League Total Modifiers Guide

[NoOne]

Informal How-To for Manipulating League Total Modifiers -> LTM


Preface:

Most of this is right.. but i'm also confident that not all of it is. Some of this is gereralized, some is spoken with an intent to normalize stats by having static LTM. regardless of your preference, you will find useful information here... just speed read to next sentence when i mention anything long-term, lol. tried to make only pertinent comments about that stuff.

About the LTM associated with a League Total -> LT

These initial concepts work extremely well when you make 1 change at a time and not many. when changing 2+ at a time, typically accounting for these things is enough... in some cases like bb/so/hits it's not always the case. did you reduce AB with the change? so, totals, not rate, will likely change for types of hits as just one example facet to consider. when rates change of something that interconnects with other stats, then that will drastically re-shape your LTM(s) on a whole.

in general, when you change 1 at a time, you can easily predict what it will do. change the LTM by 3% and you will likely change the result by 3%. notice i use a %, not .001 as a unit of measure... .001 has a different weight depending on the current value of the LTM - stick to %'s for easy calculations of small changes. %'s will not be accurate for larger changes. it's caclulated more like compounding interest than simply multiplying by a percentage 1 time. (you get to use a sigma notation, wuwu)

The Real Math:
e.g. .800 to .801 is not a .100% change. it is a .125% change (.001/,800). and so on for .801 to .802... .802 to .803... you can see why applying a simple percent change will slowly become less and less accurate the larger the change. like compounding interest, once the ball gets rolling it outpaces a flat percent easily.

Also, account for the difference in # of AB in the League totals and what you are comparing it to - either 1 year's results or an average of many. Those totales and AB equate to rates of occurence. ~2000 extra AB - 250some hits, 25hr, whatever... any calculated LT or Target result must adjust for this. (my spreadsheet does this for you - New LTM Spreadsheet i believe it's called- file attached in forums somewhere - it does not do the correct math, it half-as@es it).

do not expect LTMs that are going to center your league statistics after only 10-20 or even 50 years. # of teams does impact this, figure estimates based on 30T 162g... you'll need 100+years of data... but, this info is also good for merely making minor changes on the fly - with or without autocalcing them. i generally use 100year runs when making some static LTM for a league - if/when i put the time in. usually takes me 2.5-3 runs, now.

About the LTM not associated with an LT
(except stolen bases, those are funky to work with in combination with LTM-LT related LTM. - xbh, ba, lots of things will change SB, the others are mostly independent of the LTM-LT related stuff)

Before working on the more important ones, i'd always have a firm foundation of LTM for the ones not linked to a Totals value in the settings - just have SB/SB% in the ballpark. these are usually things that will happen in a particular frequency no matter what. get these in order first because they are A) Easy, and B) impact AB/plate apearances, which can impact totals, but not probability of various things happening (i.e. same batting average but can influence # of AB)

When i say "change one at a time," i mean to focus on 1 LTM and making any additional adjustments that may cause. if it add AB, then make those small adjustments, but then don't aslo try to adjust those LTM influenced by this change. see the results first. with more experience I "think" I can change them all at once, but it turns into more of a mess as often as an improvement, lol... some combinations are not a problem, obviously.

__________________________________


How-To begins:


Fielding:

This is more art than design for me, but you could get much more sophisticated about it by using Fielding Pct, than what i do for the sake of ease and time.

i have an idea of overall fielding pct target. i'll use an auto-calced base to start, which is generally very close to what i want... but as i check in from year to year i look at most Errors by positions in the fielding stats screen. if i get them roughly similar to what they are in RL by position (proportionately for differnet # of games), i tend to hit my Field Pct target (a target caclulated from RL data, so no mystery there)


Okay, now that that is done, on to the big stuff. Assumes you've gotten the non LTM-LT linked LTM all squared away to a result you like or simply using auto-calc as a base then change on the fly for fine-tuning that won't likely cause any issues while working with other things.


Hits:

In addition to batting average(frequency of a hit), this will influence BABIP. Increasing hits in the same amount of AB has this effect. It wil also proportionately increase hits, 2b, 3b, etc. it does not seem to effect Sac bunts or any of the LTM below the League Totals section.

changing this will change the landscape a bit. i'd get this one in line before working on distribution of those hits as doubles, triples, and hr. don't worry about sac bunts, or hits... stealing attempts may also be influenced, but not 100% on that one. i'd work on that after getting hits and XBH in order.

XBH: get your batting average and so/bb in order before fine tuning these. each time you change proportion of AB/batted balls/babip it will effect these. you can work out the math logically and consider those +/-effects, but it won't always add up as you expect. it should be pretty close, though. It's just better to get BA/BABIP/SO/BB rates all squared away, first.

Doubles / Triples / HR:

If you reduce one, that reduction is dropped back into the pool of hits... i'm assuming they are evenly distributed based on LT ratios or expected resutls due to LTM, either way. so if 5% of hits are 2b, then reducing HR by 100 should add 5 doubles and the proporitonate amount of triples and singles, too.

i would guess since it is a hit, it stays within that realm... homeruns may differ? i doubt it, but i have a sneaking suspiscioun that reducing it somehow incerases BA... maybe that's just a result of an increased BABIP? it obviously affects BABIP, though, because it's part of its calculation.

