Would the Santana deal ever happen in OOTP?

[ctorg]

I found myself wondering this today. Would OOTP's AI ever do this deal if it were controlling the Twins? I tend to think not. As far as I can tell, the AI, even in Rebuild mode, doesn't do a whole lot of cost-cutting. It is primarily concerned with getting skill in exchange for skill. It doesn't look at a guy in the last year of his contract, realize it won't be able to re-sign him, and trade him away for inexpensive, marginal talent.

Or does it? I'm going by casual observation rather than any sort of empirical test of any kind. What do you think?

To me, it's deals like this that kind of shine a light on areas of the game that could probably be made stronger. One of them - and it is very complicated, so I can see why it is not there yet - is the ability of a team's AI to assess its situation and chart a course.

[koohead]

well the first hurdle is multi-player deals, which I don't feel OOTP handles as well as it should when compared to real life. I don't recall seeing 3-4 players involved in a deal for 1 star player unless I initiate it. At most, I see 2-1 or 2-2. And usually its 1 major league player and 1 minor league player...

On the other quesiton of how the AI handles rebuilding vs win now, I see the biggest difference during the FA off season period. teams that are set to win now go crazy with spending money (obviously within budget) signing big ticket players left and right. rebuilding teams are far less active during this period. I'll add that just based on casual observation (well, okay...a bit more than casual), I do happily see alot of veterans being traded for younger talent, and its usually from a losing team to a team in contention. Star players do also tend to sign with the better teams as well, at least in my universe.

On a completely different note, I did notice a few of my Japanese players that were originally in the Majors have signed relatively large contracts to go back and play in the NPB towards the end of their career. I thought this was a nice touch by the AI and reflects on the desire of players to play near their homes I believe.

[darkcloud4579]

Umm...well, I don't think it would do it by itself. You'd have to nudge the deal.

I've found a "new" way to made more competitive deals with the computer. While a salary cap is "unnatural" to some degree, it's a good way to make sure teams don't just stockpile players for giggles.

Another thing is, it seems to induce teams around the league to spend more and to make the market more competitive by and large.

But about the dealing, what you could do...is I tend to set trades on "average" and then "hard" and see if I can make the deal on both levels and if I can, I'll pull the trigger. Or allow myself one "average" trade a year, but never to flat "cheat" the game.

I do think you're right that the game needs to have these sorts of "complex" deals, but I tend to think the game would flub it would just like it does other complex tools that it's supposed to do right like scouting and even trading.

It sucks to have to intervene. But at least it's possible. I will often go and sign a guy to a long-term deal on another team if he's a guy I can't bear to see play elsewhere. Like Derek Jeter in New York and I'll often sign the guy and give him a no-trade clause. Or I'll even do that and then prepare to deal him before his current deal expires, using it as a "window" that they give teams in real life to negotiate with a guy they're trying to trade for.

I mean, some stuff you just have to fudge if you want it to be realistic. The game won't be able to mirror that stuff and if you think about it, the "real life" game doesn't mirror that either. It takes a Commissioner's exemption to get the ability to negotiate with a guy who is on another team when you're trying to trade for him or other such things and so..I think that trying to get the game to do "too much" is where we go wrong more than anything.

But I agree with the sentiment, I'm just not confident the execution will be effective.

[boblight24]

I think in order to make these sims as true to real life you have to switch to "Commish Mode" and change things from time to time and I don't think it affects the integrity of the game. For instance, I will often go in and "edit" if a player has a career ending injury. Take last season (2007, default MLB), Kevin Youkilis ruptured his Achilles in Spring Training, boom, done for life. That's not cool, so I adjusted his injury and made it 6 months on the DL instead. He's not done for good and he has still suffered a major injury.

Speaking of blockbuster deals, how's this...
New York Mets don't get the production out of DWright they want so they put him on Waivers and eventually demote him to AAA. They need a 3B for '08, where do they go. South to their arch-rivals. They send Carlos Beltran to Atlanta for Chipper Jones and 2 prospects. I had to read the deal twice to make sure that I wasn't imagining it.

[The Wolf]

No. The trade AI needs work. In Rebuilding Mode it should be making exactly these kinds of deals. Also plain salary dumps. It doesn't.

[wlight1]

Actually, I think the biggest obstacle to this deal getting done in OOTP isn't the AI. If the Mets hadn't been able to negotiate an extension with Santana ahead of time they'd never have given up as much talent as they did, and you can't do that in OOTP. Making a huge blockbuster like this in the game, you'd run the risk that your newly acquired stud would randomly decide he didn't really want to negotiate and you'd lose him right back to FA.

[Aytumious]

Originally Posted by The Wolf

No. The trade AI needs work. In Rebuilding Mode it should be making exactly these kinds of deals. Also plain salary dumps. It doesn't.

I've noticed that teams in rebuild mode will often give up some of their best young players for more, less developed young players. That doesn't make very much sense to me, especially when the already developed young guys are still making the minimum.

[ctorg]

Originally Posted by darkcloud4579

I think that trying to get the game to do "too much" is where we go wrong more than anything.

But I agree with the sentiment, I'm just not confident the execution will be effective.

Ten years ago, this is what I found myself saying about many of the things OOTP does do well now. In time the game will get even better. With each new stage, I often find myself wondering what could possibly come next, and then something better does.

