Trade AI Discussion Master Thread (trying to gather everything in one place) - Page 3

RonnieDobbs · 04-21-2003, 12:35 PM

My suggestions, put as briefly as possible...... (if I mention something already posted, I apologize, but I am simply reiterating the previous point).

1- AI gives up valuable closers too easily, to the point that I often "raid" opposing bullpens of teams I expect to meet in the playoffs. Yes, relievers are not as valuable as starters, but I shouldn't get Billy Wagner and Octavio Dotel for a three star prospect and an extra OF.

2-AI also greatly undervalues very good young starters. I shouldn't get Mark Prior for Cliff Lee and a low level prospect, Josh Beckett for Adam Dunn. Top-shelf young starters should cost the house.

3-I strongly agree with the posts that suggest "punishment" by rival GMs for pulling out of an offer. In real life, I would have gotten a black eye by now for all of the dickering I have inflicted on rival GMs.

4-A "shop player" option should be included, but there should be a resulting penalty if you fail to trade the player after "shopping" him. Just imagine how disgruntled a normal MLBer would get after hearing that he was being shopped to any and all takers.

marc420 · 04-21-2003, 03:43 PM

First off, let me say I currently have OOTP4, and have been holding off on OOTP5. The reason I'm reading this thread is to try to find what improvements have been made, and I'm mainly concerned with areas like Free Agents and Trades.

To me, this new Trade interface sounds wonderful, and I think Markus has pretty much convinced me to go ahead and get OOTP5 by putting this in. THANKS!

As to whether other GMs get upset by backing out of a deal, it really depends on how good the interface of the game is. In OOTP4, the only way to find out what sort of trade another team would take is to propose it. If the game forces you to do that to find out what value a team places on a player, then it shouldn't penalize you for doing this.

And in general, I do believe you should be able to get to a point where you tentatively agree to a team. You'd basically say something like "That's very interesting, let me think about it and talk it over on this end and we'll get back to you." If the other team things its a good deal, they'd be willing to wait a few days. Or maybe you'd get a message back the next day that says, "California's also made us an attractive offer, get back to us today or we'll make the deal with California".

If I was a GM thinking about trading Sammy Sosa, I wouldn't do it on the spur of the moment during a phone call. I'd work out the parameters of the deal. But then I'm going to go talk to my manager and see how he likes the deal. I'm going to go talk with the Captain of my team and some other key players on the team to see what they think of the deal. I'm probably having meetings with my scouts talking about the deal. Just as pure CYA, I'm going to go talk to my owner about the deal and see what he thinks about trading away Sammy Sosa. In general, it seems very reasonable that after negotiating the parameters of a deal with another team, a GM and a club would stop and think about it and then say Go or No Go. And I can't see another GM getting too upset about that.

Now maybe if I'd worked up 15 trades this week with a GM, then said No to everyone, then I'd expect to get a "stop wasting my time" response back. But it should take something pretty severe to get there.

And besides, there's only 30 GM's to deal with. Even if one GM doesn't like another one, there's a very limited group of people you have to deal with.

As to whether there's a penalty for "shopping" a player, that would seem to depend on how much the player wants to stay with a team. If the team is 40 games out in August, and there's rumours that the player might get traded to the Yankees. Then if the deal falls through, I can see the player saying "what the heck" and tanking it for the rest the season.

But if the player is on the Yankees, and the rumour is he's about to get traded to the team 40 games out to get another pitcher, then if the deal falls through, I can see the player being elated that he didn't get traded. For some players in some cases, this could serve as an incentive to play harder.

Age of the player could make a difference too. For a young player, I can see this upsetting them. For a veteran player, they may just shrug it off. I can see them saying "Baseball's a business, this happens all the time", and then start telling stories about the time he was in Cleveland and there was a rumour .....

CubsFan · 04-21-2003, 04:32 PM

Quote:

Originally posted by marc420

...
Instead of agreeing immediately to the trade, teams would instead signal that they think this is a decent idea for a trade, but not agree to it on that day. Then team would offer the players they are offering to your team to the other AI teams, just to see what they could get.
...
So I'm thinking the AI teams could do the same. Trading should be a process that takes place over several game days. Now if you make an offer that is just incredibly wonderful and the AI thinks its a steal, they may immediately accept. Otherwise, they'll negotiate with you to work out a tentative agreement. But then they'll shop the players in that tentative agreement around the league.
...
This way, the various AI teams would act as a check on each other against stupid trades. And the game would resemble more of the marketplace it should. If a team is willing to trade a player, what they are willing to accept for that player would depend more on the value the various GM's in the league put on the player, not just on one AI's calculation.
...
In programming terms, the AI GM would set a floor value for a player. Below this level, the AI GM would not consider trading a player, as the AI GM feels that keeping the player on the team is a better option than any deal offered below this level.

The AI GM would also set an upper level. Any trade offer above this level would indeed be immediately accepted, as the GM who offered a deal this good appears to be a complete idiot and the AI decides to seal the deal before he changes his mind. (Note: at this point, the other GM, if its an AI should say something like "this is interesting, we'll get back to you." Then shop the deal around).

Between those two levels, the final value of make-up of the trade would depend on what the market allows. That would happen over several game days as teams make various trade offers and counter offers to each other. Only when a team feels like it has the best offers on the table that it is going to get should the AI GM decide which offer to take.

Just wanted to say that I think this is a sound idea, and that it seems more practical to implement into a game than many other good ideas that come around. To avoid endless AI cycles of searching for better offers, the feature could be implemented with a random but set number of offers that it would seek before finally accepting or rejecting the offer that had been made originally. An additional feature could be the AI simply deciding it did not want to trade the player(s) involved.

One concern is that such a system could lead to an undesirably high number of AI to AI trades through some sort of unintended loop. But I believe this is an intelligent and reasonable approach to expanding and improving the Trade AI (as opposed to "fixing" it, as it's far from broken).

marc420 · 04-21-2003, 04:47 PM

Two columns from ESPN's website from Trade Deadline Day last season.

I know I'm just mouthing off about the way I think things should be. Maybe I get lucky and hit a good idea once in a while.

But I'd suggest that people read these two columns, then ask themselves if in playing OOTP against the AI, they feel like they were in anything like what was described in these two columns.

Peter Gammons, Aug 3, 2002 "Going inside on Deadline Day"
http://espn.go.com/gammons/s/2002/0803/1413691.html

Jayson Stark, July 31, 2002 "Deadline Day goes quietly, but dealing not likely done"
http://espn.go.com/mlb/columns/stark...n/1412625.html

Some quotes from each
Stark ----

A few weeks ago, the Cardinals didn't seem to have much to deal. But when they were through, they'd won the Scott Rolen Sweepstakes for a very modest price

....
General manager [CLEVELAND] Mark Shapiro knew early he'd be trading a bunch of veteran players. So he did a study of all the deadline deals of the last 10 years -- and concluded the best trades were usually made for prospects who were years away.

