My NeL Player Profile Project

[luckymann]

Perhaps the first true superstar of black baseball, Grant earned his "Home Run" sobriquet by dint of his supposedly hitting 60 HR in a season playing as a young man for his hometown Findlay Giants. Like Frank Baker, his longballs over the course of his career in the majors were more timely than numerous, with SH logging him as having hit just 14 all told during the prime Deadball Era. He's widely regarded as the best SS pre-Pop Lloyd, for whom he actually shifted to the keystone sack to accommodate his spot in the Lincoln Giants during the mid-1910s.

I'm just about to start with a full 1901-onward dev-only test save to put my profiles through some more paces, with all of the NeL players who played prior to that coming into the game in one bunch via the 1902 Draft, so you'll see more of these early guys' profiles coming through for a while.

[luckymann]

The man without whom I'd hazard a guess the Negro Leagues would not now be considered major leagues. A very good, but not all-time great, pitcher and my profile strips back the hype to reflect that. EC's MLEs have given him a solid 54 career WAR and I am happy with how he has turned out here.

[luckymann]

A fine two-way player from the turn of the century. As per usual with the players of this era there's very little anecdotal and biographical data to go by, so I've had to get creative and use the imported data as my base.

I think he's come out well.

[luckymann]

A tough one, old Money Monroe. All of the anecdotals I have sourced extol his greatness and yet neither his raw stats nor EC's MLEs back this up at all, with SH listing his career slash at a very modest 279/340/344 and the MLEs marking him down further again. On the other hand, Mugsy McGraw is quoted as calling him the GOAT.

Confusion reigns. I have shut out all the noise and simply followed the stats, but ticked his POT up to 115% just to cover myself. What is agreed upon is his elite defence at 2B/3B/SS and in the end he profiles as a slightly glove-heavy slap hitter.

[luckymann]

A typical Deadball Era contact hitter who was fast and rarely struck out, "Chino" Moran played most of his career in Cuba. Fantastic on D at the hot corner with some time spent in the OF as well.

Straight up 80-110% profile. Seems there's a bug here in that for some reason I can't get a rating at 3B to show up for him. Hopefully it's just specific to this save. I'll test it elsewhere and edit here but one would suspect his defensive prowess will push his OVR and POT higher than they appear below.

Edit: Here's how he comes out in another save. Still something not quite right with that 3B rating, but at least it is better.

[luckymann]

A prototypical heavy-hitting / light-glove 1B from the Deadball Era and one of the few player the MLEs actually mark up, almost certainly because he played almost his entire career in his Cuban homeland. Pretty fast for a big man but, according to JAR, a poor baserunner.

Started quite young so a 75-108% trajectory.

[luckymann]

Another Cuban who played almost exclusively there, JAR reports Joseito was more highly regarded than Jose Mendez when they were teammates thanks mainly to his superb control, with his screwball reputed to be his best pitch. He also worked as a position player, mainly in the OF, but was a poor hitter.

75-110% profile.

[luckymann]

Pretty sure JAR has somehow conflated Luis and Juan Padron. Anyway, I had him done already for the EL so here he is. Pretty useful type, both on the hill and as a hitter / position player, and gets rated accordingly.

[luckymann]

Yet another Cubano who didn't venture far from home, playing just two seasons in the US, Rogelio was almost entirely valued for his glove and wheels. According to JAR, he was regarded as the finest defensive outfielder at the time in his homeland and could run like the wind. Also played a fine SS.

Started young, 70-105% profile. For some reason, OOTP imports him as a SP.

[luckymann]

Sadly, "El Pajaro" doesn't even get a mention in JAR's tome for whatever reason. Not so important for hitters as hurlers, with pitch array being the main element to be gleaned from these sources, but it's still nice to have these bits of extraneous information at one's disposal. By dint of him having played a single game in the bigs for the Cards in 1913 (he went 0-for-2), however, he does have a BBRef page. On this, I discovered that he was the first Spanish-born (the Canary Islands, which is a Spanish territory) major-leaguer, and with his homeland being just off the African coast, can also be considered the first African to have made it as well. It wasn't until Gift Ngoepe in 2017 that there'd be another.

Nice.

