EHM The Blue Line

On my Boston save, season 2.. I was managed to trade my 2nd pick and Gabriel Bourque (who i signed as FA), and in return i got Jonathan Drouin. Tampa also traded away Hedman for a no 1 pick to San Jose, who had lost Burns to Toronto as FA.

I feel that the trade 'feedback' AI has actually gotten worse and worse to the point where it's completely broken now.

Here's an example: My team's board confidence went from 'generally satisfied' to 'extremely disappointed' (along with the fans raging) after I traded a 33-year-old career AHLer to a 4th round pick. This lead the board to give me an ultimatum after losing the next game. I went on a 7-game winning streak and my team had reached 1st place in the conference, but because that one meaningless trade carried so much impact, I'm now getting sacked at the next game that the team loses.

Thankfully there's a "disable human gm sackings" option for these cases.

atss wrote:I feel that the trade 'feedback' AI has actually gotten worse and worse to the point where it's completely broken now.

Here's an example: My team's board confidence went from 'generally satisfied' to 'extremely disappointed' (along with the fans raging) after I traded a 33-year-old career AHLer to a 4th round pick. This lead the board to give me an ultimatum after losing the next game. I went on a 7-game winning streak and my team had reached 1st place in the conference, but because that one meaningless trade carried so much impact, I'm now getting sacked at the next game that the team loses.

Thankfully there's a "disable human gm sackings" option for these cases.

Reporting this on the SI EHM Bugs Forum/uploading saves would likely be very helpful and give Riz a chance to look into it/make improvements https://community.sigames.com/forum/292 ... ugs-forum/

atss wrote:I feel that the trade 'feedback' AI has actually gotten worse and worse to the point where it's completely broken now.

Here's an example: My team's board confidence went from 'generally satisfied' to 'extremely disappointed' (along with the fans raging) after I traded a 33-year-old career AHLer to a 4th round pick. This lead the board to give me an ultimatum after losing the next game. I went on a 7-game winning streak and my team had reached 1st place in the conference, but because that one meaningless trade carried so much impact, I'm now getting sacked at the next game that the team loses.

Thankfully there's a "disable human gm sackings" option for these cases.

Yeah, trades have a huge impact on board confidence and fan support. I've won 7 Cups in a row, but when I trade a player who I already have a better replacement for away, the board and fans will loose their collective minds and drop the board confidence from Delighted to generally pleased. All because of two pointless off-season trades, when I've single-handedly turned a franchise with a 0-8 playoff record to the longest running dynasty in NHL history. It's pretty dumb, which is in line with all the AI in this game.

Also the board/fans always look at trades in a vacuum. When you make Cap management trades they don't take your current/future cap into account at all. Which means when you dump an overpaid old player for less than their reputation might warrant (and reputation is pretty much all the AI cares about). They will rage at it as a bad trade, even though it isn't by any means.

Honestly if there was an option to disable board/fan feedback on trades I'd use that and never look back.

One thing I have noticed is the AI isn't very consistent in trades, they seem to value a player differently depending on if you or they come up with a trade offer. I was playing with the Coyotes and have bit of cap issues, with injuries happening and having to bring up players from the AHL my cap went over the limit and I was forced into making a trade to make room. So I offered one of my defenseman that was on a bit of a bad contract and not playing currently. Canucks offer me another RD (which I need + cheaper) + a 3rd. They wanted my RD + 3rd. After the trade went through my game crashed and I didn't have a chance to save it yet.

Went back in and remembered the trade so I offered them the exact same deal. Respons: Player is not of interest as he hasn't seen much action lately. Nevermind, offered him out again...and Canucks come back to me again with the exact same offer! So what has changed? I suppose it would have been fine if they wanted one of my other 3rd round picks which might be a better pick in the end.

saberhagen83 wrote: One thing I have noticed is the AI isn't very consistent in trades, they seem to value a player differently depending on if you or they come up with a trade offer. I was playing with the Coyotes and have bit of cap issues, with injuries happening and having to bring up players from the AHL my cap went over the limit and I was forced into making a trade to make room. So I offered one of my defenseman that was on a bit of a bad contract and not playing currently. Canucks offer me another RD (which I need + cheaper) + a 3rd. They wanted my RD + 3rd. After the trade went through my game crashed and I didn't have a chance to save it yet.