....
But fellow GMs complain that [TAMPA BAY]Chuck LaMar slapped Mercedes price tags on the Hyundais on his lot, and he aggravated many potential customers.

We asked about Esteban Yan, and he told us it would take two top-quality prospects, and a fringe major leaguer," grumbled one team. And we heard similar talk all over both leagues. As for LaMar's most marketable player, Randy Winn, he's so young and cheap, we probably would have driven a hard bargain for him, too. One disgruntled GM said: "What's he need him for? So he can stay there and help them win 50 games?"
....

But two major-league sources say [KANSAS CITY]Baird went down to the wire talking about a potential three-team deal that apparently would have involved the Mets and Blue Jays. The Mets would have gotten Byrd. It appears the Royals would have wound up with Toronto's big-time second-base prospect, Orlando Hudson. The sticking point may have been the Mets' reluctance to part with their best pitching prospect, Aaron Heilman.
....

Always-active Mets GM Steve Phillips had potential deals flying all over the map. Another trade reportedly would have sent relievers David Weathers and Mark Guthrie to Anaheim for Schoeneweis.
....

Names being shopped on Deadline Day
In the hours before the deadline, these were some of the names that reportedly were being dangled actively but never went anywhere:

Tom Gordon, Jeff Fassero, Doug Glanville, Delino DeShields, Roberto Hernandez, Tanyan Sturtze, Bubba Trammell, Lenny Harris, Matt Stairs, Brandon Larson, Neifi Perez, Mark Redman.

Gammons ---

The Yankees, you see, could have had Floyd, but Steinbenner so overreacted to Enrique Wilson's mishaps in right field in a game against the Mets on June 29 that he overruled the sensibilities of "his baseball people," called Toronto owner Paul Godfrey and then bought Raul Mondesi on July 2. If Steinbrenner had waited an hour, he'd have found out that as the Mondesi deal was being completed, Marlins team president David Samson had Floyd's agent Seth Levinson in his office informing him that that they could not re-sign Floyd and he would be put on the market that afternoon. Then earlier this week, Steinbrenner found out the Red Sox were the leaders in pursuit of Floyd and he tried to jump in to block the deal.

"It got down to the Yankees and Boston," says Expos GM Omar Minaya. Montreal asked the Yankees for outfielder Juan Rivera and first baseman Nick Johnson. "They chose not to do that," says Minaya, "and we went with Boston."

Oakland had the best players to offer for Floyd, but there were two roadblocks: Floyd has a no-trade clause to Oakland he wouldn't waive, and the A's also needed the Expos to take back the final $2 million in the deal; Minaya wanted the $2 million to use to acquire a closer in case, as he puts it, "we get back into the race."

Oakland then tried to get Atlanta interested in taking Floyd at the end of a three-way deal, but that didn't work, leaving only the Red Sox and Mets. And Minaya did not want to improve the team with whom he is competing in the wild card and the NL East.

....
"Even as a seller, it's more complex than many realize," says one NL GM. "It's understandable that some teams wanted to wait, but obviously the returns diminished the closer we got to July 31. It's all about buyers who could take on contracts. It'll be the same thing in August, with claimings."

....
Most GMs believe there is a major market correction working, with very few teams willing to go spend on free agents this winter. If the Indians, Cardinals and Red Sox tried to sign Thome, Rolen or Floyd right now, they would have to do so at 2001 prices. Thus they have to wait and let the market dictate their value, which in each case may be far less than they expect, despite each player's ability and character.

"A classic example is Johnny Damon," says one GM. "He essentially signed with Boston for what he turned down in Kansas City. It wasn't his fault. The market shifted downward."

WLight · 04-21-2003, 05:27 PM

Quote:

Originally posted by marc420
The agent's response should have been something like "That's a very interesting offer. We'll get back to you." Then the agent would contact other GM's to see if any wanted to top my offer. If not, the agent might call me back and agree to sign. If someone topped my offer,then the agent would not immediately sign with that team either, but would call me back and see if I wanted to raise my offer.

But they do do that. The only times, in my experience, that a free agent will sign with you on the same day you make your offer, is in one of two situations: 1) any time where you offer way more than he's asking, or way more than any other offers he's received, or 2) you're the only team making a palatable offer to a mediocre player late in the FA period. I've been burned time and time again by making a solid offer to a guy I really want early in the FA period, he 'favors' my deal for several days but doesn't sign, then later he comes back and says another team has offered him more money! I frequently end up losing out on these players.

anpham · 04-22-2003, 12:39 AM

Most people who post on this web board you can probably classify as die-hards who know baseball pretty well and want an as true to life simulation as possible. But it’s possible that the majority of people who buy this game and who don’t post on this web board may just be the occasional baseball fan who read some reviews and decided he/she wanted to try managing a team and this one was the one that got the best reviews. I looked around the Internet for baseball games and this one got good reviews on most of them and consistently. However the games this one was competing against were very different in that you used a gamepad to play those games whereas this one is more of a drag-and-drop, managerial, GM experience. I played Madden NFL 2003 and found the franchise portion of the game more enjoyable than the playing part so it was an attractive buy for me. But if I’m searching for baseball games on GameSpot or something I would kind of expect it to be the typical PS2 type of game but found that is nothing like it. However the reason it does get good reviews generally is because it isn’t your typical PS2 game and is pretty good at not being one. I sat around and thought about which game to buy whether it was High Heat 2004, MVP Baseball 2003, or OOTP5. High Heat got a 6.4, MVP 8.0, and OOTP 8.5. They are all similarly priced. I’m familiar with the first two games but never heard of OOTP until I read about it on GameSpot because you can’t find it in stores or at Blockbuster yet. Since it’s also a game that is sold over the Net then obviously the reviews of it on the Net are going to be pretty important. Of course I looked at some screenshots and tried to see what the game offered me but I did lean towards MVP at first because it was produced by a trusted name. I do think this game is trying to be different and gear itself towards a certain kind of buyer that wouldn’t buy the High Heat’s and MVP’s, but I also think they would want to avoid as many refunds as possible for people like me who read reviews and saw this one got the best and decided to buy it based on the reviewer score. Then if you’re also trying to design a game that you’re hoping will eventually start to stock the shelves at stores then you have to start developing it so that it fits the casual buyers needs, expectations, and level of experience.

