EHM The Blue Line

JeffR wrote:I'm just going to go with a list of the major changes this time, rather than a list of every fix. A lot of what we've been doing has involved a big change followed by incremental work refining it, so a huge chunk of the log would be fixes to problems you never got to see (aside from the ones in the accidental 0.9.59 release - again, sorry about that.)

Underpinning most of the AI changes is a new 10-point rating system, replacing the old OOTP-style five stars. Generally, players rated between 8 and 10 should form the core of your top two lines and defensive pairs, while those rated 6 to 8 fill the remaining roster spots, and those lower than 6 are, at best, useful as short-term replacements. As with the stars, the scale is always relative to the level you're currently at (if you're unemployed, it defaults to NHL-level.) So if you're playing with a team in the ECHL and look at an NHL roster, most of the players you see will be star-caliber by your league's standards (including, potentially, ratings higher than 10.0 - there's no cap on the top end of the scale.)

Keep in mind that it's not just a straight-line conversion of ratings to an arbitrary scale - to reach the 'first-line' levels of the scale players need skills that are relevant for the top lines (i.e., offense is weighted heavily), while getting a rating near the top of the ratings group associated with the bottom half of the roster requires more defensive skills and intangibles. That way, for example, a one-dimensional offensive forward who isn't good enough for the top 6 tends to drop right off the roster instead of getting a third- or fourth-line job - the "Keith Aucoin" effect. Finally, there's a special "goon bonus" for enforcer-type players that increases their ratings to a level where the AI will treat them realistically (the game lineup AI isn't doing that yet, though - that's coming as part of the game engine work this week.)

At the start of the game, most of the ratings you see will be fairly accurate representations of player skill. But newly-generated players can have significantly inflated or deflated apparent value. That's where your scouts come in: assigning them to a geographical area (particularly if they're from that area themselves) will reduce or eliminate the effects of inaccurate information about players in that area. (The geographical areas have also been significantly simplified.) I'll be elaborating on that in a systems discussion later this week.

The cap rules now allow the accumulation of cap space as in the NHL. We've presented this in what we think is the easiest-to-understand way, by showing your cap limit as an amount that will gradually increase during the season if you stay under it. That way, you can easily see exactly how much cap hit you can take on via a trade or free agent signing (rather than having to do a bunch of math with daily payrolls and remaining cap room.)

AI teams now have a more logical approach to cap management if they find themselves over the limit, instead of dumping half their roster if they need to make space. They'll try to resolve the problem by sending players down first, but if that fails they'll try to move their lowest-valued high-salary players to another team for a draft pick or picks. There are still some teams taking radical steps to get under the cap, but at least now they're not ditching half their roster. The ultimate source of the difficulties is the new CBA rule that deters minor-league salary dumps - when the largest amount of cap space you can gain via sending someone down is $375,000, that severely limits your options and makes writing cap-management AI much more complicated. We've also got multiple NHL rosters that are currently taking advantage of the 10% off-season cap cushion, but will need to be cut down somehow when the season starts.

The AI's trading logic has been completely overhauled, and trades between AI teams will now occur at realistic rates and times (more of them at the deadline, for example) and be significantly more varied in composition than they were before. I'm still not seeing as many draft picks moved as I'd like, so I suspect something's off in the way they're valued (also via a new method, one that takes into account the strength of the draft class), we'll be investigating that further. It also tends to overvalue older players.

AI teams should also evaluate your trade offers a little better; among other things, your trading partner will now specifically say if he's refusing the trade because you're giving him more salary than he can afford, which will let you remove some players or take salary back to fix the deal.

For now, AI teams will only make offers to you for players that you've put on the trading block (and those offers should come at a fairly dependable, but not annoying, rate.)

It's now possible for teams in leagues that allow soccer-style transfers to purchase players from other such teams; right now, this is mainly enabled in the European leagues. At the moment, pending better data on how much cash actually changes hands in these transactions, typical transfer costs will be an amount around 90% of the player's annual salary. Both the player and AI can make these deals.

