What does the AI value in trades?

majesty95 · 02-14-2018, 10:00 AM

I like to go through historical replays and try to maximize the good years of my players and get rid of them right before they have a dip in success. I recently tried to shop Tim Lollar and got no interest from any CPU teams. (I'm playing 16 right now but have seen this on 18). I have my AI eval set up to 10% ratings then 60-20-10 stats. His past two years have been great for the most part. I would think someone would want him. It's almost like the AI sees the ratings and favors them more or knows he's likely to regress although the stats do not say that. I have scouting off and trading on normal FWIW. What contributes to this? I've always been perturbed that the AI always seems to be able to see ratings and projections no matter how you set up the eval.

NoOne · 02-14-2018, 11:29 AM

how do you feel about a 27-28 y.o. with a signifcant portion of potential missing form his movement?

a 5'6" 120lbs guy hits 40hr in year one... do you think he'll continue that? with only 10% based on ratings, maybe that would be the deduction... and it would be a terrible one. some exceptions most likely exist and are virtually required for good decisions. the ai eval is a broad overarching concept that does have a major impact, nonetheless.

if his lack of reaching potential has remained consistent the last couple years (as opposed to a dip), that may not be the problem, but should be a bad omen nonetheless. if it dipped, that's a major problem... it shows a downtrend at ~27, which is going to diminish value.

even in RL when a pitcher overperforms, they don't automatically assume he's the greatest thing since sliced bread. this guy most certainly cannot maintain ~hoF quality pitching of his first 2 years (era+ ~125 is a good bet for HoF if it's scaling to the league stats properly.)

even with 10% ratings, a guy who has a career years) his first couple in the league shouldn't be an equivalent trade asset as an established player with those numbers.

can verify the ops+ potential issue by comparing to league averages the last coule years for pitching.. is he really ~20-40% better in those years? for modern stats sure looks like it, but if league era is signifcantly different than ~4.10, not so much in that environment. i'm not 100% it scales as it should to environement in ootp.

he doesn't look that sexy to me, but i am not familiar with the ratings ootp gives players of that time period. that guy may very well be a gem, and if that's the case you have a good argument.

otherwise, he looks 'replaceable' to me... not a replacement player, just numerous options available of ~similar ability. not something to give up assets in a trade when you can easily find similar quality elsewhere. so, i wouldn't value that very highly, and neither should the AI. also, he never met his potential which does not bode well.

odds, not gaurantees... if they aren't reaching potentialing 24/25 you should start looking for trade partners earlier in career and that will help. exceptions do exists, so if you can stash 1 or 2 in your 6th/7th slots in pen, no big deal... ie SP with a thirde pitch that is 1/XX scale may skyrocket, eventually (lower % as they age). Same with that movement of lollar. (historical play a bit different, but still some randomness even with strictest settings about player ratings). at 27-28 it's not going to happen and the AI knows this too in some shape or form.

one thing about 60-20-10. the 60% will be re-distributed early on in a new season. sample size is considered. in this case, if it was the offseason, it's not pertinent to Lollar. but, if it was eary on in the "3rd" season of the window, it's not weighted so heavily. if it re-distrubutes evenly, it will have a major impact on behaviour being different at various times of the calendar, due to this fact... ie early on it may be 0-40-30, that's nothing like 60-20-10. then, by end of season it's back to 60-20-10. (gist) -- what is actually done is not cetain, but what is certain is that the DO re-distribute the current year % due to sample size...

majesty95 · 02-14-2018, 03:37 PM

I wasn’t expecting an equivalent major leaguer for him really. Just thought someone would be willing to give something. However, with 16 there were only players who debuted in the current years vs full minors so that may have bring more offers. Although I would have thought at least someone would be willing to give up a mediocre relief prospect or bench player lol

NoOne · 02-14-2018, 11:45 PM

you may need to dig a bit. don't rely on "shop player" or just picking out 1 trade partner based on prospect, unless you are going to pay for it no matter the cost. if on the other hand you are looking for greatest rate of return, you have to seek highest demand. that can be a matrix of positional demands depending on complexities of the trade.

