Do AI evaluation settings affect arbitration estimates?

[drhay53]

In effect, arbitration is attempting to set a player's value in years 3-5 at some percentage of their open market value. In order to do this, one must estimate the open market value, obviously.

So my question is; do the AI evaluation settings affect the arbitration estimates, and also the player's evaluations of themselves in arbitration?

[Klaus 74]

I had to give you a 'Thanks' based on the question being so intriguing. I don't know and am anxious to hear from others that know the game far better than I. But this is something I am tempted to do.....right at the end of the season take notes on the amount offered to some of my players, go into Commissioner Mode, move them to other teams, and see what their arbitration offers are because I would have thought every team's scout would have an affect as well.

[NoOne]

well even if you turn ai eval off, this is still referencing statistical results for their contract.

i've seen highly rated palyers that performed poorly leading up to arbitration and they got a smaller contract that first year... then it jumped wildly as he had a great year the following.

So, i'd assume this is mostly it's own exclusive function to determin arbitration estimates. in RL it's all about results, so i think this is all about results, too.. a guy with more potential (maybe a 1st rd pick in RL as an rough equivalent to potential ratings in game) doesn't get a higher contract with athe same statistical results as somone from rd 30 with the same service time, nor if he has higher "current ability" liek bigger muscles.

should be inconsequential if you have player eval on or off, relative to this 1 thing.

[RchW]

My understanding is with AI evaluation off only ratings are used. Is the Arb module set up to use stats separately from AI eval.

I'd love to know the answer.

[OldFatGuy]

Originally Posted by RchW

My understanding is with AI evaluation off only ratings are used. Is the Arb module set up to use stats separately from AI eval.

I'd love to know the answer.

It really should be, shouldn't it? I mean IRL it's based only on actual performance not on the "potential" of a player. Or at least I don't recall some player winning a higher arbitration award simply because he had a high ceiling or something.

I think the Arbitration value should just be stats. Free agent value, on the other hand, makes sense to include the ratings.

[beorn]

Yeah, this was always a problem in earlier versions.

You'd have a couple rookies, each playing so-so ball, but if you looked at the arby estimates on the salary page, you could always tell which had good potential and which one did not. Very, very unrealistic.

I haven't played 18 yet, my online leagues will make the switch in the coming weeks, but I really hope that this has been improved.

[RchW]

Originally Posted by beorn

Yeah, this was always a problem in earlier versions.

You'd have a couple rookies, each playing so-so ball, but if you looked at the arby estimates on the salary page, you could always tell which had good potential and which one did not. Very, very unrealistic.

I haven't played 18 yet, my online leagues will make the switch in the coming weeks, but I really hope that this has been improved.

Never had a rookie get arbitration estimates. What's different about your set up.

[beorn]

Originally Posted by RchW

Never had a rookie get arbitration estimates. What's different about your set up.

In 17, I go to Front Office => Salaries and look at the projected salaries several years into the future.

[Klaus 74]

I did a bit of an experiment and I'm surprised of the results as it appears arbitration amount offers are not influenced by the value of the player as determined by one's scout. I thought they would be in part.

In Commissioner's Mode I forced a trade between Boston's Joe Kelly and the Mets Noah Syndergaard. Kelly's arbitration offers of 3.2M and 3.2M, and Syndergaard's 6.6M, 7.2M, 11.0M, and 14.3M are the same whether they were on their original teams or not.

[beorn]

Originally Posted by Klaus 74

I did a bit of an experiment and I'm surprised of the results as it appears arbitration amount offers are not influenced by the value of the player as determined by one's scout. I thought they would be in part.

In Commissioner's Mode I forced a trade between Boston's Joe Kelly and the Mets Noah Syndergaard. Kelly's arbitration offers of 3.2M and 3.2M, and Syndergaard's 6.6M, 7.2M, 11.0M, and 14.3M are the same whether they were on their original teams or not.

Yeah, if I am right, they are influenced by the real ratings.

[Klaus 74]

Originally Posted by beorn

Yeah, if I am right, they are influenced by the real ratings.

This is where it gets more interesting....The example I gave was right at the end of the regular season, when the arbitration process begins.The suggested amounts offered are the same as what the previous teams' owner was going to do but they only got traded that day as well and the amounts can be changed manually. So I simmed ahead to the Free Agents File date. I forced a trade between Jackie Bradley Jr. of Boston, and Jacob deGrom of the Mets. Bradley's arbitration's estimates after the season were going to be 8.9 and 11.6M if he wasn't traded and deGrom's 10.2 and 13.2 if he wasn't traded. I then simmed to the Preseason date. Now Bradley's numbers are 7.9 and 10.0 with his new team while deGrom's are 10.2 and 12.0 with his new team. No games have been played in between. So, in part, what the owner is offering a player in arbitration is influenced by the value of that player as determined by his scout as long as there was enough time for an evaluation.

