Where is Jeromy Burnitz in 1993?

[M's rule]

swede61, here's more news about Burnitz and Darvish, none of it great.

Like others who have tried, I have no problem with Burnitz. In a league that starts in 1992, he imports to the Mets roster prior to the 1993 season. In a league that starts in 1993, he's a member of the Mets on Opening Day (on the reserve roster). There's nothing in the transactions file that would prevent this from happening.

The trouble with Darvish does not show up in a league that starts in 2016; when 2016 rolls around to 2017, he remains with Texas until the trade to LA on 7/31. It's not clear what year your Darvish league begins in, but that fact is important, because . . .

In a league that starts in 2017, he incorrectly imports as a member of the Dodgers. (Did your league start in 2017?) This indicates a problem we've come across a few times before now---the historical database has some kind of error that's causing Darvish's stint with TEX in 2017 to be ignored. These mistakes are impossible to anticipate; the only way to locate one of these glitches is to stumble on it, which means you're the first person to identify this one.

To get this fixed, we need to bring it to the attention of the historical database guru. In the meantime, the only solution is to edit Darvish to be on the TEX roster before the 2017 season gets under way. Sorry about that . . .

[pstrickert]

Here's a shot in the dark. Does the historical database have the order reversed for Darvish in 2017? By that I mean, do his stats for LA appear before his stats for TEX? And if so, does it matter? I have no idea.

[thehef]

Originally Posted by M's rule

Burnitz does not have a phantom entry in the transactions file. If I were to insert one for him, it would make him inactive (“injured”) at the start of the season. So if you play with transactions on but deselect the IL option, he’ll begin as an active player... Puig does have a phantom entry that keeps him inactive at the start of 2013. Again, this will not affect his availability if you play with IL transactions turned off...
The game does ignore a player’s debut date. That’s why the phantom entries have been created: to do the work of keeping rookies off the roster until they actually joined the team.

These are all good to know. Thanks for clarifying.

Curious, though, why Puig has the phantom entry and Burnitz doesn't. Seems like very similar circumstances. If the answer is simply that the phantom entries are a work-in-progress, then I totally get that.

Originally Posted by M's rule

... he (Burnitz)’ll begin as an active player and probably start the year on the reserve roster.

In my test, he made the opening day roster as a reserve outfielder, but ya, I suppose what exactly OOTP AI does with the 1993 Mets' roster will vary for each individual league.

Originally Posted by M's rule

Combining real transactions with fictional injuries potentially leads to big problems. What happens when a player suffers a season-ending injury before he's scheduled to be traded? The trade goes through, of course, even though it makes no sense for the receiving team to consummate the deal anymore. Despite the promise of “Play it any way you want,” there are parts of OOTP that don’t cooperate with each other very well. This is one example.

I understand what you are saying, but I don't get it as being a problem. If there are those who want to avoid the "long-term, fictionally-injured player traded via the historical file" scenario, they can either play without OOTP/fictional injuries, or than can play without historical transactions. Personally, I don't see the need to take the option away (not saying you are suggesting that)... And for how I "play it my way," that scenario is not at all a problem because I use a modded injury file doesn't have long-termers (no SEI's or CEI's; nothing over 30 days; in addition, I mod the injuries so I don't have all the new, fancy names for body parts and so forth for a injury that occurs, say, in 1935... a strained ulnar collateral ligament injury in my game is probably something like sore elbow.).

So, basically, the issue you describe can be handled, if desired, by the gamers setup choices.

Originally Posted by M's rule

... the Darvish problem...

I show that the txn file has Darvish going from Tex to LA on 7/31/17, to the Cubs on 2/13/18 - both of which match reality.

Originally Posted by M's rule

I'm also curious about "important trades being made too early, or not at all." I can't imagine there's an error of this sort in the data. Can you provide some examples?

I doubt there's anything widespread to this, other than perhaps some funky things going on with swede61's specific game. Otherwise, we'd've heard about it.

Speaking of swede61's funky game, swede - have you tried a simply test version of scenario, starting up a league in 1992 and playing until opening day to see if Burnitz can be found anywhere? If not, I highly recommend you try that and report back

[swede61]

Originally Posted by M's rule

Late to the party, as usual.