I'm sure most casual users will prefer it be the way it is where they can occasionally get the players they want on their team for fairly little. If you start off as a Tampa Bay fan then your team starts off with almost nothing and there’s fairly little you can do transaction wise. However with a hole or two you could possibly trade for some talent and it'll at least make it possible for you to win using your favorite team. Although I'm not sure if this is intentional or not I've seen a game where they did make it intentional. Madden NFL 2003 is the only team management game I've ever played before this one, but in that game there was a "bug" where you could trade your first round draft pick for any teams 1, 2, and 3 picks. Even if you won the Super Bowl the previous year and the team you're trading with finished last (and the game does factor in your team records later on) you could still pull off this trade during the first 10 or so years of your franchise. After that they stop letting you trade your first pick away for any teams first 3. This bug would be pretty easy for anybody to find and it gives the user a chance to get ahead and build a better team. It's one of those things that wouldn't be missed during bug testing and doesn't make sense when you see how the computer trades with you in the later years. This helps a person build a team they could win the Super Bowl with and obviously winning makes any game more enjoyable. No one likes to lose continually so in my mind it was a marketing strategy that the Madden people decided to let you do for the first 10 years so you could build your team and be competitive. Although the percentage of people who are going to play like this could be fairly low or high, if this “bug” made 10% of the buyers happier to have it available then why not build it into the game. For the purists who like to win on their own they can just ignore it and feel better about themselves that they had the chance to cheat and didn’t, and feel better when they win because they didn’t cheat. It works both ways.

Is the ability to steal certain players a marketing strategy in OOTP? Maybe, maybe not. Is it a marketing strategy in Madden 2003? I'm pretty sure it is. The way OOTP is with the team market size and such, it leaves around half the teams at a great disadvantage when it comes to getting or keeping good players on their team. Half of those disadvantaged teams could have fans that represent 30% of the buying market, but most of those people are going to want to choose their home team to play with. So if you're a Tampa Bay fan then good luck in ever trying to win the World Series without getting fired in the first year. This is if the game was built with absolute strict trading. However an occasionally loose trading system could be the best way to approach programming a game that you're trying to sell to the public. This loose system could be a bug while the trade AI is being developed but it might not be. Is this some conspiracy theory type of situation? No it isn’t. But since this game is also different in that you can communicate with the developers as it develops then it makes it harder for someone to carry on their vision of how the game should be versus the people on this web board who see things differently.

JAttractive · 04-22-2003, 07:11 AM

Please don't trade the trade interface back to what it was. This is a step forward in my mind.... BUT....

The trade AI does seem weaker in that prospects are not being evaluated high enough. I can get five star prospects for 2 star veterans who don't even have stats to back up the deal. Admittedly this is with scouts on and YES I do think this should happen from time to time when they are used (like it happens in RL when teams scout poorly) but this is just ridiculous now.

Not only are young guys valuable for their current potential but also their potential to get better. I know they can drop as well but a good prospect falling to an average is NOT equal to a good one becoming a superstar. Why you ask? Because players are on a bell curve. Average to Good stars are fairly easy to get. Superstars are not and should be treated like GOLD!

Also extremely important is that for 5 years or so they will have a far lower salary than they probably deserve which should not be overlooked. AI teams need to start evaluating talent far higher than they do. Perhaps when it comes to talent they must consider players 25 and under to already have at least the ratings at the mid range of their talent it in terms of their value. Ex. Guys who are Brill/Brill/Brill should not be tradeable for anyone who is not at LEAST an 8/8/8 themselves. Then age and salary must be considered as well, team needs etc.

RonnieDobbs · 04-22-2003, 08:47 AM

There seems to be a wide range of opinions of what people think a trading interface should provide in terms of ease and frequency of trading. To me, any truly great game is widely customizable, and this is certainly the case with OOTP (anyone reading this should be familiar with the wide array of available customizations).

With this in mind, I think trading should be customizable. I already know that it is at least partially customizable, in that you can set the trade frequency between AI teams. Perhaps this could be extended to the willingness of AI to make trades. Personally, I would like it if I had to fight tooth and nail to get players off the AI..... but at the same time, it is a great point that many people enjoy ripping off the computer (and what isn't enjoyable about it?) The point is, this shouldn't be a one or the other option.

Perhaps the stinginess of the AI GM could be tied to the manager level? Or, is it already? I only play on "Big Leaguer". Feel free to edify me on this point.

Spielman · 04-22-2003, 09:40 AM

Quote:

Originally posted by anpham
[B]if you’re also trying to design a game that you’re hoping will eventually start to stock the shelves at stores then you have to start developing it so that it fits the casual buyers needs, expectations, and level of experience.

For the purists who like to win on their own they can just ignore it and feel better about themselves that they had the chance to cheat and didn’t, and feel better when they win because they didn’t cheat. It works both ways.

if you're a Tampa Bay fan then good luck in ever trying to win the World Series without getting fired in the first year./B]

To respond to those points in order:

The game _was_ on store shelves the last couple of years as Season Ticket Baseball.

The trouble is, this exploit is hard to ignore. I've got no trouble imposing house rules on myself, but any house rules which would counteract this AI would have to be so draconian it'd sap the fun out of the game. God help me, I want to be able to make trades, and to have my goal be to make the best possible deal for my team. Is that so ******* wrong? At this point, the challenge in trading is to play baby-sitter for the computer teams to make sure they don't get ripped off too badly. That's not what I'm looking for, and I'm flabbergasted that so many people seem to have no problem with it, given that previous versions of the game did a better job.

If you're a Tampa Bay fan who wants to cheat, play without manager mode, and edit your players. Manager mode's there as an option to provide a challenge... which it doesn't with the trade AI the way it is.

Wipethatsmirk, nice work there. The trouble isn't just that this is possible, but that it's so easy to do. The new interface makes it too easy to find the best possible deal for the best possible deal to be so very very good all the time.

-Spielman

BleacherBum · 04-22-2003, 04:11 PM

Quote:

At this point, the challenge in trading is to play baby-sitter for the computer teams to make sure they don't get ripped off too badly.

Right on. And for those who think this is acceptable, realize that you don't need any Trading AI logic at all to construct and police your own trades. You could just set teams to human, and architect any trade that you think is fair.

anpham · 04-22-2003, 05:06 PM

If you're someone who is going to edit a player or one who doesn't immediately lock the Ratings System on the first day of your first season then you're a player who loves the fact that you can steal prospects off certain teams. Editing players is way worse in my mind but it's really besides the point. The point I'm trying to get at is picture yourself as the programmer. You can spend countless hours, sleep 4 hours a day, programming. You see how your program has developed gradually and you're trying to find the most optimal way to make something that will satisfy as many people as possible. You're also testing your work as you go along and seeing the results of the code you just wrote. So if you're programming something and you start testing it don't you think he would have noticed that you could occassionally get a 4.5-5 star prospect (even a SP) for a 2 star veteran outfielder? Right when you select a player and select a team to trade with it automatically lists all the players this team would trade. If this was a bug then why would he release it this way because I'm pretty sure he would have noticed it. It's pretty hard to miss. Maybe he released it this way to see how you guys would react to a situation like this. To see whether or not you guys would accept it or react strongly against it. Some people are for it others are against it. It's just one of those things that wouldn't just slip on by and in my mind was a conscience decision to let it happen. No good programmer doesn't just write code indiscriminately without out giving it a thorough run through before releasing it to the public, especially when its something people are paying for. So if you see that trading from OOTP4 to OOTP5 has gotten looser and easier with the new trading screen then that was probably his intention. Because if you're not completely satisfied with something then you don't include it your update or patch. You continue to work on it until it's ready and then you release it. Obviously he felt that it was right in his own view and put it in there. A typo here and there obviously is accidental but with something like the trade AI that could be thousands of lines of code I doubt would be accidental.