find the teams with the red dot for that position in the mlb (prospects when trading prospects - it's about what they get). at first just start plopping guys in by potential to gauge interest from each team (obviously age, position etc play a role in value in addition to potential) ...if a couple of those teams with red dots have more positive trade feedback for distinguishably better players in return (add their prospects of various ability to resolve demand), start with those teams, otherwise, check out all the matching "red dots" have to offer.

can always kick the can down the road by getting prospects.. lots of time to move them or use them etc.

find the highest demand, hope they have some prospects or an mlb player of value. even if you have to package stuff together. not every good player you draft will fit into your plans. this is the opportunity to get somethign for them too.

heck, maybe even a bench player you have a replacement for in AAA. cut payroll at minimum loss of ability, if any, and help a trade. 2 birds, 1 stone.

1:1 trades don't usually work out well for a human player.

majesty95 · 02-15-2018, 05:20 AM

I can try that and will. I guess I thought of the shop player feature as an easier and quicker way to do that. In my mind (and for time saving purposes) shop player should do that already. Maybe not all possible trades show up (especially with scouting on) so the astute owner can still find better deals. But, for the user trying not to spend hours of his available play time searching for one trade for a mediocre player, I would hope shop payer could help out with that. Or maybe ask the assistant GM to find a trade. Would just like a more simple option to unload players.

NoOne · 02-15-2018, 11:34 AM

it can hasten the process a bit... you can scan through a shop player result for best return... then dig deeper on a trade screen with that particular team.

if you keep a incrementally maintained list of target mil prospects, you can also start there. sometimes you just can't afford the players on that list at the time of various decisions on players. hard to know where oyu are at, so tailor any advice to common sense context of your situation.

go for big name target, go for most return for future trade, etc etc. i have a pretty set way of thinking through things, so i get a bit tunnel vision when i suggest stuff.

02-14-2018, 10:00 AM	#1
majesty95 All Star Starter Join Date: Apr 2005 Location: San Antonio, TX Posts: 1,789	What does the AI value in trades? I like to go through historical replays and try to maximize the good years of my players and get rid of them right before they have a dip in success. I recently tried to shop Tim Lollar and got no interest from any CPU teams. (I'm playing 16 right now but have seen this on 18). I have my AI eval set up to 10% ratings then 60-20-10 stats. His past two years have been great for the most part. I would think someone would want him. It's almost like the AI sees the ratings and favors them more or knows he's likely to regress although the stats do not say that. I have scouting off and trading on normal FWIW. What contributes to this? I've always been perturbed that the AI always seems to be able to see ratings and projections no matter how you set up the eval. Attached Images __________________ College Football Sim League

02-14-2018, 11:29 AM	#2
NoOne Banned Join Date: Apr 2015 Posts: 7,273 Infractions: 1/2 (4)	how do you feel about a 27-28 y.o. with a signifcant portion of potential missing form his movement? a 5'6" 120lbs guy hits 40hr in year one... do you think he'll continue that? with only 10% based on ratings, maybe that would be the deduction... and it would be a terrible one. some exceptions most likely exist and are virtually required for good decisions. the ai eval is a broad overarching concept that does have a major impact, nonetheless. if his lack of reaching potential has remained consistent the last couple years (as opposed to a dip), that may not be the problem, but should be a bad omen nonetheless. if it dipped, that's a major problem... it shows a downtrend at ~27, which is going to diminish value. even in RL when a pitcher overperforms, they don't automatically assume he's the greatest thing since sliced bread. this guy most certainly cannot maintain ~hoF quality pitching of his first 2 years (era+ ~125 is a good bet for HoF if it's scaling to the league stats properly.) even with 10% ratings, a guy who has a career years) his first couple in the league shouldn't be an equivalent trade asset as an established player with those numbers. can verify the ops+ potential issue by comparing to league averages the last coule years for pitching.. is he really ~20-40% better in those years? for modern stats sure looks like it, but if league era is signifcantly different than ~4.10, not so much in that environment. i'm not 100% it scales as it should to environement in ootp. he doesn't look that sexy to me, but i am not familiar with the ratings ootp gives players of that time period. that guy may very well be a gem, and if that's the case you have a good argument. otherwise, he looks 'replaceable' to me... not a replacement player, just numerous options available of ~similar ability. not something to give up assets in a trade when you can easily find similar quality elsewhere. so, i wouldn't value that very highly, and neither should the AI. also, he never met his potential which does not bode well. odds, not gaurantees... if they aren't reaching potentialing 24/25 you should start looking for trade partners earlier in career and that will help. exceptions do exists, so if you can stash 1 or 2 in your 6th/7th slots in pen, no big deal... ie SP with a thirde pitch that is 1/XX scale may skyrocket, eventually (lower % as they age). Same with that movement of lollar. (historical play a bit different, but still some randomness even with strictest settings about player ratings). at 27-28 it's not going to happen and the AI knows this too in some shape or form. one thing about 60-20-10. the 60% will be re-distributed early on in a new season. sample size is considered. in this case, if it was the offseason, it's not pertinent to Lollar. but, if it was eary on in the "3rd" season of the window, it's not weighted so heavily. if it re-distrubutes evenly, it will have a major impact on behaviour being different at various times of the calendar, due to this fact... ie early on it may be 0-40-30, that's nothing like 60-20-10. then, by end of season it's back to 60-20-10. (gist) -- what is actually done is not cetain, but what is certain is that the DO re-distribute the current year % due to sample size... Last edited by NoOne; 02-14-2018 at 11:48 AM.

