[JeffR]

Time for an update, it's been a while since I've done one of these.

Everything is on track so far for getting back to the regular Wednesday updates next week. That includes the new selectable league setup. So, instead of having so many playable leagues that it bogs your game down, you'll be able to just the ones you want (with one limitation: the highest-level league in the game and its top farm league - i.e. the NHL and AHL in the default database - will always have to be playable.) If you've played a historical game, you know how much faster a single-league game plays. That'll also eliminate the need for us to hold off on adding playable leagues (I'd been trying to avoid that because the game was too slow already), so we may have some new ones this coming week or next - I won't say which, exactly, as I want to talk to the researchers about it first.

Coming along with that will be a significant upgrade to the way players move around non-playable leagues, so switching a league to that setting won't make it completely static in terms of player movement. That'll let players have realistic career progression and will be the first step to implementing a better player generation model.

And here's where it gets complicated, and where a big part of the recent delays have happened. Now, the easiest way to handle that sort of movement is just to hardcode specific progression from league to league - e.g., every player in the BCHL moves to a US college when he's too old to play junior. But that runs contrary to the moddable approach we want to take with the game - if we start hardcoding that stuff, you can't change it, and that's not how we want to do things: if you want to try editing in every Junior B league in Canada and making them playable, I think you should be able to do that without breaking the game.

So what we've done is made the whole non-playable transaction system work at the team level: every team will individually specify where its players go if they become too good, old, or bad to continue playing there (and there's also a chance of random movement to simulate trades and free agency.) Graduating NCAA players, for example, have several different possible destinations if they're not drafted or signed by a team in a playable league, but those vary significantly depending on their team - someone playing at Harvard is more likely to quit hockey entirely rather than make minimum wage in the low minors, teams in the northeast are a little more likely to have their players wind up in the Federal League, WCHA teams send a few more to the Central League, and so on.

Likewise, teams can restrict which players come to them. Continuing the NCAA example, the Army and Air Force teams will only take American players, and you could have a team like the University of Minnesota limit its recruits to one state or a particular region.

Getting all the data in to make that possible isn't a quick and easy job, though - we're talking about data entry for hundreds of teams in dozens of leagues, not to mention the time it takes to figure out exactly what that data should be. I've spent most of my time for the last couple of weeks on it and I'm just now at the point where it's nearly ready to go. It'll be a welcome relief once it's done, but it'll also make the game much more "alive", so I think it's worth the effort.

In other news, the Mac version is on target as well: Sebastian's expecting to have a testable build next week, so we've added a few new testers today in preparation for having a look at that. Barring any disasters, we'll be ready to go with it on the May 1 target (and since that's a Wednesday, we may be a little slower with the data update that day if there are any issues with getting it ready for purchase.)

As for other concerns: crashes and player valuation/trading remain at the top of the list. Regarding crashes, please continue sending your crash-related savegames in, those are helpful. As for making the AI value its players properly, that isn't going to be something that happens overnight, but we'll continue to make incremental improvements there. I realize that's a particular source of frustration now, but I don't want to lead anyone to think we're going to make a change one week and the AI will suddenly be brilliant. That isn't to say we're anywhere close to what I'd consider acceptable AI roster handling in a release version, just a warning that it'll get better in stages that may not be really obvious from week to week - AI teams will stop trading for three starting goalies or refusing to move a draft pick higher than the fifth round, or bottom-pair defencemen won't ask for $6 million a year and then accept $600,000 after a couple of weeks of free agency.

This is getting a bit long-winded, so I'll cut it off there, although in closing I'll add that plenty of things not mentioned above are also getting addressed in the usual day-to-day bugfixing.