[drhay53]

Originally Posted by Klaus 74

This is where it gets more interesting....The example I gave was right at the end of the regular season, when the arbitration process begins.The suggested amounts offered are the same as what the previous teams' owner was going to do but they only got traded that day as well and the amounts can be changed manually. So I simmed ahead to the Free Agents File date. I forced a trade between Jackie Bradley Jr. of Boston, and Jacob deGrom of the Mets. Bradley's arbitration's estimates after the season were going to be 8.9 and 11.6M if he wasn't traded and deGrom's 10.2 and 13.2 if he wasn't traded. I then simmed to the Preseason date. Now Bradley's numbers are 7.9 and 10.0 with his new team while deGrom's are 10.2 and 12.0 with his new team. No games have been played in between. So, in part, what the owner is offering a player in arbitration is influenced by the value of that player as determined by his scout as long as there was enough time for an evaluation.

It could also be that the market changed over the offseason as players were signed to actual contracts (which by definition change the market value). I would expect a player's market value to 'look' higher at the beginning of the offseason as all the best free agents have crazy contract demands that slowly decrease over time.

This is a somewhat intellectual exercise for me as I'm not sure it matters too much in how I'll manage my team, but if I knew the answer it might affect the way I weight AI evaluation. Glad other people are interested

[drhay53]

I also just want to point out that OOTP is somewhat unique from real MLB as regards to evaluating a player's open market value. Other than maybe the agent spouting some scouting reports, most guys who are signing free agent contracts have several years of MLB stats that set their market value. In OOTP, however, the scouting system muddies the waters a little on this point.

I do agree that the arbitration estimates fluctuate more with performance than with the scouting system's ratings. But open market value is affected by recent performance as well (depending, I guess, on your AI evaluation weights and settings). So to rephrase my initial question; does the arbitration estimate take into account your settings on how you'd like open market value to be estimated?

I think the only way to really test it would be to play around with the evaluation settings and see if there is any way to get the estimates to change, without simulating too much time in-game (which would cause the estimates to change due to other factors).

Of course, a developer could pop in and tell us how arbitration estimates interact with the evaluation settings (if at all) and make this a very easy exercise

[Klaus 74]

Originally Posted by drhay53

Glad other people are interested

Quote abbreviated. You got that right. I haven't been this intrigued since the day of those commercials with "How do they get that caramilk into those Cadbury Caramilk bars?" lol.

[NoOne]

Originally Posted by drhay53

It could also be that the market changed over the offseason as players were signed to actual contracts (which by definition change the market value). I would expect a player's market value to 'look' higher at the beginning of the offseason as all the best free agents have crazy contract demands that slowly decrease over time.

This is a somewhat intellectual exercise for me as I'm not sure it matters too much in how I'll manage my team, but if I knew the answer it might affect the way I weight AI evaluation. Glad other people are interested

this is somethign i always forget... it only affects Overall/potential rating... or just overall, whatever..

so, a high batting average won't show you a higher contact rating... it merely ups overall if they are batting better than their ratings indicates.

dont even look at the estimates if looking into this... always compare resulting contracts. estimates ahve purposely inaccurate or out-dated info - similar in concept to scouting inaccuracy.

you simply have to find a few playrs that perform in a similar statistical fashio, but one is noticeable rated more highly. see enough of those situations and you can be 99% certain weither the ratings influence arbitration...

logically, it shouldn't, but if it's used as some kind of "third rail" to guide what 'should happen' i wouldn't be surprised either.. it takes a really lucky year to outperform your ratings... it's not somethign that can happen a ton or for more than 1 year, typically.

weren't there patch notes in the past years saying arb estimates were being too influenced by ratings or the gist like that?? if you can find rhetoric like that in the patchnotes, it's hard to deny they are involved. you be able to get some good info if you can google up all the old patch notes and seasrch for various phrases relating to this.

Heck, in RL your previous contract helps set your value nearly as much as your production.... if you come out of arbitration at ~8m you expect a raise... and very likely get it, too. even if someone perfromed nearly as well, if they exit at ~6m for their last arb year, they very likely don't get as big of a bump becuae of being paid less the preivous year than the other guy, ceterus paribus (except in more extreme instances of results).

[NoOne]

Originally Posted by Klaus 74

Quote abbreviated. You got that right. I haven't been this intrigued since the day of those commercials with "How do they get that caramilk into those Cadbury Caramilk bars?" lol.

those are unfertilized bunny eggs... everyone knows that!

bet you don't eat another one anytime soon

[Klaus 74]

Originally Posted by NoOne

those are unfertilized bunny eggs... everyone knows that!

bet you don't eat another one anytime soon

OMG, I'm switching to the richest ice-cream I can get my hands on.

Lol!!, you're the best...