Sweed · 04-22-2003, 05:14 PM

Quote:

Originally posted by Spielman

I'm flabbergasted that so many people seem to have no problem with it, given that previous versions of the game did a better job.

-Spielman

Well, when v5 first came out it was somewhat hard to make trades. Some of us posted that we liked it this way. Others said it wasn't right (no trades could be made, to easy, to hard etc.) and needed to be redone and so it was. First patch came out and the posts began immediatly so it was redone again. Problem is each time it gets easier and easier and more and more house restrictions are needed. This leaves some of us that have "no problem" with it to only say "well then don't make the trade". Is there any thing else we can do? A lot of people complained about the first version but at least you had to do some work to get a trade done. I can't wait for the next version maybe I can have the computer autoscan my league and just go ahead and give me all the best players.

MannyTrillo · 04-22-2003, 06:33 PM

Quote:

Originally posted by RonnieDobbs

With this in mind, I think trading should be customizable.

Agree completely, I think this is the only solution. I really enjoyed 5.00, 5.01 was worse, 5.10 is not playable to me as a solo game. On the other hand, a loooot of people were complaining about the trading AI in 5.00, saying it made the game unplayable to them. For more experienced players, the #1 problem with OOTP seems to be that it's just too easy to build a team that wins 100 games every year, even with small market size, house rules, etc. But tighten things up too much and you make it less fun for the less experienced players. A trade AI difficulty level is the only answer that I can see.

As another specific example of silly 5.10 trades, the computer was willing to give me the #1 prospect in the league, 5 blue stars, with 5's and 6's for ratings, for a 3.5 blue star MR.

Spielman · 04-22-2003, 07:24 PM

I was mucking about with trades earlier today, experimenting with different deals.

As Detroit, I sent Todd Hundley to Florida in exchance for Ramon Castro. Hundley's more expensive, and can't make contact for crap. Castro's still too expensive, and hits for better average. Not a crazy trade for the computer to accept.

Then in my experimentation, I wound up trying to trade Castro. In shopping him around, I wound up offering him to Florida. Who do I see in the list of Florida players who would make the deal work? You guessed it... Todd Hundley.

No, Florida hadn't made any moves at all since the original deal. Their situation was exactly the same.

This. Should. Not. Happen.

Period.

Either the AI should see Hundley as more valuable to Florida under the current circumstances, or it should see Castro as more valuable. Having the ability to make the swap either way suggests that either the AI is quite broken, or that the AI isn't requiring that a trade be seen as an increase in value; just that it be sorta kinda close. That just doesn't work in a game of this type.

This appears to be a common occurance. Deals can be reversed without trouble. Players can be traded for multiple players, and one player can be swapped back for the original player, leaving one team clearly ahead on the transaction.

Any apologists have a good response on this one?

-Spielman

mtw · 04-22-2003, 11:34 PM

Quote:

Originally posted by Spielman
Either the AI should see Hundley as more valuable to Florida under the current circumstances, or it should see Castro as more valuable. Having the ability to make the swap either way suggests that either the AI is quite broken, or that the AI isn't requiring that a trade be seen as an increase in value; just that it be sorta kinda close.

-Spielman

As one of the many who bemoaned the difficulty of trading in prior OOTP versions, I'd be happy to recite a litany of trades that the AI should have taken, but didn't. There were also some trades that the AI shouldn't have taken, but would have without "house rules". This can be debated back and forth all day without resolving anything.

Compared to prior versions, I love the AI the way it is now. Is it overly permissive? Sure, but I remain impressed with the improvement of the value the AI places on players, especially the higher rated ones. That said, there are certainly some adjustments which could and should be made, already detailed in other posts, including within this thread, that would IMO address a lot of the problems with "easy trading" without resorting back to the "We will not trade X" AI of versions past. Examples include: giving more value to high-rated prospects, reducing the AI's emphasis on needs and increasing it's emphasis on value received, and firm rules about the value the AI needs to receive for players of higher ratings (i.e. one two-star veteran is not worth one five-star prospect).

Instead of complaining about how easy the trade AI is to take advantage of, how about occasionally posting some specific and constructive ideas of how to make it better? Also, if you're going to post examples of bad trades, how about also posting all relevant details surrounding the trade so that areas for improvement can be pinpointed rather than just maligning with general disapproval. Examples of information to provide would include: What did the AI list as its needs? What was its depth at the positions it was trading? What was the financial situation of the AI team? How old were the players invloved? What were their ratings? For the purposes of the above quoted example, how much time passed between the first trade and the later attempts to trade? With that additional information, perhaps the cause of the problem can be identified, and legitimate and considered suggestions constructed for Markus to implement, if he so chooses.

Not specifically picking on Spielman, just using this quote as a typical example of posts I've seen so far complaining about the AI, I don't see any of this additional information provided--only the names of two players and some general statements that the trade AI is flawed.

IMO the goal should be to get the Trade AI (and perhaps the entire game), as close to real life as possible and to the satisfaction of all users, not mindlessly reverting to an AI that by my recollection generated more dissatisfaction than the current one.

Maple Leafs · 04-23-2003, 09:24 AM

Excellent thread. Here's a few comments, many of which fall into the "me-too" category since they've already been raised.

1. OOTP desperately needs for teams to recognize whether they're in the race or not. Things like long-term strategies and GM personalities would be nice, but they're not must-haves. However, teams that are contending should not be trading stars for prospects, period. This is the #1 problem with the trade AI right now.

OOTP clearly has some mechanism in place to determine if a team is a contender, because it affects fan interest. Maybe this could be factored into the trade AI?

2. Like most of you, I'm seeing the "five-star prospect for one-star scrub" deals on the table. My hunch is that this is not new to OOTP 5.1, but rather that we're only noticing it now that we can see every player the AI would consider taking in a deal. It seems as if the AI algorithm has some sort of built-in "margin of error" to allow for the occasional lopsided deal, and this is showing up every now and then.

3. Not really an AI thing, but trade-related: please let me choose my own "weaknesses" if it's going to determine what players I'm offered. I'm tired of getting a half dozen offers involving 2B because OOTP decides that's where I'm weak. I like my 2B. I think I'm weak a 3B, but OOTP doesn't so I'll never get an offer.

4. As several people have pointed out, there needs to be a "submit offer" button on the trade screen, and the human GM should be held to his offer (no taking back players and trying again).