Burnitz does not have a phantom entry in the transactions file. If I were to insert one for him, it would make him inactive (“injured”) at the start of the season. So if you play with transactions on but deselect the IL option, he’ll begin as an active player and probably start the year on the reserve roster.

Puig does have a phantom entry that keeps him inactive at the start of 2013. Again, this will not affect his availability if you play with IL transactions turned off.

The game does ignore a player’s debut date. That’s why the phantom entries have been created: to do the work of keeping rookies off the roster until they actually joined the team.

Combining real transactions with fictional injuries potentially leads to big problems. What happens when a player suffers a season-ending injury before he's scheduled to be traded? The trade goes through, of course, even though it makes no sense for the receiving team to consummate the deal anymore. Despite the promise of “Play it any way you want,” there are parts of OOTP that don’t cooperate with each other very well. This is one example.

swede61, I need more time to figure out the Darvish problem. I'll get back to you.

I'm also curious about "important trades being made too early, or not at all." I can't imagine there's an error of this sort in the data. Can you provide some examples?

M's - Thanks for responding...

The one that stood out was the Darvish one...I had just started a historical replay of 2017, (had NOT played 2016, so this was the START of a game) when I saw all the weirdness with TEX pitchers, and had to do digging....I have played a few replays, different eras, and I think I saw something like this before, though can't recall specifically who it was...So I started writing down all transactions for the year I'm about to replay, so I can put guys on their appropriate teams to start the season...If I recall who it was, or see a similar situation, I will certainly mention it

Swede

[swede61]

OK, just started a new game, from 1992, ran the season, up to end of 92 playoffs, checked for Burnitz, and he IS there in this game, on the Mets..

This is so bizarre; why wouldn't he be in the game I also started with 92 season, and am now in May 93 with?

And more importantly, what the heck do I do? Can I copy and paste a player?

Urghhh

Swede

[M's rule]

hef, I feel you. If you're willing to do all the work needed to get these systems to play nice with one another (and you obviously are), then kudos. I guess I should have said that anyone who tries to combine them out of the box, without taking other steps, is going to have problems.

Phantom entries are a work in progress, as is the entire transactions file. I can put one in for Burnitz, but it won't show up until testing starts for 22.

pstrickert, you're on to something.

I called up an Excel spreadsheet of the content of the Pitchers2 file, which I got from Lukas Berger a while back. Not the same as the .odb file that the game uses.

The entry for Darvish in 2017 has lines for TEX and LAN, but there's a column for "stint number" that's blank in both of these rows. Clearly those values should be 1 and 2, respectively. As it happens, the line for LAN appears ahead of the TEX line, so in the absence of a stint number the program just looks at the LAN line and puts Darvish on the Dodgers right away. That's my best guess.

I have no idea why this problem does not also crop up in a league that begins in 2016 or earlier.

[swede61]

Thinking outside the box, I suppose I could , using the editor, turn a player from the 1800s into Burnitz, transposing all his ratings, though he would have different player ID, and logically, I would have to do re-do ratings for him every year?

Any thoughts? Am I right, that there is no way to get him into my ongoing game otherwise, if I can't find him on search?

I really don't want to lose where I'm at....

[thehef]

Originally Posted by swede61

Thinking outside the box, I suppose I could , using the editor, turn a player from the 1800s into Burnitz, transposing all his ratings, though he would have different player ID, and logically, I would have to do re-do ratings for him every year?

Any thoughts? Am I right, that there is no way to get him into my ongoing game otherwise, if I can't find him on search?

I really don't want to lose where I'm at....

Have you tried simply importing him? Obviously, that's not going to revert your game to the 1992 start, but at least you can move forward with him - assuming he imports successfully - from 1993 (or wherever you are) on...

[swede61]

How do you import a player?

[pstrickert]

FWIW, the OOTP team is planning to consolidate databases used in the game, including the historical database. I believe Lukas is working on the project, now that Spritze has died. Hopefully, it'll keep things consistent in the game and -- who knows? -- may help us solve certain problems more quickly.

[thehef]

Originally Posted by swede61

How do you import a player?