Went back in and remembered the trade so I offered them the exact same deal. Respons: Player is not of interest as he hasn't seen much action lately. Nevermind, offered him out again...and Canucks come back to me again with the exact same offer! So what has changed? I suppose it would have been fine if they wanted one of my other 3rd round picks which might be a better pick in the end.

Loading a save probably results in the game re-running various calculations leading to a bunch of changes in AI behaviour.

It's always a hit or miss.

In my league, the AI Canucks are a powerhouse that in five years, was finalist to the cup in year 1, won 3 Stanley Cup in a row and was again finalist in year 5. They trade key assets around trade deadline for their playoff push and gets "younger" around draft. I don't even know if I'll be able to beat them this year as they are still ranked #1 by the league. So quite please to see the aI doing his job. The Sabres are another team that does well.

I did raise the cap significantly thought... because I was tired of seeing Crosby on waivers after a few years. So teams have a lot of room to work with which does help.

I hev to say that trading is much harder now in 1.4.1.

Building a team from draft seems to be only way when playing with challange rule set ( no trade for ufa,rfa, max sign 1m ufa and maximum 2:1 trades).

I can offer top 10 NHL prospect(19 year winger) and solid 21 year 2nd line center for a good center prospect and they now say no all the time. Its very hard configure the team to be solid in all areas (2 good players, low liners, and prospects even between the positions).

I used a database from here with Vegas added along with Seattle and I have noticed it is an awful lot better trading on EHM now than it used to be, no more using the trade block to pick up great players for 7th rounders...

Anyway before the season I thought I'd make a trade with Seattle and wondered what it would take to get their 1st rounder for 2017 off their hands, the answer was a MASSIVE win for me as while I'm currently 14th of 16 in the West I am competitive... Unlike Seattle who haven't won one game in 26 so far, bear in mind I made this trade BEFORE game 1 of the regular season. I give up my 2017 1st round pick, for Seattles 1st (which probably is #1 as long as they stay awful), 2nd AND 3rd from 2017, so literally for my 1st rounder which isn't as good as my expansion team rivals pick they give me their first 3 rounds of picks... but besides that it is very difficult to trade which is what I myself really like.

there it is, took me a while to get understanding how to put that up right and correction needs to be made by me, it was after I had played 7 games and Seattle had played 6

The team that finishes last won't probably be the #1 pick. In fact it's highly unlikely that it will be. If the lottery isn't rigged at all (but it seems to be) there's just a 20% chance of it happening. I'm currently 0 for ~16 with getting the 1st overall pick when you finish dead last as the only human player. I've picked 4th (so losing all 3 rounds of the lottery) in all but 2 where I picked 3rd (losing 2/3 lottery rounds) which is getting more and more statistically unlikely, but it keeps happening.

Regardless when a season hasn't started and you're trading picks the main factor behind a team's evaluation of their pick's worth vs. yours seems to be the team's reputation. If you won the President's Trophy last year but your team has 7500 reputation and they finished dead last but have 8500 reputation they will think that their 1st round pick is worth a lot less than yours (because clearly a team with their reputation will do better than you) leading to them overpaying so much for your pick. The trade AI is dumb and highly abusable which is why trading 1st round picks is banned from challenge saves, because you can just fleece the AI over and over.

Personally I've won more draft lotteries with the traded for pick of the team 1 spot outside the playoffs than I have by finishing dead last (by a lot). Which is just outrageously unlikely if the lottery worked as it should, so it probably doesn't.