He was quick and appears to have been a deft bunter, but didn't hit for much and relied on his excellent glovework, mainly at SS but also at 1B with limited time at 2B and 3B.

Straight-up 80-110%.

[luckymann]

One final Cuban for today. Emilio was a good enough player to have made it to the Cuban HOF in 1956, but is fairly representative of his time and place. A decent enough hitter with an excellent eye, but no superstar, fast and well-rounded. Pitched a bit, but not enough or well enough for me to rate him there.

Standard profile.

[luckymann]

Well I need not explain to you how disappointed I am to have to write this but I am experiencing unexpected and unacceptable problems with this project.

It all stems from something I have long suspected that has now been reconfirmed by further testing. When a player is imported, he comes with a predetermined imprint that decrees how his ratings will behave over the course of his career. In the case of these NeL players, this by and large means that, as they progress they move toward the profile you see for them when they import.

For hitters, this is low contact / high power. For pitchers it is high stuff / low movement and control.

I had not been seeing this in a couple of my dev only saves and thought it only happened when recalc is being used. But now I have also seen it in the dev only test environment I was using to - what I thought, at least - fine tune my profiles.

Instead it has shown they are basically unusable in a practical sense. They'll work for a season or two but this imprint will gradually assert its influence and you'll end up not getting anything near the correct statistical output for the players.

I really thought I had done enough testing and seen enough evidence that these would work to start sharing them with you. Apparently not, for which I apologise. You don't know how gutted I am that this is the case.

They are still able to be applied but only using a variety of workarounds that make them unpractical for general use.

I understand why the game has to have this preordained imprint - player DNA if you will - for players when they import. I just don't understand why the game allows you access under the hood if the work you do there ends up having no influence. It's a bit like swapping in a souped-up engine only to discover when you start the car the crappy old one is back in there. Makes no sense at all.

So for now this project has to be put on ice. I'll still be working on the profiles but as I said they serve no practical purpose for the general OOTP gamer, for whom there is already a vast number of play options to suit their needs. Meanwhile, as per usual, the poor old NeL player has to just suck it and see.

For anybody willing to put even more work into the process than these already take to implement and accept a few compromises of realism, hit me up on the DM and I'll walk you through what is necessary.

Again, my apologies for shooting the gun too early here, I really thought I had it nailed down.

G

[weeqat]

Sorry to hear that you're shutting this down! I've been looking forward to digging in to try some of these players out.

I have a suggestion though which might prove to be a workaround. If you were to identify a roughly equivalent MLB player for each player that you want to create, and then import that player instead and edit him, that might solve the regression problem that you are seeing. A lot more editing to do though. I don't think it would even need to be that close of an equivalency, just a general career shape would probably be enough since you are going to modify all of the ratings anyway.

[luckymann]

Originally Posted by weeqat

Sorry to hear that you're shutting this down! I've been looking forward to digging in to try some of these players out.

I have a suggestion though which might prove to be a workaround. If you were to identify a roughly equivalent MLB player for each player that you want to create, and then import that player instead and edit him, that might solve the regression problem that you are seeing. A lot more editing to do though. I don't think it would even need to be that close of an equivalency, just a general career shape would probably be enough since you are going to modify all of the ratings anyway.

Thanks mate, I'm sure it's obvious to all how much love, time and effort I've been putting in here so to say I am gutted is an understatement.

Yes, the workaround you mention is one of those I have up my sleeve. There are similarity scores that can be used to find the right substitute player, then you just import that player, clone them to remove the direct link, and use my profiles to build the NeLer. It works well, as long as you don't mind knowing that Grant Johnson in your game is actually Luke Appling reborn anew.

I actually use a more involved methodology in which I have allocated a specific performance each season for the player, which is added to the editor the day before each Opening Day. These yearly stats add up to the total career target for the player in question, subject to the usual vagaries in-game. Again, it works well apart from certain instances where there is some regression to that imprint over the course of a season.

If you're interested in seeing this just DM me and we can swap email addresses and go from there.

The thing is, both of these involve even more work and I'm not sure many would be interested in following that methodology. Even the one I've set out here is probably beyond the interest of all but a few. I haven't even contemplated how to fix the pitchers - they have, on the whole, worked brilliantly in my Bucs save. But now I'm seeing all sorts of hinky stuff when retesting them.