Go to the dropdown for your league (likely "MLB") and select Transactions > Free Agents, and then on the far right, under Actions, select Import Historical Player. At that point the Historical Player Import Options pop-up appears and you'll need to supply the Historical Player ID and select other options...

(I believe Burnitz' historical ID is burnije01, but I am not 100% certain.)

[thehef]

Originally Posted by M's rule

hef, I feel you. If you're willing to do all the work needed to get these systems to play nice with one another (and you obviously are), then kudos. I guess I should have said that anyone who tries to combine them out of the box, without taking other steps, is going to have problems.

Well, not to get too sidetracked on this... But I don't see this as being much different that using historical transactions with development on, and that resulting in some head-scratching trades. Take, just as a random example, the Dec '64 trade of Frank Howard by the Dodgers to the Senators in exchange for Claude Osteen (other players were also involved). This was - at the time and in hindsight - a relatively balanced swap: A decent starting pitcher in exchange for a decent power-hitter. But if you play OOTP with development on - especially if your development settings are high - this could very well turn out to be a swap of a hall-of-famer for a scrub. (Or two HOFers. Or two scrubs. Or whatever else.)

OOTP has A LOT of settings & variables, some that in combinations with others can be headscratchers to some gamers. But what might lead to whacky results (or trades) for some might be considered desired results for others As long as the options are there, we're good!

Originally Posted by M's rule

Phantom entries are a work in progress, as is the entire transactions file.

Got it. Of course that makes sense

[thehef]

Originally Posted by pstrickert

FWIW, the OOTP team is planning to consolidate databases used in the game, including the historical database. I believe Lukas is working on the project, now that Spritze has died. Hopefully, it'll keep things consistent in the game and -- who knows? -- may help us solve certain problems more quickly.

Interesting, I hope that is result in addition to it not restricting the ability to use certain custom files.

[thehef]

duplicate

[swede61]

Thanks all-

I imported , and now at least Burnitz is in my season...will have to do things manually for him I'm sure...

Appreciate all input

[swede61]

Just thinking...

I assume it HAS come up, the issues with players being on rosters too early in a season? I've seen plenty of players accumulating ABs or IPs in April, then you check BBREF gamelogs, and they didnt play til August?

Also, seen pitchers who had 25 games pitched, ALL starts, being put on the roster as a RP, and accumulating relief appearances...

And in '42 replay, seen relief pitchers with 1 SV opp, get the CL designation, while the Rp with 10 or so saves did not?

Are these just things we accept and deal with in what is otherwise a great game?

Thanks

[Lukas Berger]

Originally Posted by M's rule

hef, I feel you. If you're willing to do all the work needed to get these systems to play nice with one another (and you obviously are), then kudos. I guess I should have said that anyone who tries to combine them out of the box, without taking other steps, is going to have problems.

Phantom entries are a work in progress, as is the entire transactions file. I can put one in for Burnitz, but it won't show up until testing starts for 22.

pstrickert, you're on to something.

I called up an Excel spreadsheet of the content of the Pitchers2 file, which I got from Lukas Berger a while back. Not the same as the .odb file that the game uses.

The entry for Darvish in 2017 has lines for TEX and LAN, but there's a column for "stint number" that's blank in both of these rows. Clearly those values should be 1 and 2, respectively. As it happens, the line for LAN appears ahead of the TEX line, so in the absence of a stint number the program just looks at the LAN line and puts Darvish on the Dodgers right away. That's my best guess.

I have no idea why this problem does not also crop up in a league that begins in 2016 or earlier.

The pitching1 file, has the stints correct but, as you mention, the pitching2 file does not, for whatever reason.

I've corrected this now, but it won't show up in game until OOTP22.

[thehef]

Originally Posted by swede61

Just thinking...

I assume it HAS come up, the issues with players being on rosters too early in a season? I've seen plenty of players accumulating ABs or IPs in April, then you check BBREF gamelogs, and they didnt play til August?

This sounds like what M's rule is working on... work in progress.

Originally Posted by swede61

Also, seen pitchers who had 25 games pitched, ALL starts, being put on the roster as a RP, and accumulating relief appearances...

You can generally mitigate these types of things with your settings. But there could be legitimate things that also explain them.