Generally the AI won't start assigning draft picks their "correct" values until after the lottery has taken place, with only slight changes in evaluations prior to the trade deadline in most seasons. I have a lot less data on that though as I used to always make my trades in the pre-season, and these days I won't trade for 1st round picks at all.

Does the game recognise bad contracts when trading and rate the player accordingly?

I.e would Ottawa be happy to get rid of Bobby Ryan or would you the AI just view him on ability rather his contract??

Most of the time they do not care about the contract at all. I've seen teams trading for a specific player but when the deal's done, they have to put another of their key players on waivers or on the trade block to free some cap space.

It doesn't happen that often though. Mainly when your saves goes over 4-5 years and the teams are allowing ridiculous contracts to anyone.

That's a shame. So you can't take a bad contract then if you have cap space to get a deal done.

I'm 30+ years deep in my current save with BKArkitecht's NHL 32 rosters, but I've played a lot of EHM (1100h+ on Steam) on different rosters and the AI, of course, acts more or less the same way regardless of the database. I always play with human GM sacking disabled. In this current save I've been managing the Canucks from the get-go and God knows how many times I would've been fired because of the trades and board/fans disagreeing with them.

Anyway, here's some observations I've made about the AI in general in my many hours of EHM:

• I don't know how the AI "sees" the player/personnel attributes, but on NHL level the teams keep hiring personnel with awful attributes, even when there's a lot better options available.

• In the draft, again, I don't know how the AI handles "scouting", but they sometimes seem to make some unbelievably stupid picks, especially in the early rounds of the draft. OFC, in real life hits and misses happen but these are usually just ridiculous attribute- and stats-wise.

• After the draft, give it a few years - AI doesn't sometimes sign their high picked prospects even when they're more than capable to play on the AHL/NHL, and let them expire. Or, they trade their high pick player away to a new team, and that team lets him walk as well. AI tends to suck bringing new kids to the roster, and young guys get traded around a lot.

• AI is very bad at bringing younger players/draft picks from the KHL to the NHL. KHL draft picks tend to go unsigned, play some years in the KHL and sign an NHL contract in their 30's.

• As stated, you can easily game the AI in trades. Also, more often than not, the AI doesn't value draft picks very much.

• Too many "good" players are placed on waivers, usually because of bad contracts. And nobody takes them, because everyone already has their salary cap full with said bad contracts.

• AI doesn't seem to make enough effort to sign their key-core players, and often the best players go to Free Agency, even if it doesn't really make much sense. If they happen to make a contract extension with someone when he's still on the team, it's usually a good contract, or decent at least.

• During Free Agency, players aged 27+ end up changing teams every 1-2 years, why? Because the AI makes a ridiculous 9 mil $ contract to a player and that same player walks, come next free agency. Rinse and repeat. You very rarely see contracts of 3+ years, on stars anyway, after that point.

• The AI is very bad at "plucking holes" in their system/roster depth. Also, the more the years pass in your save, teams tend to become "empty" as in there are some years when there's barely 30 players on AI roster.

Of course there might still be more I'm missing, but these are from the top of my head just now!

I've come to a conclusion that all this is happening because the AI doesn't value a player's contract, his attributes, or game ratings/stats nearly as much as it does value the players REPUTATION within the league. A player might be a beast on, say, KHL or SM-Liiga, with good attributes, but his NHL/world reputation is low, so he never gets signed to "better" clubs.

I'm no game developer, but addressing this issue might help with the AI problems all around the board. These are just my 2 cents from personal observations - I very rarely even visit this forum to engage in talks like these - but feel free to copy-paste and use if you find this useful at all!

Reading about this thread and trying to create/modify some database... am I right thinking I should recalculate every player reputations to decrease the likeness of weird trades from ever happening? I was trying to find a solution to the trade logic and reading this thread, it seems the AI puts player reputation on the frontline of every player negotiation. Unless I am wrong?

Yeah they indeed overhype players judging their reputation or stars level or the opposite. I don't know if it would work to tweek that though. I'm not pretty good with modifying database.