It is this inconsistency I am experiencing that is the hardest factor to overcome, because it doesn't let you isolate the problem(s) and attack them.

I haven't given up by any stretch, but for now there's no point wasting people's time by putting them up here and claiming they work.

G

[weeqat]

Wow, the year-by-year adjustments sound like a l-o-t of work! ;-)

Because I play in a few online leagues I'm used to letting the development engine do its thing, for better or worse. I'm hoping to seed a few of these guys into a random debut league to see what happens...

[luckymann]

Originally Posted by weeqat

Wow, the year-by-year adjustments sound like a l-o-t of work! ;-)

Because I play in a few online leagues I'm used to letting the development engine do its thing, for better or worse. I'm hoping to seed a few of these guys into a random debut league to see what happens...

Actually, for the hitters at least, it isn't. Takes less than a minute per player to enter in the new data. But, as I said, I haven't even considered how to attack the pitchers. Because of the complications brought about by the interaction of individual pitch ratings, this approach would be a nightmare.

[weeqat]

In one of my online leagues I've had Turkey Stearnes for his whole career, he just retired in 1943. 2700+ career hits, 300 HR (and 300+ triples!), hopefully he'll be a 1st ballot HOFer. That's without any editing of his ratings. It seems like he and Oscar Charleston and a handful of other hitters have the right "hard coding" as you called it, but so many other players are ruined. I can't think of a single pitcher that has been successful.

[luckymann]

Originally Posted by weeqat

In one of my online leagues I've had Turkey Stearnes for his whole career, he just retired in 1943. 2700+ career hits, 300 HR (and 300+ triples!), hopefully he'll be a 1st ballot HOFer. That's without any editing of his ratings. It seems like he and Oscar Charleston and a handful of other hitters have the right "hard coding" as you called it, but so many other players are ruined. I can't think of a single pitcher that has been successful.

Buck Leonard is another, as is Josh Gibson. But this is the thing, I have also seen plenty of problems with these guys, although you're right in saying they are generally more reliable than others.

I've just had Dick Redding reach 350 wins in my Bucs save and the other big pitchers from that era - Williams, Rogan, Foster, Mendez, Pedroso - have all done well. Satch and Ted Trent have just come in and are killing it.

But again, it's the inconsistency that kills me because I'll import them into another save and they'll tank.

So infuriating.

[No Pepper]

I'm sorry to see this happen. Does it occur the same way for all imported, historical minor leaguers, i.e., non-NeL? Does the same ratings progression occur if you import the same player multiple times in the same universe at the same time to see if it is consistent? Has this occurred in previous versions of OOTP?

[luckymann]

Originally Posted by No Pepper

I'm sorry to see this happen. Does it occur the same way for all imported, historical minor leaguers, i.e., non-NeL? Does the same ratings progression occur if you import the same player multiple times in the same universe at the same time to see if it is consistent? Has this occurred in previous versions of OOTP?

To be honest NP, I know the comprehensive answer to none of these questions.

I've never mucked about with other minor league players, in this version or any previous ones.

My experience is that it is consistent with players in the same universe. For example I did some testing over the weekend with Grant Johnson and he behaved the same way when he came into the 1895 universe - I'd edit him to about 48 / 64 on CURR / POT and within a year or so he'd be 45/45 with very flat ratings across the board.

In the course of that testing, I did a few sims using cloned versions of players who rank at the top of the NeLer's similarity scores.

I did Grant with Luke Appling and then Josh Gibson with both Mike Piazza and then Jimmie Foxx. The results were OK but you are still vulnerable to the cloned player's imprint having too strong an effect on the NeLer, despite you having entered in a whole new set of stats for the ratings to be generated.

This, for mine, is where the problem lays. As I said earlier - why allow us to go into the editor in the first place if the changes we make aren't going to hold? I fully understand the imprint is necessary in lieu of having recalc to cross check each season. But just don't see why the edits aren't retained by the engine. Makes no sense whatsoever.

So as it stands my yearly edits are the best way I have and this continued testing has allowed me to hone them further so that now I have them incorporated into the spreadsheet of each profile. Even though it's less than a minute's work per player per season, I don't think there's any point in sharing them because few, if any, would be interested and those that are can PM me and I'll share them that way.