League Totals and Modifiers

[crowhop]

Originally Posted by RchW

I keep LT static and use LTM's to change up various outputs if I want. I can have HR year(s), SB, doubles, more IP by starters etc..

I use an excel spreadsheet to track year to year output and give LTM's for each. I find autocalc is prone to bigger swings in output than I like.

At what point in the game calendar do you input the data each season?

[crowhop]

Is it known how much the League Evolution "more/less offense/pitching" swings the modifiers?

One work-around to getting the Totals to work the way I (wrongly) thought they worked naturally would be taking the resulting totals at the end of each sim'd season and inputting them. That would create a kind of natural league evolution as opposed to always using the defaults.

[Syd Thrift]

Originally Posted by crowhop

At what point in the game calendar do you input the data each season?

IMO the ideal time is during preseason. I guess you could take care of that at any time during the offseason; I just find the first day of the preseason to be a good time in general to stop and take a long look at what's happening in your league. That's also when I make any franchise moves / division realignments / etc. because that's when the "new year" starts in the encyclopedia that's part of the game.

Otherwise, if you do it during the season you'll wind up with league totals at the end of it that aren't based on the numbers you entered in and so if you don't like how they came out, you won't really be able to know how much to adjust. Doing it in preseason allows you to have a very, very vague idea of how things might look by looking at spring training, although IME the situations are so much removed from how regular season games play out that I think this works as a very general guideline at best.

[RchW]

Originally Posted by crowhop

At what point in the game calendar do you input the data each season?

On or before opening day, but recently ie last time I did this I ran AutoCalc on OD and recorded those LTM vs the spreadsheet. In some cases the AutoCalc number was close to the second decimal so I used that.

Also note; It's better to run AutoCalc on OD as the AI teams must get the MLB roster set. Running it early includes players not likely to be in MLB and may skew results.

[crowhop]

Originally Posted by RchW

On or before opening day, but recently ie last time I did this I ran AutoCalc on OD and recorded those LTM vs the spreadsheet. In some cases the AutoCalc number was close to the second decimal so I used that.

Also note; It's better to run AutoCalc on OD as the AI teams must get the MLB roster set. Running it early includes players not likely to be in MLB and may skew results.

Very cool. Do you keep tabs on run totals to keep them in line? Though I'm guessing after 60+ years of doing this, you can pretty well predict what you'll get..ha

I never even thought about doing it for the minors/other leagues.

[KBLover]

Strange question but does the game take any form of the league totals into player creation?

I ask because players now have a different mindset and are developed differently than players even in the 90's, let alone the 70's and 80's. I believe that's borne out in the league averages.

I know there's settings for player creation, but if the league totals have some kind of influence, I could let it shift organically (more/less evolution settings and not me having to come up with some kind of system) and not 100% randomly (my league just so happens to keep producing X type of player which is different than the usual 2019 MLB player).

I was getting excited in an earlier version of the game where my oldest fictional league in 2029-2033 went on this binge of Gap + Speed + Av. K hitters. They were everywhere, shaking up the game, the leaderboards, it was pretty exciting to have to re-evalute players to keep up.

I was wondering if that was "5 years of just randomness that somehow persisted" or if there was something else happening...and if that something else was in the league totals changing in that league.

[Syd Thrift]

Not in the league totals but yes there are player creation modifiers and they do change over time.

[RchW]

Originally Posted by crowhop

Very cool. Do you keep tabs on run totals to keep them in line? Though I'm guessing after 60+ years of doing this, you can pretty well predict what you'll get..ha

I never even thought about doing it for the minors/other leagues.

LOL does my OCD slip show?

On or about May 31 or 60 games played I take a look and have been known to make changes. Yeah both MLB and the minors are based on RL totals for 2009 taken from BR.

[andyhdz]

Originally Posted by Syd Thrift

So... if you're out in the future, you might actually want to turn recalc off. Recalc is kind of great for historical seasons in getting any kind of player either generated by the game or imported to look like baseball from that period. In the future, recalc is just going to recalculate the stats so that they look like the last year available. If you turn that off, the engine will just keep churning out stats based on the last recalc, and truth be told that probably won't change much in an individual year, but it could change around a little...