BB:

This will affect AB totals which will obviously domino into other things, too. if you add 400 more ab by reducing BB by that much, you get your expected # of hits,doubles, hr, outs etc...

this can shift what other LTM's result in, significantly. definitely get this along with hirs/so worked out before bothering much with XBH.

***doing the math on this and other things while chanign multipl LTM will not always yield predictable results- even when you use the cumulative change from all the modified LTM. 1 at a time it works more consistently and when you hit that jump you can recognize and adjust when it occurs more easily.

HBP:

small potatoes, shuold be easy to predict and stay relatively consistent regardless of other statistics being whacky or near what you want to see. unusual for one tied to a total, but it's so small relative to 165k, which could explain it.

SO:

Again, this one has numerous side effects to be accounted for. reducing this by 400 will likely increase hits/XBH proportionately to batting average (should have noted before: the resulting long-term BA not the target, but they should align eventualy, lol).

BABIP:

This should be calculated from your League Totals on this screen. Only a concern if you use your own values for the LT/LTM. unfortunately, you need to know how many sac flies you are likely to hit per year, given an infinite sample size. if you choose this and hone it, it's safe to use that BABIP here... it will fall into line for certain.

_____ Non LT linked LTM ___________

Changing these is easy as pie... expect them to work as predicted from LTM change for all except SB and maybe DP's

GB%

Your guess is as good as mine.. i've read things in forum but never from an ootp source. i just rely on autocalc getting it in the ballpark and leaving it there.

Wild Pitches

I use this to offset the wonky balks.

Balks:

These are the most poorly refineable results, lol... absolutely the poorest resolution in the tuning.

e.g. in a 40team 162g leageu(1.33 x MLB) .494 = ~200 balks, .495+ = ~250. a .001 change resulted in a 25% incease in balks... lol!! in a mlb-like league it's 180 / 120 - a 33% change from a .001 tick, if i recall. this is 1000's of years of verifiable data, at this point... no arguing this. changing it further in either direciton has likely zero to very little effect. i did not investigat far enough to see if there were other "large steps" from a .001 change. (in another league i thought it was .594/.595, so the line in the sand may shift, but it still works the same basic way)

so, whtever the +/- balks have to my target, i offset with that many +/- to Wild Pitches. it affects the same player, so it sounds good to me.

PB / SF / SH:

Simple... works as expected without concern for anything else.

SB / SB%:

This LTM typically requires alot more time than others to get a good prediction - if using long-term sims to create a baseline statistical enviroment for your league to operate under in perpetuity. some LTM are clear within 20-30years, this one takes 100+... more is better, but i jsut don't care much about hte exactness of hitting my SB attempt target as i do with other statistics. if i'm off by 50-100 that's fine, if the % is spot on.

you may want this in the ballpark before working on other LTM, but more importantly, fine-tune it after the other LTM are all setup and as you want them.

changing success rate can affect attempts and vice versa... i've not found a good way to predict # of attempts relative to these to forces. if you change one at a time they are very predictable in nature... although i think there is an element similar to Balks here... could be wrong.. definitely not as prevalent.

so, if you reduce sb% LTM, this generally causes a slight uptick in steals. same inverse relationship with adjusting attempts. i just recently dropped SB attempts a bit and my SB went up by ~10% for no obvious reason. i wasn't changing 1 thing at a time, dummy, so i'm not 100% certain why it jumped an insane amount when considering i lowered the LTM .010. but, batting average didn't change more than .254-.255 or so.

this isn't unusual in my experience, though. this is why i think there is some tiered, stepping behaviour regardin SB/SB% and various levels or ratios of the 2.


Conclusion:

Above all esle, don't get too picky about hitting a target. if you are using long-term sims, the resutls will still be off by a certain % error.

secondly, being off by 50-100 on something in a mlb league is 1-3 occurrences per team.... that's not a big deal if talking about 25,000hits... but maybe important to you when dealign with 300 passed balls.

Put things in proper perspective... being close but not perfect will still create a static baseling environment... you'll know it can't deviate into anything extreme like .230BA or 5ERA, unless that's your target of course, lol.

i'll post some pictures of what 100years looks like with LTM that do not change. just Slash and ERA line graphs. i think they are in my LTM spreadsheet in forums.. should auto-fill as you add data to the table. fully scaleable - must be a 2 league setup, though, or you have to make changes to formulas and fill down.

[NoOne]

This is the 2nd long-term run for a league. the third one is in progress now and looking pretty. reduced HR and final touches on Stolen Bases and %.

just obp / slug / era, coudln't fit BA. anyway, looks like the Y-axis got cutoff a bit, the #'s aren't important. the fluctuations and other visual traits are important. in a well-balanced league you should not encounter any large shift in talent throughout - once its near an equillibruim point.

This data starts after running a leageu for 35 years (2016-2050). 2051-2151 and no statistical drift... just fluctuation from aging, development, and player creation randomness. all seed players are gauranteed gone. all players arose from natural creation processes. lowest minor league rung (2 leagues) won't dip below 28/30, most will have 28/50 (28 limit on 1 of the leagues). any changes to # of new players created per year were instituted in 2016 and had 35 years to near an equillibrium # of players in the MiL. should be enough?

Sometimes you get distinct decades or even 20+ years... sometimes it's quite unpredictable year-to-year... barring the historical changes that have caused most of the statistical changes in RL baseball, this is how it is supposed to work. you want a mound change? than adjust your ltm to make offense lower - just like league evolution does it!