trading preferences REALISTIC

[Orcin]

Perhaps the Nationals don't want McCutchen's selfish attitude. Keep in mind that the game does not accurately model personalities.

[SirMichaelJordan]

If they aren't in win now at best I don't understand why would they make that trade. Thats 13-15 mil on the books tied up 3 years for a 29 year old who will end up blocking a young prospect. Even then you'd have a better chance trading Clutch to a contender if he had 1 year left on his contract (rental). You'll probably be able to make that trade on the hardest setting on MLB The show but this is not The Show. Also most uneven deals like that you would have to offer up money along with the player. Trading in this game (on harder levels) is much more complicated than OVR matching or trade this guy who is miles better for a average guy philosophy other things matter outside of OVR. Like the human player, the AI is going to try to buy low and sell high and that does not mean the trading is broken on harder levels. People are just used to dumb trading AI.

At the end of the day just lower the difficulty...

[Elektranaut]

This is interesting....for years now standard advice has been to trade on Very Hard,and favor prospects otherwise its too easy to fleece the AI.I've noticed in OOTP 16 trading has almost dried up completely in my league.I am even beginning to miss those ludricously urealistic trade offers you used to get.I think I'll try Hard/Neutral,see how it goes.

[Orcin]

Originally Posted by Elektranaut

This is interesting....for years now standard advice has been to trade on Very Hard,and favor prospects otherwise its too easy to fleece the AI.I've noticed in OOTP 16 trading has almost dried up completely in my league.I am even beginning to miss those ludricously urealistic trade offers you used to get.I think I'll try Hard/Neutral,see how it goes.

Standard advice?

[IsaacR]

Probably 90% of real life proposed trades never come to fruition either, even if they're fair to both sides.

[Orcin]

Trading is a matter of circumstances, not trade difficulty. You can fleece the AI on any setting if you find the right set of circumstances.

The trade below is accepted on Very Hard / Heavily Favor Prospects. The AI heavily favors my prospect more than his, and the AI is very wrong.




This is a recalc league. so the AI should be able to do a quick check of its own historical data and learn that Sonny Jackson accumulates 2.2 WAR over the remainder of his career (1966-74) while Steve Carlton has a cumulative WAR of 83.3 to go with his 4 Cy Young awards. John Hoffman, the other player in the trade, played a total of 8 MLB games in his career, none after 1965.

Clearly, this is a lopsided rip-off of the AI, achieved on Very Hard / Heavily Favor Prospects. I suspect that the AI accepts it because Jackson is capable of playing shortstop (albeit poorly) right now while stealing 50 bases this year, and Carlton is only a reliever at age 20. So the "win now" AI wants Jackson.

There is no setting that protects the AI from being fleeced in a historical league. I have posted before that the AI should be able to "see" the same future as the player. In lieu of that, I have to play with house rules to prevent myself from assembling a stable of HOFers. At that point, I might as well use Hard / Neutral so that I can at least make reasonable competitive trades for non-star players.

It is just as easy to fleece the AI on Very Hard / HFP (or Neutral) in a fictional league, except that you don't have the stats to prove it. But it can be done, the usual ways being multiple mediocre players for one superstar or catching a team in rebuild mode to give it 4 middling prospects for its best player.

[SirMichaelJordan]

Originally Posted by Orcin

It is just as easy to fleece the AI on Very Hard / HFP (or Neutral) in a fictional league, except that you don't have the stats to prove it. But it can be done, the usual ways being multiple mediocre players for one superstar or catching a team in rebuild mode to give it 4 middling prospects for its best player.

In a league with a high TCR those 4 prospects could easily become 4 MLB starters. I was a victim of trading away prospects that became multiple time AS's plenty of times. In a league like mine you may feel like you got a steal but 5-10 years later you may have regrets. But at the end of the day that was one of the main points of playing with a high TCR because the human is aways going to have an advantage over the AI.

Like you just proved, trading on very hard is not impossible it just takes harder work which can make it frustrating because of the trading mechanics in the game.

In a historical league I would think the user already has an advantage over the AI by default. Does the AI have the ability to read future stats? How is potential ratings handled in that mode? I also think the fact Carlton was listed as a RP made that trade possible as RP really do not have much value in OOTP.

[Westheim]

The question is whether you can really compare a historical setting with a MLB quickstart or fictional setting. You know about Carlton's future, but the AI doesn't. In the other two settings neither of you will know about the future (unless if you actually can see the future, then I need some expert opinion on a few of my cubs with striped tails, and tomorrow's lottery numbers).

