Feeder leagues are broken -- what's wrong and how to fix it

[stealofhome]

So after some more playing around and analyzing the real MLB draft, I have to update some of my targets. I plan on updating the OP eventually.

First, the 25% high school number is just wrong. I was looking at the MLB.com analysis of the 2017 draft and they used that number. However, when I do the numbers myself over the past 20 years, I find that the ratio between high schoolers and college is closer to 40/60 over the first 5 rounds. This means the OOTP MLB Quickstart percentages are right on target. The default feeder is still way off, but it means my custom feeder settings need some more tweaking.

Also, I noticed that many of the "1B" in the draft are actually many other positions disguised as 1B and once drafted actually get used by their team at their more valuable spots. This is a bug and I have submitted it as so. I have been analyzing the top ~300 players in each OOTP draft and comparing the ratios with the real MLB draft. College/HS, Hitter/Pitcher. The cutoff for the top 300 players is typically around 30/35 potential which I guess is fine but still seems low to me. I guess in my world I'd rather have a bunch of high potential prospects who bust than pick through a bunch of 30 potential players in the 5th round.

Once I get a better understanding of how the number and potential of players changes with feeder league settings I hope to post a "how to" guide that will work for different universe types.

[ra24mpa]

Originally Posted by stealofhome

I guess in my world I'd rather have a bunch of high potential prospects who bust than pick through a bunch of 30 potential players in the 5th round.

Agreed. After a few playthroughs reaching the year 2024 & 2025 draft, I can honestly say that I lost interest in drafting way too early. I mean, before the tenth round I'm actually drafting my scouting director's recommended pick just to get it over with.

[stealofhome]

Originally Posted by ra24mpa

Agreed. After a few playthroughs reaching the year 2024 & 2025 draft, I can honestly say that I lost interest in drafting way too early. I mean, before the tenth round I'm actually drafting my scouting director's recommended pick just to get it over with.

Yep, I usually try to get through at least the 10th rd on my own and then autodraft the rest. With those low potential players, it's really hard to do any of that as they all look the same at some point.

[stealofhome]

I had success with 56 high school teams, 40 college teams, feeder league roster limits of 28, 5 man rotation, 10 relievers, 13 position players. Seems like you can manipulate the percentage of hitters/pitchers by fully defining the roster limits in the feeder leagues. The high school/college percentage is a little harder to nail down perfectly as it changes with total number of teams in each league as well as the ratio.

My goal based off of the first 5 rounds in MLB over the past 20 years is 0.93 hitter/pitcher and 1.39 college/high school. With the settings I mentioned, I achieved 0.87 hitter/pitcher and 1.54 college/high school with a total of 1,022 draft eligible players. I still want to look at the percentage of relievers to starters and total quality.

[stealofhome]

I've started looking into fielding ratings. The year one MLB Quickstart is made of real players so I'd like to recreate that repeatedly over many years as it is the highest quality draft. What I've found is that with the exception of outfield and infield arm, the players in the Quickstart draft have on average 15% better raw fielding ratings than feeder league players. The infield and outfield arm ratings are about the same for both leagues.

[stealofhome]

Another update on fielding ratings - a fielding pcm setting of 1.5 seems about right in initial tests. However, the tdp rating is much too low and will not change with this modifier.

[stealofhome]

Yes, I just verified at 10 years out that changing the fielding rating modifier to 1.5 gets everything just about right with the exception of the tdp rating which is simply a bug. If that is ever fixed I will have a feeder league setup that mimics the mlb quickstart draft.

So to recap: quickstart drafts are okay, the default feeder system is too large, feeder league players are not created or used properly, and I have found and reported three separate bugs related to feeder leagues. Two aren't game breaking (improper usage and wrongly labeled high schoolers) but one is (tdp ratings too low and not adjustable).

[itsmb8]

Originally Posted by stealofhome

Yes, I just verified at 10 years out that changing the fielding rating modifier to 1.5 gets everything just about right with the exception of the tdp rating which is simply a bug. If that is ever fixed I will have a feeder league setup that mimics the mlb quickstart draft.