5. In trades, the AI must be able to consider a player's secondary positions. A team that needs a closer should know that it can use a five-star MR. If a team needs a RF and I'm offering a guy who can play all three OF spots, I shouldn't have to go in and change his position to make the AI "see" that he would fit their needs.

RonnieDobbs · 04-23-2003, 10:14 AM

A point raised most recently by Maple Leaf, but also raised numerous other times, that the AI must recognize where it is in the standings when trading, is very true. Here is a quick example....

Using the 2003 rosters/season, I am managing the Red Sox. My record is something like 72-28. Looking at the standings, and using commons sense, the Yankees are my likely opponent in the ALCS. With this in mind, I went shopping for a closer near the trade deadline, with lots of money and prospects to spare. I immediately targeted Mariano Rivera, and coaxed the Yankees into giving him up for a decent, but not great, three star prospect RF, who might be of some use in a couple years.

Now, not only do I have a very good closer, but my primary threat to AL supremacy, the Yankees, must try to beat me with Steve Karsay closing out games, a double advantage. I basically ripped off their arm, and am now planning to beat them over the head with it.

I think it goes without saying..... if the Yankees dealt Rivera to the Sox in the middle of a heated race, New Yorkers would probably tear down Yankee Stadium.

Do I expect the AI to be so subtle as to recognize the rivalry between two cities, the history of Sox-Yankees trades (ie Mr. Ruth), etc.? No. But, should the AI be able to recogize that the Yankees have the second best record in baseball, and thus shouldn't be trading ANY four or five star player for ANY prospect to ANY team, no matter how enticing. It would be nice.

Just a suggestion.

mtw · 04-23-2003, 02:06 PM

Quote:

Originally posted by Spielman
The only relevant point would be whether game time went by before the trade back. I thought that the answer to that would be evident from context. Evidently not. No, no game time went by between the original trade and the trade back.

Thank you for now providing a crticial detail which helps evaluate the trade you mentioned. No game time passing is certainly relevant and a simple statement to that effect would have clarified most of the other information that I suggested providing. From the way your original post was worded, it seemed to me like some game time had passed, so I did not make an assumption to the contrary.

What I will assume is that catcher was a need for Florida. If that is true, then my response is that having also seen similar scenarios, the AI needs to focus more on "value added" and less on "need", which I and others have said already. If Todd Hundley has Value X to the AI and Ramon Castro Value Y, then the AI should prefer one player to the other. If both have equal value, then the AI should not be willing to give up any additional players to get one or the other. Player valuation is an another area where several people have posted probably needs some adjustments.

Quote:

I've posted suggestions earlier in the thread. I don't feel the need to reiterate them every time I post something slightly different in order to protect myself from the criticism of an over-zealous apologist. Apparently I'm wrong there too.

Paraphrasing, you've said within this thread that 1) prospects are undervalued and 2) the AI should deal more critically with users who propose silly trades, both statements with which I agree. Good suggestions.

I find it very amusing that after posting on numerous and varied aspects of OOTP in an effort to get changes made, including the current trade AI, I am now an Over-Zealous Apologist. I like it, and will now use that as my handle for the forseeable future.

Edit - SV - Removed the non-Trade AI related stuff.

ghulten · 04-23-2003, 02:16 PM

Quote:

Originally posted by Maple Leafs
Excellent thread. Here's a few comments, many of which fall into the "me-too" category since they've already been raised.

1. OOTP desperately needs for teams to recognize whether they're in the race or not. Things like long-term strategies and GM personalities would be nice, but they're not must-haves. However, teams that are contending should not be trading stars for prospects, period. This is the #1 problem with the trade AI right now.

OOTP clearly has some mechanism in place to determine if a team is a contender, because it affects fan interest. Maybe this could be factored into the trade AI?

2. Like most of you, I'm seeing the "five-star prospect for one-star scrub" deals on the table. My hunch is that this is not new to OOTP 5.1, but rather that we're only noticing it now that we can see every player the AI would consider taking in a deal. It seems as if the AI algorithm has some sort of built-in "margin of error" to allow for the occasional lopsided deal, and this is showing up every now and then.

Two quick reponses:

1. Please think this through all the way. How is the AI going to evaluate it's current situation? Is the AI going to be good enough to recognize whether it's a contender or a rebuilder? In real life, many team fail to recognize where they are in the "success cycle" (think NY Mets) and insist on acting like contenders when they should be rebuilding. This is being discussed in a little more detail here but needs to be remembered. REAL LIFE GMS SCREW UP all the time while evaluating thier own teams and their chances of contending. I don't know if it's realistic to expect Markus to program an AI that thinks more intelligently than Steve Phillips or Dave Littlefield do (or maybe not ;-)
EDIT: When these AI GMs screw up their self-evaluation and make trades that are inappropriate for their team's current position, arent' we right back where we started? With teams making trades that make no sense for them and their situation? Such mistakes are realistic, so how do we address that?

2. I wish people would stop talking about stars. The star rating assigned by the game are wrong all the time. There can be so much difference between two three-star players. Please remember this when evaluating trades.

Thanks,
Gordy

Spielman · 04-23-2003, 03:11 PM

"What I will assume is that catcher was a need for Florida. If that is true, then my response is that having also seen similar scenarios, the AI needs to focus more on "value added" and less on "need", which I and others have said already."

No, that'd be a faulty assumption. Florida has Pudge Rodriguez; catcher was not considered a need position by Florida before or after the first trade. In fact, Hundley wasn't assigned to the major league roster at all, but was sent to AAA.

Edit - SV - Removed Non-Trade AI stuff (Take the personal disagreements elsewhere please!)

04-21-2003, 12:35 PM	#41
RonnieDobbs Bat Boy Join Date: Apr 2003 Posts: 10	My suggestions, put as briefly as possible...... (if I mention something already posted, I apologize, but I am simply reiterating the previous point). 1- AI gives up valuable closers too easily, to the point that I often "raid" opposing bullpens of teams I expect to meet in the playoffs. Yes, relievers are not as valuable as starters, but I shouldn't get Billy Wagner and Octavio Dotel for a three star prospect and an extra OF. 2-AI also greatly undervalues very good young starters. I shouldn't get Mark Prior for Cliff Lee and a low level prospect, Josh Beckett for Adam Dunn. Top-shelf young starters should cost the house. 3-I strongly agree with the posts that suggest "punishment" by rival GMs for pulling out of an offer. In real life, I would have gotten a black eye by now for all of the dickering I have inflicted on rival GMs. 4-A "shop player" option should be included, but there should be a resulting penalty if you fail to trade the player after "shopping" him. Just imagine how disgruntled a normal MLBer would get after hearing that he was being shopped to any and all takers. __________________ "Ya'll been brutalizin' me....."