Some loopholes in free agency that allowed the signing of ineligible players (other teams' draftees, underage players, etc.) have been closed; you now should only be able to open the offer screen for a player if it's possible to sign him.

Players who are RFA's on the expiration of their contracts should now be properly designated at the start of the game, but there still appear to be some issues with the AI's qualification-offer logic.

The historical database is now complete through the 2008-09 season. I also spent some time today making a few changes that will model high-PIM players more accurately. On the downside, the new trade engine isn't working in historical mode yet, but that's coming.

The match engine now includes more varied text for shot and faceoff events; there'll be a lot more coming, but you can see the basic outline of how it'll work - we just need to add the text triggers to things that are already occurring in the game, but not being reflected in the text (e.g., all the giveaways, takeaways, hits, penalties, etc. that you see in the stats.)

The new researchers are getting up to speed and there's been much more editing activity in the last three weeks. That'll continue post-release, we'll be working continuously on the database and posting regular updates.

YZG continues to add to the player name-generation files, we're gradually taking advantage of the new naming-by-state ability to get more accurate regional names.

The Norwegian GET-Ligaen is now playable; thanks to Gorane, our new Norwegian researcher, for his work there. We have a couple more leagues that are very close to being ready for playable status, but those will come post-release, since there's not enough time to test them properly right now. And there are more that will follow them as the research gets completed.

Various crash bugs have been fixed, although we discovered a new source (or sources, rather) of lineup corruption today that may take a few attempts to eradicate completely. The good news is, it doesn't tend to produce repeatable crashes, so if you get one (typically in November-December) you can usually progress through it on the next attempt.

***What's Left***

With a week to go, the three areas we're going to be concentrating one are the full implementation of custom play, adjusting several things in the match engine to improve stats generation for star players and defencemen and increase the effects of toughness and leadership, and getting facegen working. Those items are already in various stages of completion, but weren't suitable for inclusion in this update in a partial state.

Facegen ran into an aesthetics problem: when you take the baseball caps off OOTP's players and replace them with the hair options we have available, the results vary from funny to hideous, leaning towards the latter. So for now facegen players will get draft-day style portraits, with the ballcap on. (Update 8/27: I wrote that last night, and with today's delays and the amount of time needed to get the necessary art assets finished, it's looking like facegen is the top candidate to slide past the release date. Since it's cosmetic, I think it makes the most sense to cut there rather than lose func

The match engine work will also include some non-stat fixes, including, I hope, the return of the shot view screen (it'll be viewable by clicking items on the highlights list.) We'll also be recording some additional stats, including things that will allow us to calculate Corsi/Fenwick - you can see the beginnings of this in the bottom section of the player season stats screen, with some dummy values (PDO is there and calculated correctly, though.)

See the description of custom mode I posted a few days ago for the outline of what we'll be doing there.

Aside from those big-ticket items, we'll be continuing the usual database work and have time blocked off before the deadline to deal with last-minute issues, particularly stability-related ones. Sebastian worked out something today that will make one category of crash bug a little easier to find, so that should help on that front.

Viking wrote:Will FHM be shipped with a pre-game editor?

archibalduk wrote:Am I right in thinking there isn't actually anything but quickstart games for FHM? I don't think I've seen any other options in the game.

archibalduk wrote:The Commissioner Mode quick start creation is pretty neat. But it's no pre-game editor replacement.

helmespc wrote:FHM seems like it will pretty much be "always" editable, that is, once it is mature.

dave1927p wrote:If i were the FHM devs, as much as they don't want to do it, I would delay the games release. I know they've "carved the date into stone" but if any reviews come out early they could be potentially crippling let alone to people purchasing the game not knowing what they're getting into at this point in time. The fans/potential customers will get over it, some quicker than others. It is no longer a double edged sword...on end is just a dagger.

dave1927p wrote:Yes, all modes are crashing. Modern, Historical, and especially custom. I'm sure they are working day and night on this game but it's a mess at this point. I would be absolutely shocked if they cleaned up most of this mess by Tuesday. Even if they had an additional month it would be difficult.

dave1927p wrote:Hate to say it, but potential is worthless on a broken product.