[Fantasy draft bug and a few suggestions]

Yay! New version of EHM, how exciting! However, all is not perfect yet so I thought I'd give some feedback.

First of all, the bug I submitted a few years ago regarding the fantasy draft with more than 40 teams is indeed fixed, as well as all directly related bugs! Awesome!
However, with that being fixed, another bug appeared, although I've technically known about it for a while now.
Anyway, in that old 48 team database that I previously submitted, if one runs the fantasy draft it works fine for a while, until it suddenly just end for no obvious reason. At first I thought it was a player rights issue, that as soon as a team hits the maximum allowed player right the draft stops, even though that's something that would sort it itself out. So in an attempt to circumvent that problem I released all NHL player rights from all players. It didn't help though. Instead of ending abruptly in the 30-35th round, it now consistently ends in the 15th round, without exception. Always different picks, but always the 15th round. I tried to rearrange some of the players between different teams to try and circumvent it, but it seems determined to stop in the 15th round which, you know, that's a bit of a bummer. I don't know if you need me to upload a new file, since you might have the old file still that I submitted a few years ago, but if not I'll submit it later.

Beyond that I have some suggestions, some probably harder to implement than others.
First suggestion is regarding the waivers. I decided to turn Waiver Draft on in one of my databases to see how it works and yup, looks good. However, one thing I think it's a bit weird is that even with Waiver Draft enabled, teams seem to still put players on waivers BEFORE the Waiver Draft. This is something that makes no sense to do, unless you're desperately trying to do a cap dump, since all player exposed before the Waiver Draft will have to be exposed during the Waiver Draft again and thus you're just giving all other teams a chance to pick up your player for free twice instead of once. Perhaps this is something someone could circumvent by tweaking for rules in the database, but in any case it makes no sense for teams to do this.

Second thing is player management, NHL teams are using way too few players. During the 18-19 NHL season, the last complete NHL season we have to compare to, the 31 teams in the league used a total of 1050 players. During a test run of the latest TBL roster using the beta version a mere 824 players were used, over 200 less. The reason for this is simple, teams almost never assign farm players to the first team unless they don't have enough to fill out the game day roster. It doesn't matter if two of their players are out for the season, they'll almost certainly be playing with a 21 man roster for the rest of the season. This of course makes no sense since unless there are cap concerns, those roster spot would be filled. Beyond that, teams usually juggle tweeners in and out of the squad most season to test them out. During the previously mentioned 18-19 season Anaheim managed to have 49(?!) players dressing for them, although most teams usually finish with a final player count somewhere in the 30s. Whereas in my test run, the team that managed to use the most players during the season was Chicago with a paltry 32... And Vegas literally just finished the season with the 23 players they started with. No roster moves what so ever, every single farm player was staying put that season (they didn't even make a trade)... Which really feels just awkward... And the only way to fix it right now would probably be to set the roster limit to 20 to force the computer teams' hand... Although the more elegant solution would probably be tweak the AI's likelihood to make a roster move...
(Did a second test and no team used more than 29 players this time... And Vegas just used 23 players this season too! They must really hate their farm team...)

Although part of the problem with the lack of roster moves is really another problem altogether which is that player are just too durable, there just isn't enough injuries. Now this is admittedly something that could be rectified atleast a little bit by changing the injury proneness of players in the database, but one regens come around I suspect there will be alot of iron men around. During the previously mentioned 18-19 season, after ignoring all players that for all intents and purposes have retired, there were 6867 "man-games lost" as they say. Now there isn't any easy way to measure that number in EHM, but after checking my test sim and ignoring all players that were preset in the database to be injured (thus not something that the game itself had anything to do with) a rough estimate is about half that. And this ignoring the fact that alot of the time when players get injured in EHM they only get a "yellow status" and they're still fit for play so the actual number of games missed is likely lower much lower. The result of this is quite clear. The previously mentioned Chicago is the team with the most "man-games lost" with an estimated number of 180ish "man-games lost" (after discounting preset injured players) compared to the 18-19 Anaheim which ended up with 428 man-games lost, which is partially the reason why they used so many players. Ergo, it's likely that if the injure numbers were more realistic, the roster usage numbers would improve as well... Although not completely seeing how there are team like Vegas how just doesn't like any of their prospects and refuse to assign any of them to the first team, no matter what... or something of the sort...
Poking through the editor, in the injuries section, I did see that there's a field named "Chance". I don't know what that actually means though, like if a higher "injury chance" increases the likelihood of that injury overall or if a higher chance just means it's more likely that that specific injury occurs after the game has first decided that an injury has happened at all. If it's the first option then that would be a solution to the problem I guess, if it's the second though it wouldn't actually make injuries more likely, and all you could do is change the ratios... Although I guess one could just mod all minor injuries out of the game to ensure the more serious injures are more likely... Either way, higher likelihood of injury is needed.

I know, these are fairly late in the day things to poke at probably, but I feel they are things that needs to be adjusted. I have over 1200 hours in game according to Steam, but 'til this day I still haven't be able to finish a second season in the game, because the unrealistic behavior of the computer AI just pulls me out of it. So I've mostly just been doing experiments for the last couple of years... Meanwhile I've played good old EHM 07 until I hit the save file corruption bug twice... That game isn't perfect either, the AI makes strange choices frequently and veteran's skill doesn't diminish fast enough for example... But it still feels more realistic. I mean, the new version sans balancing is a better game hands down, but those things are still obstacles to my enjoyment...

Maybe what we need is the return of the sliders from the good old freeware game. That game also had a problem with players being too healthy, but once you maxed the injury bar it was perfect! Probably too late for implementations like that though.

[Provisional changelist from the work-in-progress patch 1.5]

On 13/05/2020 at 13:42, Riz Remes said:

- Enabled waiver draft support for National League in historical databases (waiver draft needs to be set to "not extinct" in the database to enable it)

Somehow this is the thing I'm the happiest about... I always liked the idea of the waiver draft and I especially like how it can work as a makeshift expansion draft. But just in general it looks great!

[EHM Wishlist thread]

One decimal game player ratings as it is in FM would be kinda neat too.

[EHM Wishlist thread]

It's a small detail but personally I'd like a stat for goal/points per 60 minutes of ice time. Just seems like it would be a simple and effective stat to evaluate the offensive performance of a player.