daahdeedaa Posted March 30, 2015 Share Posted March 30, 2015 I mentioned this in the big thread but something is definitely off. The AI does not value your picks compared to theirs properly. Also I feel the value of players compared to draft picks is deflated for the most part though they will pay a 4th rounder for anything that breathes. Some scenarios that are clearly broken 1. You can offer say pick #40 (mid second rounder) in the draft and teams will give you at best a 3rd round pick in the same draft. Most of the time they'll only give you a 4th or 5th round. It's not just using the shop around feature I have offered multiple teams a second round pick for various lower draft picks and they just refuse. 2. I can offering picks #30, #40, #50 around pick #25 and the AI refuses almost every time (tried with multiple teams). It's not that the AI loves a certain player and refuses to trade. They just clearly over value the pick or under values yours. Thoughts on other valuations that need work but require more research/external analysis 1. Teams will give you 4th round picks for almost anything including really crappy prospects or 30 year olds buried in the AHL. 2. 3rd round picks and higher are a lot harder to acquire 3. If you offer a 2nd liner/2nd pairing defenseman usually gets you a 3rd round pick, occasionally a second rounder (2nd liners fetch at least 2nd rounders and usually mid-late 1st rounders in real life, especially at the trade deadline in the NHL). Link to post Share on other sites More sharing options...
dugger999 Posted March 30, 2015 Share Posted March 30, 2015 Funny, because in my game its the opposite. The AI gives no worth to draft picks. Got buffalo and edmonton 1st pick for for almost nothing. Also when i try to trade if i offer a first pick the answers is no but if i substitute the first pick for the right of a mediocre unsigned prospect the trade will be accepted. Link to post Share on other sites More sharing options...
Mars_Blackmon Posted March 30, 2015 Share Posted March 30, 2015 Funny, because in my game its the opposite. The AI gives no worth to draft picks. Got buffalo and edmonton 1st pick for for almost nothing.Also when i try to trade if i offer a first pick the answers is no but if i substitute the first pick for the right of a mediocre unsigned prospect the trade will be accepted. Im with you, AI give no worth to draft picks. It may get altered slightly based on team needs and direction but draft picks need more value. Link to post Share on other sites More sharing options...
CapEm Posted March 30, 2015 Share Posted March 30, 2015 3. If you offer a 2nd liner/2nd pairing defenseman usually gets you a 3rd round pick, occasionally a second rounder (2nd liners fetch at least 2nd rounders and usually mid-late 1st rounders in real life, especially at the trade deadline in the NHL). Not that I disagree, but deals at the trade deadline are difficult to truly value, because most times there is a premium on specific types of trades, and also most trades are situational, based on a teams position in the standings (hence the term "rental player"). Example - a team that is expected to make the playoffs, might give up a 2nd round draft pick for a depth player, if it feels it makes sense for a playoff run, but come summer time, specifically around the draft, teams place a much higher premium on draft picks in late June, when that same depth player might only command a 4th round pick. I'd be curious to know how EHM model's this, and if the AI's draft pick valuation dynamically changes as the trade deadline approaches, and its position in the standings. And then how it changes again in June. Link to post Share on other sites More sharing options...
daahdeedaa Posted March 31, 2015 Author Share Posted March 31, 2015 Not that I disagree, but deals at the trade deadline are difficult to truly value, because most times there is a premium on specific types of trades, and also most trades are situational, based on a teams position in the standings (hence the term "rental player"). Example - a team that is expected to make the playoffs, might give up a 2nd round draft pick for a depth player, if it feels it makes sense for a playoff run, but come summer time, specifically around the draft, teams place a much higher premium on draft picks in late June, when that same depth player might only command a 4th round pick.I'd be curious to know how EHM model's this, and if the AI's draft pick valuation dynamically changes as the trade deadline approaches, and its position in the standings. And then how it changes again in June. Like I said the last 3 things are something that need more research to determine the value of a player and prospects vs pick and in different rebuilding stages/times of the year. Not a small task at all. The first 2 are just straight up broken. Here's an example below. It doesn't matter if you're trying to get veterans or rebuilding with young players. There's no way in any scenario the AI should value your picks vs theirs in this manner (unless Craig MacTavish is the GM of the team). Link to post Share on other sites More sharing options...
AaronV Posted March 31, 2015 Share Posted March 31, 2015 Theres obviously no difference in the code for AI vs AI trades or AI vs player trades with this example. Its set so the AI doesnt rip itself off, not a big issue but would be nice to have it coded both ways. In all of these hockey games past and present the trade/picks has always been off. All we can do is post screens and examples of it over the next few months so the developer's can make adjustments and hope the code is smart enough to handle it. As far as draft picks I see 2nd threw 7th round picks being seen as just that but 1st round picks are another monster. 1st pick over all should have its own weight and after that maybe every 5 picks in the first round be weighted differently so you have the 1st and then 2-5, 6-10 and so on. I think this would go along way to straightening out draft picks and what would be left is setting weights on players. Even better would be 1-5 have there own weight and then 6-10, 11-15 and so on. Link to post Share on other sites More sharing options...
SI Staff Riz Remes Posted March 31, 2015 SI Staff Share Posted March 31, 2015 Some of the discrepancies in draft pick evaluations in trades are due to some data not being exactly up-to-date in all databases, especially with last season performance of the club etc. These obviously affect the evaluation of draft picks in trades. There will be some new developments in this area in the first update. Link to post Share on other sites More sharing options...
DJLM82 Posted March 31, 2015 Share Posted March 31, 2015 Just as a test I offered my 3rd round pick (61 overall) for a 4th pick from a team I could not remember that was mid pack (around 105 overall) and they did not accept and wanted another 3rd to part with their fourth. I've got the gave saved on this very day Riz if you wish me to send you the file. Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.