04-21-2003, 04:47 PM	#44
marc420 Minors (Double A) Join Date: Oct 2002 Location: Terrapin Station Posts: 112	suggested reading Two columns from ESPN's website from Trade Deadline Day last season. I know I'm just mouthing off about the way I think things should be. Maybe I get lucky and hit a good idea once in a while. But I'd suggest that people read these two columns, then ask themselves if in playing OOTP against the AI, they feel like they were in anything like what was described in these two columns. Peter Gammons, Aug 3, 2002 "Going inside on Deadline Day" http://espn.go.com/gammons/s/2002/0803/1413691.html Jayson Stark, July 31, 2002 "Deadline Day goes quietly, but dealing not likely done" http://espn.go.com/mlb/columns/stark...n/1412625.html Some quotes from each Stark ---- A few weeks ago, the Cardinals didn't seem to have much to deal. But when they were through, they'd won the Scott Rolen Sweepstakes for a very modest price .... General manager [CLEVELAND] Mark Shapiro knew early he'd be trading a bunch of veteran players. So he did a study of all the deadline deals of the last 10 years -- and concluded the best trades were usually made for prospects who were years away. .... But fellow GMs complain that [TAMPA BAY]Chuck LaMar slapped Mercedes price tags on the Hyundais on his lot, and he aggravated many potential customers. We asked about Esteban Yan, and he told us it would take two top-quality prospects, and a fringe major leaguer," grumbled one team. And we heard similar talk all over both leagues. As for LaMar's most marketable player, Randy Winn, he's so young and cheap, we probably would have driven a hard bargain for him, too. One disgruntled GM said: "What's he need him for? So he can stay there and help them win 50 games?" .... But two major-league sources say [KANSAS CITY]Baird went down to the wire talking about a potential three-team deal that apparently would have involved the Mets and Blue Jays. The Mets would have gotten Byrd. It appears the Royals would have wound up with Toronto's big-time second-base prospect, Orlando Hudson. The sticking point may have been the Mets' reluctance to part with their best pitching prospect, Aaron Heilman. .... Always-active Mets GM Steve Phillips had potential deals flying all over the map. Another trade reportedly would have sent relievers David Weathers and Mark Guthrie to Anaheim for Schoeneweis. .... Names being shopped on Deadline Day In the hours before the deadline, these were some of the names that reportedly were being dangled actively but never went anywhere: Tom Gordon, Jeff Fassero, Doug Glanville, Delino DeShields, Roberto Hernandez, Tanyan Sturtze, Bubba Trammell, Lenny Harris, Matt Stairs, Brandon Larson, Neifi Perez, Mark Redman. Gammons --- The Yankees, you see, could have had Floyd, but Steinbenner so overreacted to Enrique Wilson's mishaps in right field in a game against the Mets on June 29 that he overruled the sensibilities of "his baseball people," called Toronto owner Paul Godfrey and then bought Raul Mondesi on July 2. If Steinbrenner had waited an hour, he'd have found out that as the Mondesi deal was being completed, Marlins team president David Samson had Floyd's agent Seth Levinson in his office informing him that that they could not re-sign Floyd and he would be put on the market that afternoon. Then earlier this week, Steinbrenner found out the Red Sox were the leaders in pursuit of Floyd and he tried to jump in to block the deal. "It got down to the Yankees and Boston," says Expos GM Omar Minaya. Montreal asked the Yankees for outfielder Juan Rivera and first baseman Nick Johnson. "They chose not to do that," says Minaya, "and we went with Boston." Oakland had the best players to offer for Floyd, but there were two roadblocks: Floyd has a no-trade clause to Oakland he wouldn't waive, and the A's also needed the Expos to take back the final $2 million in the deal; Minaya wanted the $2 million to use to acquire a closer in case, as he puts it, "we get back into the race." Oakland then tried to get Atlanta interested in taking Floyd at the end of a three-way deal, but that didn't work, leaving only the Red Sox and Mets. And Minaya did not want to improve the team with whom he is competing in the wild card and the NL East. .... "Even as a seller, it's more complex than many realize," says one NL GM. "It's understandable that some teams wanted to wait, but obviously the returns diminished the closer we got to July 31. It's all about buyers who could take on contracts. It'll be the same thing in August, with claimings." .... Most GMs believe there is a major market correction working, with very few teams willing to go spend on free agents this winter. If the Indians, Cardinals and Red Sox tried to sign Thome, Rolen or Floyd right now, they would have to do so at 2001 prices. Thus they have to wait and let the market dictate their value, which in each case may be far less than they expect, despite each player's ability and character. "A classic example is Johnny Damon," says one GM. "He essentially signed with Boston for what he turned down in Kansas City. It wasn't his fault. The market shifted downward."

04-22-2003, 08:47 AM	#48
RonnieDobbs Bat Boy Join Date: Apr 2003 Posts: 10	There seems to be a wide range of opinions of what people think a trading interface should provide in terms of ease and frequency of trading. To me, any truly great game is widely customizable, and this is certainly the case with OOTP (anyone reading this should be familiar with the wide array of available customizations). With this in mind, I think trading should be customizable. I already know that it is at least partially customizable, in that you can set the trade frequency between AI teams. Perhaps this could be extended to the willingness of AI to make trades. Personally, I would like it if I had to fight tooth and nail to get players off the AI..... but at the same time, it is a great point that many people enjoy ripping off the computer (and what isn't enjoyable about it?) The point is, this shouldn't be a one or the other option. Perhaps the stinginess of the AI GM could be tied to the manager level? Or, is it already? I only play on "Big Leaguer". Feel free to edify me on this point. __________________ "Ya'll been brutalizin' me....."

04-22-2003, 07:24 PM	#54
Spielman All Star Starter Join Date: Feb 2002 Location: South Carolina Posts: 1,668	I was mucking about with trades earlier today, experimenting with different deals. As Detroit, I sent Todd Hundley to Florida in exchance for Ramon Castro. Hundley's more expensive, and can't make contact for crap. Castro's still too expensive, and hits for better average. Not a crazy trade for the computer to accept. Then in my experimentation, I wound up trying to trade Castro. In shopping him around, I wound up offering him to Florida. Who do I see in the list of Florida players who would make the deal work? You guessed it... Todd Hundley. No, Florida hadn't made any moves at all since the original deal. Their situation was exactly the same. This. Should. Not. Happen. Period. Either the AI should see Hundley as more valuable to Florida under the current circumstances, or it should see Castro as more valuable. Having the ability to make the swap either way suggests that either the AI is quite broken, or that the AI isn't requiring that a trade be seen as an increase in value; just that it be sorta kinda close. That just doesn't work in a game of this type. This appears to be a common occurance. Deals can be reversed without trouble. Players can be traded for multiple players, and one player can be swapped back for the original player, leaving one team clearly ahead on the transaction. Any apologists have a good response on this one? -Spielman __________________ Spielman was at one time the smartest person on these boards. http://www.ootpdevelopments.com/boar...martest+Person I don't believe in AnotherAlias. Last edited by Spielman; 04-22-2003 at 07:37 PM.