Originally Posted by swede61

And in '42 replay, seen relief pitchers with 1 SV opp, get the CL designation, while the Rp with 10 or so saves did not?

Well, we can look at actual stats and find real-life examples where the roles don't match the performance. Take the 1977 Dodgers, just as an example. Their three primary 'pen guys posted these numbers:

a) 3.32 ERA, 1.319 WHIP, 7.4 K's per 9 innings, 1.5 K-to-BB ratio
b) 2.73 ERA, 1.309 WHIP, 4.2 K's per 9, 1.32 K-to-BB
c) 1.98 ERA, 0.848 WHIP, 6.6 K's per 9, 3.92 K-to BB

Who among those three would you want as your closer? I'm going with the last guy, no question. In reality, though, who where these guys and what were their roles?

a) Charlie Hough, closer thru June, then mostly setup guy (22 saves)
b) Mike Garman, setup guy thru June, then mostly closer (12 saves)
c) Elias Sosa, middle-reliever/setup guy all season (1 save)

As many on this board have said over the years, OOTP doesn't know names, only stats. So my guess is that if OOTP were to evaluate these players on those 1977 stats, it would put Elias Sosa in the closer role.

[swede61]

Originally Posted by thehef

Well, we can look at actual stats and find real-life examples where the roles don't match the performance. Take the 1977 Dodgers, just as an example. Their three primary 'pen guys posted these numbers:

a) 3.32 ERA, 1.319 WHIP, 7.4 K's per 9 innings, 1.5 K-to-BB ratio
b) 2.73 ERA, 1.309 WHIP, 4.2 K's per 9, 1.32 K-to-BB
c) 1.98 ERA, 0.848 WHIP, 6.6 K's per 9, 3.92 K-to BB

Who among those three would you want as your closer? I'm going with the last guy, no question. In reality, though, who where these guys and what were their roles?

a) Charlie Hough, closer thru June, then mostly setup guy (22 saves)
b) Mike Garman, setup guy thru June, then mostly closer (12 saves)
c) Elias Sosa, middle-reliever/setup guy all season (1 save)

As many on this board have said over the years, OOTP doesn't know names, only stats. So my guess is that if OOTP were to evaluate these players on those 1977 stats, it would put Elias Sosa in the closer role.

Excellent points- but wouldn't saves be one of the stats to look at? At least in terms of role designations? We can all agree the "value" and usefulness of the save is way overrated, certainly proven by your example where Sosa is clearly the best RP of that triad; but in assigning players to roles, such as closer, if one is attempting to make a historical replay as historical as possible, Sosa wouldn't be the closer.

I have gone into "pitching" to alter roles to better fit the actual historical circumstances, and have noticed that OOTP likes to "put things back" the way they see it...For example, if I have a closer who averaged getting 4 outs per appearance, I change his "usage option" to "8th+ or later" rather than "9th or later"...Unfortunately, it doesn't stick...seems like as soon as the next day comes, or there is an AI roster move, (not sure which), everything reverts back to the prior...have noticed the same under "lineups", when I try to alter order of PHs, especially in a case where a PH may have gotten 60 ABs..

Again, I think you are probably correct in that OOTP is choosing who may be the best hitter in each spot, but if so, why have settings if they don't stick?

[Lukas Berger]

Originally Posted by swede61

Excellent points- but wouldn't saves be one of the stats to look at? At least in terms of role designations? We can all agree the "value" and usefulness of the save is way overrated, certainly proven by your example where Sosa is clearly the best RP of that triad; but in assigning players to roles, such as closer, if one is attempting to make a historical replay as historical as possible, Sosa wouldn't be the closer.

The roles are chosen by the ai based on the player's abilities as it sees them in game. The stats are only going to affect the role the AI chooses for each pitcher in so far as they're taken into account in your league's settings of the mix to use of stats and ratings. So if you're using 100 stats for the evaluation, then the stats would be all the ai uses. If you're using 10%, it's going to be based 90% on ratings. etc. etc.

Even with 100% stats though, afaik the saves stat is not going to play into things at all, as all the ai is using to make its determination is actual performance stats, not 'reporting' stats like W-L, Saves, etc.