So to recap: quickstart drafts are okay, the default feeder system is too large, feeder league players are not created or used properly, and I have found and reported three separate bugs related to feeder leagues. Two aren't game breaking (improper usage and wrongly labeled high schoolers) but one is (tdp ratings too low and not adjustable).

Thanks for your findings! Will definitely utilize this once everything is fixed and/or you have the guide up.

[stealofhome]

Here's an image of the average fielding ratings - first year quickstart, 10th year of a normal feeder league setup, and 10th year of a feeder league setup with 1.5 fielding ratings. You can see how every single attribute is much lower in the feeder league with the exception of IF arm and possibly C ability. The fielding modifier fixes all of this except it increases the IF arm when it shouldn't (doesn't impact OF arm much) but definitely does nothing for the turn double play rating.

[stealofhome]

The ratio between starting pitchers and relievers is also very different in every other draft compared to the MLB Quickstart first year draft.

I'll define a possible starter as someone with at least 3 pitches with 45+ potential and at least 40 stamina (20-80 scale).

In the MLB Quickstart, 78% of pitchers with 40+ overall potential meet those criteria and can be classified as starters. In feeder league drafts, that number is somewhere near 30%. Not enough pitchers are getting a third pitch. The movement and control numbers are also a bit low, but the primary reason is that so many relievers get created in the feeder league system with no chance of ever being a starter in the minor leagues.

To get around this problem, I have created feeder leagues with no relievers on staff in the league roster settings screen. Even with a definition of 0 relievers in the feeder league system, the percentage of starting pitchers coming into the draft is 58% - still nowhere near the 78% to start.

I'm still playing with the proper league roster settings to balance hitting/starting/relieving, but this most recent league (somewhere around the 60th iteration) is getting close to acceptable - once the TDP issue gets fixed.

[jfb8300]

I'm having an issue with 0 relievers. With a 28 roster limit comprised of 0 relievers and 14 position players, the roster settings page says "total players on roster: 19 of 28."

Filling the rosters gives each team only 19 players. The 14 position players and 5 starters (or 6 if I choose expanded rotations). How are you getting more starters in this set up?

[stealofhome]

Basically that. I put one reliever, 6 starters, I think 13 hitters in my last attempt. Then put the roster limit at whatever that total sum is - 20 in this case.

I mess with the other roster rules too - allow starters in relief, increase their stamina, use of relievers low, etc. It's more so removing relievers than adding starters. There's still some play in getting the quality up with the pcm that I haven't nailed down yet. I have a 50 year sim going right now just to see long term effects.

[TheMaus2]

Originally Posted by BusterKing

Here's my 2019 draft with OOTP19

my thoughts on this draft can be found here.

[BBGiovanni]

Originally Posted by stealofhome

Yep, I usually try to get through at least the 10th rd on my own and then autodraft the rest. With those low potential players, it's really hard to do any of that as they all look the same at some point.

Not to derail an excellent thread, but I do this too, except that I draft a lot of one-dimensional players (speed/stealing or outstanding defense) on the off chance they learn how to hit. Not particularly fruitful but gets me through 10-12 rounds then I let the AI mop up.

[Rain King]

Just want to jump in and say I also have noticed that there are way too many 2-pitch pitchers being created. I don't think that is just a Feeder League issue though, I think it happens in all instances.

[stealofhome]

Originally Posted by Rain King

Just want to jump in and say I also have noticed that there are way too many 2-pitch pitchers being created. I don't think that is just a Feeder League issue though, I think it happens in all instances.

I wonder if it's done this way on purpose since more pitchers are relievers than starters. However this begins to cause a problem when most starters are failures and even the starters become relievers.

[Rain King]

Originally Posted by stealofhome

I wonder if it's done this way on purpose since more pitchers are relievers than starters. However this begins to cause a problem when most starters are failures and even the starters become relievers.

Yea, in my opinion very few pitchers should come into pro ball with only 2 pitches. They should have at least a fastball, breaking ball and changeup of some kind. Those don't all have to be good pitches and maybe if the pitcher isn't very good in the first place, they never use some of the pitches (i.e. they are career minor league relievers) but all pro pitchers can at least throw that combination of stuff in a bullpen. I have no problem with the pitches dropping off at some point or something if they have been used only in the bullpen for a certain period of time.