02-14-2018, 03:37 PM	#3
majesty95 All Star Starter Join Date: Apr 2005 Location: San Antonio, TX Posts: 1,789	I wasn’t expecting an equivalent major leaguer for him really. Just thought someone would be willing to give something. However, with 16 there were only players who debuted in the current years vs full minors so that may have bring more offers. Although I would have thought at least someone would be willing to give up a mediocre relief prospect or bench player lol __________________ College Football Sim League

02-14-2018, 11:45 PM	#4
NoOne Banned Join Date: Apr 2015 Posts: 7,273 Infractions: 1/2 (4)	you may need to dig a bit. don't rely on "shop player" or just picking out 1 trade partner based on prospect, unless you are going to pay for it no matter the cost. if on the other hand you are looking for greatest rate of return, you have to seek highest demand. that can be a matrix of positional demands depending on complexities of the trade. find the teams with the red dot for that position in the mlb (prospects when trading prospects - it's about what they get). at first just start plopping guys in by potential to gauge interest from each team (obviously age, position etc play a role in value in addition to potential) ...if a couple of those teams with red dots have more positive trade feedback for distinguishably better players in return (add their prospects of various ability to resolve demand), start with those teams, otherwise, check out all the matching "red dots" have to offer. can always kick the can down the road by getting prospects.. lots of time to move them or use them etc. find the highest demand, hope they have some prospects or an mlb player of value. even if you have to package stuff together. not every good player you draft will fit into your plans. this is the opportunity to get somethign for them too. heck, maybe even a bench player you have a replacement for in AAA. cut payroll at minimum loss of ability, if any, and help a trade. 2 birds, 1 stone. 1:1 trades don't usually work out well for a human player. Last edited by NoOne; 02-14-2018 at 11:56 PM.

02-15-2018, 05:20 AM	#5
majesty95 All Star Starter Join Date: Apr 2005 Location: San Antonio, TX Posts: 1,789	I can try that and will. I guess I thought of the shop player feature as an easier and quicker way to do that. In my mind (and for time saving purposes) shop player should do that already. Maybe not all possible trades show up (especially with scouting on) so the astute owner can still find better deals. But, for the user trying not to spend hours of his available play time searching for one trade for a mediocre player, I would hope shop payer could help out with that. Or maybe ask the assistant GM to find a trade. Would just like a more simple option to unload players. __________________ College Football Sim League

02-15-2018, 11:34 AM	#6
NoOne Banned Join Date: Apr 2015 Posts: 7,273 Infractions: 1/2 (4)	it can hasten the process a bit... you can scan through a shop player result for best return... then dig deeper on a trade screen with that particular team. if you keep a incrementally maintained list of target mil prospects, you can also start there. sometimes you just can't afford the players on that list at the time of various decisions on players. hard to know where oyu are at, so tailor any advice to common sense context of your situation. go for big name target, go for most return for future trade, etc etc. i have a pretty set way of thinking through things, so i get a bit tunnel vision when i suggest stuff.