I was anyways under the impression turning off recalc sets your league in a sort of wild mode not tied to any baseline because over time the different mix of player ratings can not produce the same baseline. I have mine set to 2019 totals but the cool thing is you could pick any historical baseball year and even set it to change every year incrementally even if you are way in the future say you want the 1990 to 1999 stats in the 2030s.

[Syd Thrift]

Originally Posted by andyhdz

I was anyways under the impression turning off recalc sets your league in a sort of wild mode not tied to any baseline because over time the different mix of player ratings can not produce the same baseline. I have mine set to 2019 totals but the cool thing is you could pick any historical baseball year and even set it to change every year incrementally even if you are way in the future say you want the 1990 to 1999 stats in the 2030s.

You can do that as well, yes. In theory, if you turn recalc off then what you say can happen; however, for a 30 team league with full minors you have a couple thousand players in your system and so the chances of so many players being created with weird ratings that it skews the league averages is very, very low.

[RubeBaker]

If I'm doing a historical simulation, I select the year of the season I'm starting with in OOTP. For example, if I start in 1981, I'm going to use the totals and strategy from that season. If I start in 2021, I will select the totals from 2019 because it was the last full MLB season and the numbers are pretty close to real-life.

[crowhop]

Originally Posted by RubeBaker

If I'm doing a historical simulation, I select the year of the season I'm starting with in OOTP. For example, if I start in 1981, I'm going to use the totals and strategy from that season. If I start in 2021, I will select the totals from 2019 because it was the last full MLB season and the numbers are pretty close to real-life.

But what is your plan for 2121?

[Ratbelly]

Originally Posted by crowhop

Seems like last year's game was more dynamic...but it almost always tended to having SPs get like 400+ Ks per year, so it was kinda deadballish...ha.

I can't get 22 to automatically evolve at all.

After a few years, I had the DH rule implemented in the National league and Rosters were decreased to 25. I like it. Makes me rethink roster building strategies. I loaded up my bullpen with high stamina pitchers.

[KBLover]

Originally Posted by Ratbelly

After a few years, I had the DH rule implemented in the National league and Rosters were decreased to 25. I like it. Makes me rethink roster building strategies. I loaded up my bullpen with high stamina pitchers.

One of my saves has rosters at 24. Which is going to be interesting with the modern bullpen usage.

So far in the save I started in 22, I've had a relocation (Year 2) and Expansion (Year 4).

[markprior22]

Originally Posted by RchW

On or before opening day, but recently ie last time I did this I ran AutoCalc on OD and recorded those LTM vs the spreadsheet. In some cases the AutoCalc number was close to the second decimal so I used that.

Also note; It's better to run AutoCalc on OD as the AI teams must get the MLB roster set. Running it early includes players not likely to be in MLB and may skew results.

Let me see if I understand...you're saying set up your league with the totals you want from bb reference and note the modifiers in each league. Then, around OD of next season, you run auto calc and see how the modifiers compare to what you started with. If something looks like it's starting to get out of whack, you readjust the modifier. Is that correct/


Also...will it work to use baseball reference totals from different years? (i.e. hr's like 1987, stolen bases like 1996, doubles like 2002, etc). Or is it better to pick totals from the same year and ride with that?

[RchW]

Originally Posted by markprior22

Let me see if I understand...you're saying set up your league with the totals you want from bb reference and note the modifiers in each league. Then, around OD of next season, you run auto calc and see how the modifiers compare to what you started with. If something looks like it's starting to get out of whack, you readjust the modifier. Is that correct/

No, I'm not quite that smart. Originally I started these leagues using MLB 2007 totals before autocalc was in the game unless my memory is faulty. I simmed 5 seasons one season at a time and manually did the first one then realized that it would be a lot easier to do a spreadsheet in excel. Somehow I decided I liked 2009 output better and used them from 2010 and on in several versions through v22.

Your description above looks like something worth trying though. I may have time to try later.

