[JeffR]

One more update from me:

Development changes should, barring anything unexpected, be all but finished now. I'm seeing more or less what I was aiming for, aside from some display issues (e.g. when scouts overestimate a player, it results in them being shown as ridiculously good, so good that it's obvious the scout has to be wrong.)

As for scouting, I think that's pretty much done, I just need to add a few more phrases (for variety) to the text the scouting report uses.

We wound up scrapping the part of the retirement system that tried to get guys to quit when they'd declined too far to play in the league they spent most of their career in; it just wasn't very effective, probably because it was a little too rigid (using games played totals) in how it determined the league in question. Rather than trying to come up with an elaborate fix for that, we're instead going to track what the player's peak ability was, and then have a retirement possibility trigger when it declines a certain amount (the probability increasing as the decline continues.)

Research update deadline was last night, so we can now get the various systems-affecting database changes (more farm teams, changes to junior roster rules, a bunch of adjustments I made to player movement in nonplayable leagues to break up bottlenecks that were starving some leagues of players, etc.) that depended on that update into the next build. Also waiting on that build to evaluate the recent changes to contract systems.

We've added one more big overhaul to the patch: the coach and GM hiring and firing system, which was some of the oldest code in the game, is getting some significant changes. The frequency and timing of firings should be a lot more realistic now. End-of-season firings will still come without much warning; the computer will evaluate the GM's recent performance based on a variety of factors and make a change if it doesn't like the direction the team is going in. In-season firings, however, will work a little differently - when the team is in serious trouble, underachieving and in a slump, the GM and/or coach's job has a chance of becoming in jeopardy. There'll be a news item announcing that (the text may be a little sparse for now, we'll try to add a little variety later.) If that happens, every subsequent loss will bring the chance of a firing, until it either happens or the team manages to win a couple in a row, ending the jeopardy status (for the moment.)

The hiring improvements will try to emulate real hiring patterns: some of the time, teams will simply decide to promote from within. If they don't, they'll shortlist five candidates, the highest-reputation ones that make sense for their level (i.e., no NHL coaches showing up in the Federal League) and have appropriate resumes (e.g. not choosing unemployed assistant coaches for GM jobs, and favouring nationalities from the country or countries where their league is based), and then choose one of them. Hiring directly from other teams' staff is possible in the offseason, but will be limited to hiring from teams in lower-level leagues and staff that will be getting a promotion within the same level of league (e.g., one NHL team hiring another's assistant as its head coach.) Human GM's will be at a slight advantage in the hiring process, but you won't be getting badgered with job offers from lower-level teams; computer teams will avoid you while you're employed in a better league.

Staff generation is getting some small changes, mainly to give them lower starting reputations and limiting the full 1-20 range of attributes to coaches in the major hockey countries - smaller countries get a more restricted range. Reputations will also be driven more by the ability to stay employed - as long as a GM/coach/scout keeps their job, their reputation will tend to go up (but not beyond a level governed by the quality of the league they're currently in.) New staff will also only generate as scouts or assistant coaches - getting the "Head Coach" or "GM" role requires being hired for that position by someone. I've got some more long-term plans for generating staff and handling their movement on non-playable teams, but there won't be time to get those into this version.

Finally, the match engine: Sebastian continues to make progress, and I continue to write text for it (watching the Olympics is handy for that, I can watch the CBC feed and tab over to the text editor whenever I hear him use a description that'll work in the game.) However, it's going to be a while before it's ready, which leads me to something I hinted at last time: we've decided to do this big update in two patches, one with everything that's been done up to this point, and one that'll be mainly the match engine. It doesn't make sense, given all the changes and bug fixing that's been done in the last few months, to leave all of that on the shelf while we wait for one more thing to get done - even if it's a pretty huge addition. So we're going to combine the planned trading deadline data update with a patch containing all the new stuff except the match engine. It won't be out immediately on deadline day, which is March 5, since we'll need to do some testing to make sure nothing went wrong when the database is updated with the late roster moves, but it'll be very soon after that.