Yet, Orcin, there's still points to be made for the broken state of the AI. Since I know you're following the Raccoons, you might also remember the only deal for a first-rate prospect I managed to pull off while perpetually rebuilding since migrating to 16: the Ryan Miller deal with the Titans (Mays doesn't grade first-rate, and I probably overpaid for Quebell)

To sum it up for non-followers of the Furball cult, the Titans had their fourth outfielder, Christian Greenman, on the trading block for a full season, a slugger with a few holes in his swing, and I tried to trade for him during that season in July (when I couldn't come to terms with the demanding AI) and August (when a waiver deal fell through), and the trade was finally done in January. The offer that the Titans, who merely romped up 117 wins and a championship in that season, took to think about:

Titans get:
26-yr old MR Manuel Martinez (6-6, 3.53 ERA, 9 SV) - high K, 1 stamina, closer potential
28-yr old C Mark Thomas (.240, 11 HR, 90 RBI) - a dud

Raccoons get:
28-yr old OF Christian Greenman (.244, 77 HR, 226 RBI) - rarely a regular starter for them
20-yr old A SS Ryan Miller - looks like stellar D, high OBP, very good speed, AND some pop

In that situation, the Titans come back and ask for another dud catcher, who was a desperate waiver claim by the Raccoons during the season, 27-yr old C Freddy Rosa (.214, 16 HR, 65 RBI - those are all career numbers, btw.), while they already have *three* catchers under contract that are all by bounds and leaps phenomenally better by magnitudes than the two duds, including a regular All Star with a power bat, and one of which they had just signed as free agent two weeks prior.

I did that deal, because I've gotten enough grief from the trade AI, and it's not my job to look after a 117-win championship team when my perpetual rebuilding project is heading into its second decade. In the end, the Titans waived both those catchers (no takers, surprise). But they surrendered a top notch shortstop prospect (AND a capable batter!) for a right-handed reliever (a very good one, but still...), while a dozen much fairer deals I worked on were just plainly refused. Hence: broken.

I won't go into the 15-some trade offers I got from the AI during that offseason, where every single one followed the same principle, dumping an underwhelming, overpaid catcher onto me for a good major leaguer AND a second-rate prospect (Raccoons have no first-rate prospects......).

Of course we don't know yet whether Miller will ever make it to the majors and whether he'll be a Carlton or a Hoffman, but the AI does value Miller highly, because the second I traded for him, other teams wanted to get him, and we won't know for some months with my speed of playing.

Unless I can win that lottery.

[Orcin]

Originally Posted by SirMichaelJordan

Like you just proved, trading on very hard is not impossible it just takes harder work which can make it frustrating because of the trading mechanics in the game.

Yes, my point is that trading is not impossible on any setting. It can be tedious and it sometimes doesn't happen exactly when you want. But trades that favor the human are there if you look hard enough.

Originally Posted by SirMichaelJordan

I also think the fact Carlton was listed as a RP made that trade possible as RP really do not have much value in OOTP.

I just changed Carlton to a starting pitcher using the editor, and the trade still goes. Of course, he is still a 1-star SP at age 20 with low stamina, but the AI should be able to see his future and know what he becomes. If not, there is a huge improvement waiting to happen for those of us that play historical leagues.

FYI Jackson's potential is 2.5 stars as a SS, overall is 4 stars. Carlton's potential is 4.5 stars as a SP, overall is 1 star. The overall should not matter if the setting is Heavily Favor Prospects, because Carlton is obviously the better prospect even if you only know the ratings (and not names or future stats).

Originally Posted by SirMichaelJordan

In a historical league I would think the user already has an advantage over the AI by default. Does the AI have the ability to read future stats? How is potential ratings handled in that mode?

I believe the AI can read potential - at least I hope it can. I don't know why the AI would be unable to see potential. It appears to me that the AI cannot see future stats, but only a dev could answer that one for sure.

Yes, the user has an advantage over the AI in a historical recalc league IF the AI cannot see future stats, but there is no reason for that. In a recalc league, the AI should be as good as Deep Blue.

[Orcin]

Originally Posted by Westheim

The question is whether you can really compare a historical setting with a MLB quickstart or fictional setting. You know about Carlton's future, but the AI doesn't. In the other two settings neither of you will know about the future (unless if you actually can see the future, then I need some expert opinion on a few of my cubs with striped tails, and tomorrow's lottery numbers).