Also...will it work to use baseball reference totals from different years? (i.e. hr's like 1987, stolen bases like 1996, doubles like 2002, etc). Or is it better to pick totals from the same year and ride with that?

No because changing totals instead of modifiers has the reverse effect. If you start with 5000 HR and set LT to 5500 a 10% increase, you will actually get approx 10% less HR or 4500 HR.

You can still get what you want, or more accurately get close to what you want by adjusting modifiers. It's what I do.

However it's advisable to consider the proportionality ie total hits split into singles doubles triples and HR. If you raise HR significantly those will affect league wide BA OBP SLG but not BABIP. Changing the other hit types will affect both.

Not sure how clear that is.

[SirMichaelJordan]

One thing I’ve been having a problem with is that later in the years I’ll get some starters hitting .360+ and a bunch of sub .200. However, the overall League averages are on point.

I wish I could condense that a bit.

[NoOne]

Originally Posted by crowhop

Seems like last year's game was more dynamic...but it almost always tended to having SPs get like 400+ Ks per year, so it was kinda deadballish...ha.

I can't get 22 to automatically evolve at all.

this probably explains the high babip. If the modifier or total is too high for strikeouts, then hits take up a greater portion of balls in play with a smaller pie as more outs are occurring 'not in play' so to speak.

Did you change BABIP value in game without editing the other totals? The other numbers have to jive or it won't work as well as it should when you autocalculate the modifiers. Or, if it is a custom game, ensure it's got the same values as an MLB standard/quickstart league totals.

[NoOne]

i have a spreadsheet buried in the boards somewhere that can help with keeping track of results and sets of LTMs. has to have 2 sub leagues or requires some editting. If the csv output files have changed... also another hiccup, lol.

1 year isn't enough to be reliable long-term for sure.. with a 30-team league. Larger the league, the more reliable 1 year's results become.. or reverse with a smaller league.

need a good 50-100 years to get the averages of various stats. Some come quicker than others. If i recall, stolen bases is a good example of one that takes forever to find a steady long-term average. A whacky start to any league can make it take longer to find an average with confidence in any 1 statistic.

These are dials that augment or shrink various outcomes. I always preferred they stayed constant, but to do that you need to run multiple long-term sims to derive a set that will work relative to current league settings

funky things happen too.. these aren't solid 'curves'. Some, like balks, have very little resolution. i guess that could have changed last couple years. But, if you change that Balks LTM value drastically, it potentially does nothing if you don't cross a threshold. Yes there are 1,000 points below 1, but only a few large ranges of values that cause a change.

Anyway, quicker, less precise way would be to simply autocalculate to start and make minor adjustments over time if you see yearly totals dip too low in various categories.

Just understand these dials are shifting around bits of the same total pie. if you add 2000 hits, relative to previous years you will also get proportional additions to 2b, 3b, and hr, also fewer ~2000 'other' outcomes in proportion from previous years all other factors remaining the same... If you add a bunch of walks, that'll reduce total number of hits per PA, so you get fewer hits 2b, 3b and hr, even if same slash. Anyway, it's fluid.. like squeezing one of those stress balls.. some equal amount pops out somewhere else.

Add to doubles without changing total number of hits, then you end up with fewer of the other types of hits than you did in previous seasons (on average). slugging affected but avg and obp is not.

The more often you adjust one at a time, the better off you will be at getting what you want. The ones that have very little to do with each other being an exclusion.

don't react to any mid season results. you'll just chase your own tail on that one, unless you run a very, very large league. Remember, 1 year is bad info to work on, so half a year is horrendously bad info to make a decision solely based upon.

You can find a constant set that will work well, and allow the natural random creation of players dictate year to year volatility in stats, which imho is how it should work.. not dials. -- This excludes intended 'evolution' or anything that is akin to that... maybe you just want a big change in statstical output, it doesn't matter why or how it is role-played.

[andyhdz]

Originally Posted by SirMichaelJordan

One thing I’ve been having a problem with is that later in the years I’ll get some starters hitting .360+ and a bunch of sub .200. However, the overall League averages are on point.

I wish I could condense that a bit.

I notice that too.