few of these types of insights ever since we started printing those boxes. A lot of them came from  as a result of the closed beta we’ve been organising and we learnt a lot from seeing all those players in action. So I’m actually quite pleased that we decided to have a mandatory zero day patch. I guess there is still some discussion about these type of patches, but really, this is 2013 and the very reason why some games aren’t as good as they could’ve been is because developers couldn’t touch their games anymore once they went gold, or because they were allowed only one “free” patch which is what happened to us on the X360. (I know it’s changed now, but this particular stupid rule has been in place for many years)

Yours truly once even heard that his publisher’s games were perfect and that therefore they had a zero patch policy. Even showing said publisher a mantis database with over a 1000 flagged bugs in it didn’t make them change their opinion 😉

However, there is one nagging and very practical publishing problem with this modus operandi, and it’s one that needs to be solved by the self-publishing developer who wants to make his game as good as he can: when to send out review code.

If you send it out on the day you release the game, you risk not having enough hype and people thinking that you have something to hide. (I’m thinking of you Angry Joe) There’s also the risk that if it’s already old news, certain outlets won’t write about it. But,  if you send it out to soon, the risk is that you have  a game reviewed which is significantly different from what you are actually going to be shipping, because perhaps one of those last minute changes actually makes all the difference in the world. (I don’t remember if I ever explained here how there was 10% metacritic difference in Divine Divinity reviews because of one change, but if I didn’t, let me know)

Dragon Commander for instance is a much better game this week than it was last week, and those changes I was referring to in a previous paragraph will probably make it even better (it could actually also be worse, that happens, but I have high hopes).  I certainly want this week’s changes to be in a version that gets reviewed, but of course, I don’t know what’s going to happen next week. Perhaps there will still be one last change in there that makes it a classic! 😉

So the very real problem poses itself – when should we send out review code of Dragon Commander? As it happens, I needed to answer this question today, because we’d announced here and there that today was going to be the day.

Obviously, if you intend on continuing to make changes to make your game better, you want to send out review code as late as possible, so that reviewers have the most up to date code. Preferably, you do give them sufficient time to create their review and the consensus seems to be that 10 days is sufficient.

Sadly, that’s not always feasible.

There are still print magazines out there, and despite their numbers dwindling fast, they still have some influence, especially among retailers. If you want to be featured in those magazines and you are not one of their colleagues turned developer, you really need to send out your review code about a month before release. A month is an eternity in game development, especially if you adopt the change-if-it-improves-the-fun dogma, and so the problem of when to send out review code is amplified when dealing with print.

I don’t really know how to handle this problem but I do know how to not do it. Don’t differentiate between publications. We for instance gave out Dragon Commander review code early to a few print magazines to take into account their deadlines, and we regretted it almost instantly.

From the very moment we gave out the code to the moment the magazines had to go to print we‘ve been bombarding them with new code updates, sometimes twice a day, causing stress on our build pipelines and stress at the journalists side. It was our choice to hand out the review code (well, if we didn’t, we wouldn’t get a lot of coverage by them) so we only have ourselves to blame, but in hindsight, I wouldn’t do it again if only because those reviews won’t be talking about the same game the other reviewers will be talking about. Which is a pity (for the record, I really have no idea what our scores are or what those reviewers thought)

Actualy, once we realized we made a mistake, we actually tried to pull back the reviews, but then of course we created a problem for the publications who already counted on having those pages dedicated to Dragon Commander, so we had to continue with it, if only to live up to our promise to them that we’d provide them with review code. Yes, we made a mess of it and I’m not sure they’re used to devs changing their game so fast.  In any case, we tried explaining their editorial staffs what it is that we were still doing and why we considered this important for the review version, but for them too the proof is in the eating of the cupcake, and so I’m not really sure if that really made an impression. I guess they hear stories like that all the time from developers running out of time and I guess that more than often if they choose to believe what a developer or publisher says, they end up disappointed because a promised change got lost somewhere. So it’s probably just easier to review what you have and ignore all the rest.

Luckily (perhaps, who knows), there are only three publications where we made this mistake, so I think it’s safe to say that for the remainder, the same version of Dragon Commander will be reviewed. They’re just not going to get it today as promised – it’s more going to be like end of this week, perhaps beginning of next week. Not because we’re trying to hide anything, the beta & preview versions are out there in the open anyway, but because… well, we’re still making some changes, and we think they’re important enough to withhold sending out review code.