Did the devs make one trading AI for historical leagues and another for MLB quickstart and fictional? If not, then you can compare them. The AI makes poor judgments at times regardless of playing mode. The difference is that they can be seen immediately in a historical recalc league, whereas it takes several seasons of play to know in a fictional mode.

[PSUColonel]

Is this a fair trade?...Just want to see what people think. The players are all as they would be at the start of the 2015 season in the MLB roster set.

[rpriske]

Originally Posted by Orcin

This is a recalc league. so the AI should be able to do a quick check of its own historical data and learn that Sonny Jackson accumulates 2.2 WAR over the remainder of his career (1966-74).

Uh, no. The AI should never look at data for years that have not yet happened.

[Spritze]

Originally Posted by rpriske

The AI should never look at data for years that have not yet happened.

Then the humans should not either and so Carlton for Jackson/Hoffman should not an eyebrow raise.

[ConStar]

I've been having similar issues. As soon as I bought 16 and moved my league up to it, trades just flat-out stopped. And that was going from "hard" setting in 15 to "average" in 16.

I finally had to move it to "easy" just to get anything done. I've got a lot of self-control and I'm not prone to trying to fleece the AI, so I could set it to automatically approve any trade if such a setting existed and I still wouldn't cheat (what's the point? It's a computer game against a box). I still get turned down about half the time on "easy," and that's with me sometimes purposefully giving up too much in order to assure myself I'm not fleecing the AI.

[NoOne]

if the team has a good prospect, they have less interest, too. their scouting results are different than yours. maybe it looks at next year's FA crop in its evaluation. expected role on the target team may not be CF. depending on the complexity of the ai, you might need alot more info to understand why they don't want your vet. even the position will be significant in the decision. an aging CF probably has less value than most position, cetrus paribus.

normal difficulty seems to limt our ability to take advantage of the AI. there are multiple ways to mine players, and a human can do a better job at every one. imo, i don't think it is skewed on normal to the AI - i think it still favors the human, slightly. i still find ways to take advantage to maximize deals, lol. i just can't help myself sometimes. but, that's a matter of perspective on value. i definitely have a different idea than most on that topic. which is why i have less frustrations from the trading on normal.

i've been debating chanigng the favors option, if it helps the small market teams get more prospects as a way to even the playing field a bit. i think favoring vets would accomplish that, since small market teams sell talent before their contracts explode. as of yet, i still use neutral.

however, i'd be worried about its derivative effects beyond my trade screen. would it make some AI gm's significantly worse at their jobs? would the ones that favor vets be at a disadvantage by always buying high? also, a small market doesn't always hire a gm that favors kids, so changing that setting might not be as localized as i think.

i think it's a great option to have, but neutral doesn't require research to see if it has any adverse effects.

if you want easier trades, do it. there's no rule that you have to make it like the real world. imo it defeats the purpose of a simulation, but opinions are like a**h**es. everyone has one.

[Orcin]

Originally Posted by rpriske

Uh, no. The AI should never look at data for years that have not yet happened.

Why not?

[rpriske]

Originally Posted by Orcin

Why not?

Because it HASN'T HAPPENED.

The AI shouldn't have some magical crystal ball that lets them see the future.

Human players shouldn't either, but hey... it's your game. Do what you want.

How can this be a simulation if both the game and you don't act as if you are there... and then?

For my historical games, I get around the fore-knowledge thing by not overriding my scouting director when it comes to the draft. (I make a specific exception in my Blue Jays dynasty for players who I consider 'classic' Blue Jays. Just because I want to. As I said, your game...)

Where I step past the 'it's your game' mantra is of you want the AI to use real world future history, because if the AI does THAT the game is absolutely destroyed as a simulation. Historical play would become completely pointless.

[Orcin]

Originally Posted by rpriske

For my historical games, I get around the fore-knowledge thing by not overriding my scouting director when it comes to the draft. (I make a specific exception in my Blue Jays dynasty for players who I consider 'classic' Blue Jays. Just because I want to. As I said, your game...)

So you use a house rule to level the playing field. I prefer to have the AI fully capable of beating me, and play to see if it can.

[rpriske]

Originally Posted by Orcin

So you use a house rule to level the playing field. I prefer to have the AI fully capable of beating me, and play to see if it can.

Completely eliminating the ability to use the game as a simulation.

[Orcin]

Originally Posted by rpriske

Completely eliminating the ability to use the game as a simulation.

This makes no sense. Why would this option eliminate the ability to use the game as a simulation?

Regardless, there is no harm if the option is made available to allow the AI to see future performance in a recalc league. You merely don't select it and continue to play the game your way.