[stealofhome]

Another update (didn't mean to have this thread as a research blog, meant to have the issues found before posting but I keep finding new things as this goes).

PCM changed - Pitching stuff 0.85, Movement - 1.25, Control, 1.15, Fielding 1.25

Feeder Leagues Teams - 64 HS, 48 College
Eligible - HS ages 14-18, eligible if max age reached. C - 18-23, eligible 2 yrs prior to max age
Roster Rules - 20 players, 6 man rotation, allow starters in relief, 2 relievers, 12 position players
Strategic tendencies - very slow hooks, very rare use relievers/closers, very high stamina, very rarely pinch hit for pitchers

With the very reduced roster amounts this league setup now only supplies 814 players for the draft so I'll have to increase the number of teams.

In the quickstart draft, very very very few players (3/1401, 0.2%) are listed as fragile. The standard for feeder leagues is an order of magnitude higher (~5%) but in this case that almost seems more realistic. I've noted it but not tried to change anything about it.

The ratio between college:HS is 1.45 which is right on. The ratio between hitters and pitchers is now thrown way off because there are only 8 pitchers for 12 hitters in the feeder league. However, this allows for the ratio between starters and relievers to go up to around 2. Not quite the 4:1 in the quickstart draft but better than having more relievers than starters.

The fielding ratings are on average good, but there are a few ratings in particular which are too low (TDP, C ARM, C ABL) which lead to way too few SS and C.

The potential pitching ratings are still a bit high but I have seen those come down over time and change a little between drafts and leagues.

With the exception of the total number of draftees which has now changed due to the roster limits and the fielding ratings which need to be fixed by the devs, this league is starting to look like a real feeder system. I've included the 1st round of the draft (a little short due to free agent draft picks losses) and it looks worlds different than before.

[stealofhome]

Okay last post on this until I see some movement with the fielding ratings in player creation. Here's a comparison of the "golden draft" - 2018 quickstart - with my best shot at a feeder league (same settings as above but 84 High School/64 College, the 10th year of a quickstart league, and the 10th year of a standard MLB league (New Standard Game, 2018 MLB).

I've put in red any value that is ~90% or lower or ~110% or higher than the 2018 draft. TDP and C ARM/ABIL are issues even in the standard leagues with no feeders.

[Lukas Berger]

Originally Posted by stealofhome

Okay last post on this until I see some movement with the fielding ratings in player creation. Here's a comparison of the "golden draft" - 2018 quickstart - with my best shot at a feeder league (same settings as above but 84 High School/64 College, the 10th year of a quickstart league, and the 10th year of a standard MLB league (New Standard Game, 2018 MLB).

I've put in red any value that is ~90% or lower or ~110% or higher than the 2018 draft. TDP and C ARM/ABIL are issues even in the standard leagues with no feeders.

I mentioned this in the other thread where we were discussing this, but the issues you're finding in comparing the feeder league classes with the quickstart draft class if anything, will actually mean that I'm overrating the players in the quickstart and need to adjust their values down, rather than that there's something wrong with how OOTP is generating other classes and that their ratings need to be raised.

Same with the amount of players that are fragile etc.

The debate as to whether draft classes should have higher or lower potential players is kind of an eternal one around OOTP (you can find plenty of threads about this with really spirited debate from previous years on the forums) and it's probably a debate we can't really win in that no matter what we do, someone is not going to be happy.

Obviously we want the draft to be fun but given that in reality there's little to no difference between the players being drafted in the 15th round and the 35th round (or signed as UDFA's) it probably doesn't make much sense to create an artificial difference between those players in OOTP just to make the draft a little more fun, especially since that could lead to issues with the overall talent levels in leagues over a longer period of time.

In the end, a lot of it is just personal taste. It's part of the reason the .pcm's exist, since they allow you to tweak things more to your liking.

That being the case, my biggest concern here is if adjusting the .PCM's is not changing the dp rating or C arm/abil or is changing the ratings incorrectly. Do you have a chart that documents that (maybe I just missed it)? EDIT: Markus fixed this, so never mind!