The risk of course is that we’ll get less coverage by the larger publications who usually need more time to do their thing, but I don’t want to run the risk of not having that one feature that makes the game that much better than it already is. (Yes, I had to put that in there, but I actually do believe it too ;))

So, I’m guessing we’ll be sending out review code by the weekend, and there’ll probably be a few updates during the week 🙂 Yes yes, I know.

  • Stabbey

    Get better soon, Swen! As always, this is a really interesting inside look at game development.

    “Our code of conduct is that whenever somebody posts a bright idea it
    gets on our list, and then put it in, as long as it remains feasible for

    Okay! So here goes:

    On the RPGCodex, Forktong said that in single-player, the gold costs for certain decisions, like healthcare was a flat fee. Now, I’m not sure if the fee scales the greater your income, so if it already does, never mind this.

    The Healthcare one flat fee seemed to be 2 gold. That might be a lot per turn when your total income is under 20 gold, but when it’s like 40+, you’re no longer really going to notice 2 gold more or less. It’s not even a single squad of Troopers.

    If you get the healthcare question randomly about 30 turns into the game, that 2 gold fee isn’t going to be much. To me, that kinda seems to not really be much of a trade-off or a hard decision. Unless you really are going for a faction that doesn’t like it, there’s no real downside to paying that 2 gold, there’s no long-term consequence. If there’s no meaningful long-term or ongoing consequence, it seems a bit of a missed opportunity.

    So, healthcare should probably cost more. But just tying it to income doesn’t sit right with me. Healthcare costs are tied to amount of people, not amount of money. If you play a “Revenue increase” card or plant a gold mine, then your healthcare costs would go up without your population going up. So my idea is to tie healthcare costs to the total population under your control. Healthcare costs a flat 1 gold, plus 1 gold per every 5000 or so population under your control (adjust as needed). It would still hurt more at the start, and the cost would increase as you start controlling more people and land – but the increased revenue from the new territories would be more than sufficient to cover it.

    There, now universal Healthcare is not just a token free approval bonus with a token gold cost, but an actual, noticeable drain on the war economy, no matter if the healthcare question comes up on turn 1 or turn 100. What do you think? (Note: I am not actually expecting an answer.)

  • Fox

    I don’t remember reading about Divinity 2’s 10% shift… but it’s possible I simply missed it. I don’t think I’ve read 100% of the posts on here (but I do read a lot of ’em–very informative stuff).

    Anyway, really looking forward to Dragon Commander and Divinity OS.

    As for review code, I say hold it back as long as possible and–if necessary–delay the game. Last year Baldur’s Gate Enhanced Edition launched with a TON of bugs… and even though most of them were fixed within two weeks, the game everyone reviewed was the one that launched. Even now, with the game virtually bug-free, it still has a very perceptible reputation of being a buggy mess. Very unfortunate stuff.

    • Raze

      It was Divine Divinity that got the 10% metacritic jump, but I’m not sure which blog post that was mentioned in. Essentially, after the German release of the game Larian tweaked the random loot generation before the English release, without telling the publisher (so they were surprised by the difference, since RPGs tend to be more popular in Germany).

      • http://demovit.tumblr.com/ Movit

        I don’t know exactly, but I think they also fixed some problems with level interiors that made you get stuck on pebbles and other small things, what I thought was the most annoying thing. Not counting the psychic AI that remembered it has attacked you even when you loaded a save game previous to this encounter.

        @Sven: I would consider to host preview code on your own servers and give journalists a gaikai like access.

  • AlexF

    I think you should be prepared for some serious backlash due to the required day 1 patch for boxed copies. You should at least host the patch on a fast server and let people download it without using the game client. There are even nowadays people that don’t have internet at home or people like me who have a much faster connection at work than at home and I’d prefer it if I could download any patches here and then take them home on my flash drive.

    As for reviews, it is generally thought that a company that doesn’t send review copies ahead of release is trying to hide something. However your case is different. You have sent the preview copy to a great many sites and people and it contains a good chunk of gameplay. You’ve already shown the game and your development process is one of the most transparent. I’d like to believe that if you explain to those mainstream sites that you didn’t forget about them but you were developing till the last moment they won’t take have a less favorable look on the game.

    • Stabbey

      There may be some backlash, but it is an RTS game, so balancing patches will be expected by anyone familiar with the genre. Better to have a little flak for requiring a day 1 patch than a lot of flak for things not working properly.

  • melianos

    Mardi Mardi Mardi Mardi Mardi Mardi Mardi Mardi Mardi Mardi Mardi Mardi Mardi Mardi Mardi Mardi (if you understand what I mean :D)