04-23-2003, 10:14 AM	#57
RonnieDobbs Bat Boy Join Date: Apr 2003 Posts: 10	A point raised most recently by Maple Leaf, but also raised numerous other times, that the AI must recognize where it is in the standings when trading, is very true. Here is a quick example.... Using the 2003 rosters/season, I am managing the Red Sox. My record is something like 72-28. Looking at the standings, and using commons sense, the Yankees are my likely opponent in the ALCS. With this in mind, I went shopping for a closer near the trade deadline, with lots of money and prospects to spare. I immediately targeted Mariano Rivera, and coaxed the Yankees into giving him up for a decent, but not great, three star prospect RF, who might be of some use in a couple years. Now, not only do I have a very good closer, but my primary threat to AL supremacy, the Yankees, must try to beat me with Steve Karsay closing out games, a double advantage. I basically ripped off their arm, and am now planning to beat them over the head with it. I think it goes without saying..... if the Yankees dealt Rivera to the Sox in the middle of a heated race, New Yorkers would probably tear down Yankee Stadium. Do I expect the AI to be so subtle as to recognize the rivalry between two cities, the history of Sox-Yankees trades (ie Mr. Ruth), etc.? No. But, should the AI be able to recogize that the Yankees have the second best record in baseball, and thus shouldn't be trading ANY four or five star player for ANY prospect to ANY team, no matter how enticing. It would be nice. Just a suggestion. __________________ "Ya'll been brutalizin' me....."

04-21-2003, 03:43 PM	#42
marc420 Minors (Double A) Join Date: Oct 2002 Location: Terrapin Station Posts: 112	First off, let me say I currently have OOTP4, and have been holding off on OOTP5. The reason I'm reading this thread is to try to find what improvements have been made, and I'm mainly concerned with areas like Free Agents and Trades. To me, this new Trade interface sounds wonderful, and I think Markus has pretty much convinced me to go ahead and get OOTP5 by putting this in. THANKS! As to whether other GMs get upset by backing out of a deal, it really depends on how good the interface of the game is. In OOTP4, the only way to find out what sort of trade another team would take is to propose it. If the game forces you to do that to find out what value a team places on a player, then it shouldn't penalize you for doing this. And in general, I do believe you should be able to get to a point where you tentatively agree to a team. You'd basically say something like "That's very interesting, let me think about it and talk it over on this end and we'll get back to you." If the other team things its a good deal, they'd be willing to wait a few days. Or maybe you'd get a message back the next day that says, "California's also made us an attractive offer, get back to us today or we'll make the deal with California". If I was a GM thinking about trading Sammy Sosa, I wouldn't do it on the spur of the moment during a phone call. I'd work out the parameters of the deal. But then I'm going to go talk to my manager and see how he likes the deal. I'm going to go talk with the Captain of my team and some other key players on the team to see what they think of the deal. I'm probably having meetings with my scouts talking about the deal. Just as pure CYA, I'm going to go talk to my owner about the deal and see what he thinks about trading away Sammy Sosa. In general, it seems very reasonable that after negotiating the parameters of a deal with another team, a GM and a club would stop and think about it and then say Go or No Go. And I can't see another GM getting too upset about that. Now maybe if I'd worked up 15 trades this week with a GM, then said No to everyone, then I'd expect to get a "stop wasting my time" response back. But it should take something pretty severe to get there. And besides, there's only 30 GM's to deal with. Even if one GM doesn't like another one, there's a very limited group of people you have to deal with. As to whether there's a penalty for "shopping" a player, that would seem to depend on how much the player wants to stay with a team. If the team is 40 games out in August, and there's rumours that the player might get traded to the Yankees. Then if the deal falls through, I can see the player saying "what the heck" and tanking it for the rest the season. But if the player is on the Yankees, and the rumour is he's about to get traded to the team 40 games out to get another pitcher, then if the deal falls through, I can see the player being elated that he didn't get traded. For some players in some cases, this could serve as an incentive to play harder. Age of the player could make a difference too. For a young player, I can see this upsetting them. For a veteran player, they may just shrug it off. I can see them saying "Baseball's a business, this happens all the time", and then start telling stories about the time he was in Cleveland and there was a rumour .....

04-22-2003, 12:39 AM	#46
anpham Bat Boy Join Date: Apr 2003 Posts: 10	Most people who post on this web board you can probably classify as die-hards who know baseball pretty well and want an as true to life simulation as possible. But it’s possible that the majority of people who buy this game and who don’t post on this web board may just be the occasional baseball fan who read some reviews and decided he/she wanted to try managing a team and this one was the one that got the best reviews. I looked around the Internet for baseball games and this one got good reviews on most of them and consistently. However the games this one was competing against were very different in that you used a gamepad to play those games whereas this one is more of a drag-and-drop, managerial, GM experience. I played Madden NFL 2003 and found the franchise portion of the game more enjoyable than the playing part so it was an attractive buy for me. But if I’m searching for baseball games on GameSpot or something I would kind of expect it to be the typical PS2 type of game but found that is nothing like it. However the reason it does get good reviews generally is because it isn’t your typical PS2 game and is pretty good at not being one. I sat around and thought about which game to buy whether it was High Heat 2004, MVP Baseball 2003, or OOTP5. High Heat got a 6.4, MVP 8.0, and OOTP 8.5. They are all similarly priced. I’m familiar with the first two games but never heard of OOTP until I read about it on GameSpot because you can’t find it in stores or at Blockbuster yet. Since it’s also a game that is sold over the Net then obviously the reviews of it on the Net are going to be pretty important. Of course I looked at some screenshots and tried to see what the game offered me but I did lean towards MVP at first because it was produced by a trusted name. I do think this game is trying to be different and gear itself towards a certain kind of buyer that wouldn’t buy the High Heat’s and MVP’s, but I also think they would want to avoid as many refunds as possible for people like me who read reviews and saw this one got the best and decided to buy it based on the reviewer score. Then if you’re also trying to design a game that you’re hoping will eventually start to stock the shelves at stores then you have to start developing it so that it fits the casual buyers needs, expectations, and level of experience. I'm sure most casual users will prefer it be the way it is where they can occasionally get the players they want on their team for fairly little. If you start off as a Tampa Bay fan then your team starts off with almost nothing and there’s fairly little you can do transaction wise. However with a hole or two you could possibly trade for some talent and it'll at least make it possible for you to win using your favorite team. Although I'm not sure if this is intentional or not I've seen a game where they did make it intentional. Madden NFL 2003 is the only team management game I've ever played before this one, but in that game there was a "bug" where you could trade your first round draft pick for any teams 1, 2, and 3 picks. Even if you won the Super Bowl the previous year and the team you're trading with finished last (and the game does factor in your team records later on) you could still pull off this trade during the first 10 or so years of your franchise. After that they stop letting you trade your first pick away for any teams first 3. This bug would be pretty easy for anybody to find and it gives the user a chance to get ahead and build a better team. It's one of those things that wouldn't be missed during bug testing and doesn't make sense when you see how the computer trades with you in the later years. This helps a person build a team they could win the Super Bowl with and obviously winning makes any game more enjoyable. No one likes to lose continually so in my mind it was a marketing strategy that the Madden people decided to let you do for the first 10 years so you could build your team and be competitive. Although the percentage of people who are going to play like this could be fairly low or high, if this “bug” made 10% of the buyers happier to have it available then why not build it into the game. For the purists who like to win on their own they can just ignore it and feel better about themselves that they had the chance to cheat and didn’t, and feel better when they win because they didn’t cheat. It works both ways. Is the ability to steal certain players a marketing strategy in OOTP? Maybe, maybe not. Is it a marketing strategy in Madden 2003? I'm pretty sure it is. The way OOTP is with the team market size and such, it leaves around half the teams at a great disadvantage when it comes to getting or keeping good players on their team. Half of those disadvantaged teams could have fans that represent 30% of the buying market, but most of those people are going to want to choose their home team to play with. So if you're a Tampa Bay fan then good luck in ever trying to win the World Series without getting fired in the first year. This is if the game was built with absolute strict trading. However an occasionally loose trading system could be the best way to approach programming a game that you're trying to sell to the public. This loose system could be a bug while the trade AI is being developed but it might not be. Is this some conspiracy theory type of situation? No it isn’t. But since this game is also different in that you can communicate with the developers as it develops then it makes it harder for someone to carry on their vision of how the game should be versus the people on this web board who see things differently.