I think that one of the shakier areas of the game now is the AI's ability to assess its needs. This is a complicated thing for sure and has to take into account finances, position in the standings, and even more complex things like relative positional strength compared to other teams in the division/league. Self-assessment affects trades, signings, player promotions, and a variety of other aspects of the game, and it's something that can always be fine-tuned.

[The Wolf]

Originally Posted by Aytumious

I've noticed that teams in rebuild mode will often give up some of their best young players for more, less developed young players. That doesn't make very much sense to me, especially when the already developed young guys are still making the minimum.

Yes. The trade AI really needs work.

[james17]

Originally Posted by The Wolf

No. The trade AI needs work. In Rebuilding Mode it should be making exactly these kinds of deals. Also plain salary dumps. It doesn't.

I know it would be hard to do but I agree that it needs serious work.

[Who Killed Kenny?]

Markus could spend a year on AI and it would likely not reach the point of reality. AI is a hard problem. Unfortunately too, would anyone (other than the ones responding in this thread) pay $35 for a new game where the only "new" feature is better trade AI? I doubt it and Markus prolly cannot spend all his time on this problem. Other bells and whistles are needed in order to sell the game. 2 cents.

[The Wolf]

Nonsense. You have ridiculously overestimated the amount of AI programming required to fix the trade AI.

[darkcloud4579]

Originally Posted by ctorg

Ten years ago, this is what I found myself saying about many of the things OOTP does do well now. In time the game will get even better. With each new stage, I often find myself wondering what could possibly come next, and then something better does.

I think that one of the shakier areas of the game now is the AI's ability to assess its needs. This is a complicated thing for sure and has to take into account finances, position in the standings, and even more complex things like relative positional strength compared to other teams in the division/league. Self-assessment affects trades, signings, player promotions, and a variety of other aspects of the game, and it's something that can always be fine-tuned.

Well elucidated.

[Mike Donlin]

I remember tracking this--that is, rebuild, win now, neutral--in the last beta go-round and lobbying that it really needed work. I think some tweaks were made, but obviously more needs to be done.

[Who Killed Kenny?]

Originally Posted by The Wolf

Nonsense. You have ridiculously overestimated the amount of AI programming required to fix the trade AI.

Well thats good to know! Markus, Wolfie says this should be a easy fix! Make it so!

[Riverdome]

Originally Posted by wlight1

Actually, I think the biggest obstacle to this deal getting done in OOTP isn't the AI. If the Mets hadn't been able to negotiate an extension with Santana ahead of time they'd never have given up as much talent as they did, and you can't do that in OOTP. Making a huge blockbuster like this in the game, you'd run the risk that your newly acquired stud would randomly decide he didn't really want to negotiate and you'd lose him right back to FA.

BINGO. I just decided against a trade that would have brought the best SP in my league to my team. I was giving up solid hitting talent but I didn't have any way to know if the SP would even talk to me, let alone actually sign an extention.

A sign and trade module would be a nice addition to the sim.

[darkcloud4579]

Originally Posted by Riverdome

BINGO. I just decided against a trade that would have brought the best SP in my league to my team. I was giving up solid hitting talent but I didn't have any way to know if the SP would even talk to me, let alone actually sign an extention.

A sign and trade module would be a nice addition to the sim.

It would be a disaster. This stuff isn't "realistic". Sign and trade deals are a new advent related to the modern economics of baseball. It's where the commissioner allows a team a negotiating window to deal with another team.

So to allow the game to do it intuitively is simply a bad idea. If people want to do this, you go into commish mode and negotiate the deal. Of course, the problem is, in the game it's not that hard to do because they'll accept it if you meet their demands.

[The Wolf]

Originally Posted by Who Killed Kenny?

Well thats good to know! Markus, Wolf says this should be a easy fix! Make it so!

If you think AI programming of something like that is all that hard, then talk to a real AI programmer about what they do every day. Or even in a day. There are consulting programmers that I work with who could knock an AI mod to an existing system like that out in a week. Tested, too.

You think that it's a lot harder than it really is.

[The Wolf]

Originally Posted by Mike Donlin

I remember tracking this--that is, rebuild, win now, neutral--in the last beta go-round and lobbying that it really needed work. I think some tweaks were made, but obviously more needs to be done.

And I'm sure that they will be in Version 9. Especially if we all keep pointing out that the trade AI needs work and you keep lobbying it.

[ctorg]

Originally Posted by Who Killed Kenny?

Markus could spend a year on AI and it would likely not reach the point of reality. AI is a hard problem. Unfortunately too, would anyone (other than the ones responding in this thread) pay $35 for a new game where the only "new" feature is better trade AI? I doubt it and Markus prolly cannot spend all his time on this problem. Other bells and whistles are needed in order to sell the game. 2 cents.

I've seen this sort of reasoning get used to knock down a lot of ideas over the years. It's not that I want Markus to spend all his time on the AI to the exclusion of all else. I know there are other aspects of the game that need work, improvements, etc., and there need to be new features. But I really think that the very core of the game should always be improved. Sure, it will probably never be as good as a real human manager (at least not until computers reach sentience or something ), but it can always be a bit closer than it is. Markus knows this and makes updates accordingly.

Originally Posted by The Wolf

And I'm sure that they will be in Version 9. Especially if we all keep pointing out that the trade AI needs work and you keep lobbying it.

This is the key, and it's why I think it's important to talk about where the game needs improvement. It's a given to me that the game is great. In my opinion, it's the best game ever, in any genre, because I've gotten more enjoyment out of it than any other, by a long shot. But it can always be even better.