04-22-2003, 07:11 AM	#47
JAttractive All Star Starter Join Date: May 2002 Location: St. Catharines, Ontario Posts: 1,140	Please don't trade the trade interface back to what it was. This is a step forward in my mind.... BUT.... The trade AI does seem weaker in that prospects are not being evaluated high enough. I can get five star prospects for 2 star veterans who don't even have stats to back up the deal. Admittedly this is with scouts on and YES I do think this should happen from time to time when they are used (like it happens in RL when teams scout poorly) but this is just ridiculous now. Not only are young guys valuable for their current potential but also their potential to get better. I know they can drop as well but a good prospect falling to an average is NOT equal to a good one becoming a superstar. Why you ask? Because players are on a bell curve. Average to Good stars are fairly easy to get. Superstars are not and should be treated like GOLD! Also extremely important is that for 5 years or so they will have a far lower salary than they probably deserve which should not be overlooked. AI teams need to start evaluating talent far higher than they do. Perhaps when it comes to talent they must consider players 25 and under to already have at least the ratings at the mid range of their talent it in terms of their value. Ex. Guys who are Brill/Brill/Brill should not be tradeable for anyone who is not at LEAST an 8/8/8 themselves. Then age and salary must be considered as well, team needs etc.

04-22-2003, 05:06 PM	#51
anpham Bat Boy Join Date: Apr 2003 Posts: 10	If you're someone who is going to edit a player or one who doesn't immediately lock the Ratings System on the first day of your first season then you're a player who loves the fact that you can steal prospects off certain teams. Editing players is way worse in my mind but it's really besides the point. The point I'm trying to get at is picture yourself as the programmer. You can spend countless hours, sleep 4 hours a day, programming. You see how your program has developed gradually and you're trying to find the most optimal way to make something that will satisfy as many people as possible. You're also testing your work as you go along and seeing the results of the code you just wrote. So if you're programming something and you start testing it don't you think he would have noticed that you could occassionally get a 4.5-5 star prospect (even a SP) for a 2 star veteran outfielder? Right when you select a player and select a team to trade with it automatically lists all the players this team would trade. If this was a bug then why would he release it this way because I'm pretty sure he would have noticed it. It's pretty hard to miss. Maybe he released it this way to see how you guys would react to a situation like this. To see whether or not you guys would accept it or react strongly against it. Some people are for it others are against it. It's just one of those things that wouldn't just slip on by and in my mind was a conscience decision to let it happen. No good programmer doesn't just write code indiscriminately without out giving it a thorough run through before releasing it to the public, especially when its something people are paying for. So if you see that trading from OOTP4 to OOTP5 has gotten looser and easier with the new trading screen then that was probably his intention. Because if you're not completely satisfied with something then you don't include it your update or patch. You continue to work on it until it's ready and then you release it. Obviously he felt that it was right in his own view and put it in there. A typo here and there obviously is accidental but with something like the trade AI that could be thousands of lines of code I doubt would be accidental.

04-23-2003, 09:24 AM	#56
Maple Leafs Minors (Triple A) Join Date: Mar 2002 Posts: 234	Excellent thread. Here's a few comments, many of which fall into the "me-too" category since they've already been raised. 1. OOTP desperately needs for teams to recognize whether they're in the race or not. Things like long-term strategies and GM personalities would be nice, but they're not must-haves. However, teams that are contending should not be trading stars for prospects, period. This is the #1 problem with the trade AI right now. OOTP clearly has some mechanism in place to determine if a team is a contender, because it affects fan interest. Maybe this could be factored into the trade AI? 2. Like most of you, I'm seeing the "five-star prospect for one-star scrub" deals on the table. My hunch is that this is not new to OOTP 5.1, but rather that we're only noticing it now that we can see every player the AI would consider taking in a deal. It seems as if the AI algorithm has some sort of built-in "margin of error" to allow for the occasional lopsided deal, and this is showing up every now and then. 3. Not really an AI thing, but trade-related: please let me choose my own "weaknesses" if it's going to determine what players I'm offered. I'm tired of getting a half dozen offers involving 2B because OOTP decides that's where I'm weak. I like my 2B. I think I'm weak a 3B, but OOTP doesn't so I'll never get an offer. 4. As several people have pointed out, there needs to be a "submit offer" button on the trade screen, and the human GM should be held to his offer (no taking back players and trying again). 5. In trades, the AI must be able to consider a player's secondary positions. A team that needs a closer should know that it can use a five-star MR. If a team needs a RF and I'm offering a guy who can play all three OF spots, I shouldn't have to go in and change his position to make the AI "see" that he would fit their needs.

04-23-2003, 03:11 PM	#60
Spielman All Star Starter Join Date: Feb 2002 Location: South Carolina Posts: 1,668	"What I will assume is that catcher was a need for Florida. If that is true, then my response is that having also seen similar scenarios, the AI needs to focus more on "value added" and less on "need", which I and others have said already." No, that'd be a faulty assumption. Florida has Pudge Rodriguez; catcher was not considered a need position by Florida before or after the first trade. In fact, Hundley wasn't assigned to the major league roster at all, but was sent to AAA. Edit - SV - Removed Non-Trade AI stuff (Take the personal disagreements elsewhere please!)