What is the best way to reverse this trend?

[Tyler87898]

So, my fictional MLB has been in existence from 2014 to 2090. I know in the beginning of OOTP leagues many people sim for 10-20 years because stats are a bit wonky, but I did not do that. Anyways, in 2014 the AL avg ERA was 4.10 while the NL was 3.57 (that difference is a bit weird, but that's not the point). Now in 2089, the AL avg ERA was 4.84 (!) and the NL's was 4.53. By the way, AL had the DH for the whole time while the NL adopted the DH in 2016 and removed it in 2053.

Over the years, the league expanded with a total of 6 more teams, so now there are 36 more teams. I'm not really sure what caused this extreme rise in offense. It probably has to do something with the expansions. I noticed that when the DH was repealed in the NL in 2053, the avg ERA dropped from around 4.30 to closer to 4.00 for about a decade and then started to rise quickly. The AL's ERA just continually rose.

The only league evolution that is enabled is turn on/off DH so no mounds were raised or lowered.

Another result of this increased offense/decreased pitching is that only about 1 SP in the AL and about 3 SP in the NL end the year with an ERA under 3.00 (with at least 180 IP).

Is there a good way that I can gradually tone down the offense? If you guys want any more details I can try to grab them for you.

[bwburke94]

Try messing a bit with PCMs, and generating an extra few rounds of draftees.

[NoOne]

"League Total Modifiers" will be the easiest way to fix it.

if you don't like the distribution and min/max of various ratings in your league, you would also want to touch on PCMs.

e.g. if you started with real players, you may notce some differences. fictional players have more power on average, strike out less on average, and have lower babip ratings. which is why you see the elevated ERA when compared to early seasons with real players. (assuming you played with real players based on your description)

PCMs can reduce/augment these changes in ratings distribution, if you want to do that. just becareful, you don't want the distribution to be skewed too high or too low compared to where it started.

[Tyler87898]

I don't know how much to change these numbers, and how often so that it is still realistic.

[actionjackson]

I know NoOne will disagree with me , but you could also use auto-calc to come up with different totals by selecting an historical year to base your totals on. I use 1984, but you could use any year from history that you want. Just don't use a steroidish year, or various 1920 and 1930ish seasons, as they have pretty high offensive numbers as well.

[NoOne]

Originally Posted by actionjackson

I know NoOne will disagree with me , but you could also use auto-calc to come up with different totals by selecting an historical year to base your totals on. I use 1984, but you could use any year from history that you want. Just don't use a steroidish year, or various 1920 and 1930ish seasons, as they have pretty high offensive numbers as well.

it's alot of work to do it without using auto-calc. surprised i didn't mention it. if someone wants to look up the thread with the spreadsheet, more power to them. there will be no tech support given, lol (or at least very little).

if the default numbers for HR was 4300 (or there abouts), those are probably good numbers already iin the League Totals. if it's 5k or more you may want to select a new year to auto-fill the values with better numbers. (5k is the value for a fictional league default, i think -- 4300 is what a modern mlb-like league starts with default)

so, just click auto-calc and it will overreact to yearly randomness :P

i had to get one jab in. it will prevent anythign from being too crazy as far as results.

otherwise run some 100+ year sims and start changing the ltm's. it might take 3-6 runs or more, if you do it in the right order. sounds like fun right?

[Tyler87898]

Originally Posted by bwburke94

Try messing a bit with PCMs, and generating an extra few rounds of draftees.

Why would extra rounds of draftees help? Wouldn't it also create more better batters?

[NoOne]

it should be proportional, yes.

the other side of that coin is what occurs when you add rounds. e.g. if you add 10 rounds to a 20 round draft, do you get 50% more of all quality of players, or is it just players that fit the quality of rounds 21-30? i'd guess it is the former, rather than the latter. either way, it should be proportional -- exlcuding the expected volatility of randomly generated players.

the league gets out of whack due to the difference in players. the LTMs affect a real player's distribution of talent (various raitngs) differently than they affect the fictional player's distribution of talent. because they have better eyes, more power and less babip, the same LTMs will cause very different statistical output with this new distribution of talent.

the dh/no dh evolution is probably why the nl didn't follow the exact same path as the AL. the teams had to convert lineups and strategies on the fly - so an extra dip to start after the dh removal seems reasonable. a non-dh league will put up less offense on average, regardless.

the offense isn't skyrocketing because of lack of pitching talent or too much batting talent. it's because the LTMs were calculated using a league of players that are significantly different than what gets created as the league progresses. whatever the differences are, they clearly favor the batter in regard to the original LTMs for a real-world league. it willhappen without fail. start any real league with default settings (no auto-calc) and once they are replaced by fictioanl players, offense explodes by 20-25% and fluctuates around that new average.

if i understand your situation well enough, hitting auto-calc just one time would significanlty change your stats the next year. the LTMs will be adjusted based on how far "off" they were the previous year when compared to the League Totals values. (not telling you to use it once or none or all the time -- that's just what it does)

going to the statistical accuracy report will show year-by-year the % off from an average. i don't know if that average is based on hardcoded numbers or your the league total values you use, so use that info with caution. (don't ask where it is. i don't recall. check the manual - i think it's only available after the season ends?)

-------- i write this below just to stir a thought or two - not trying to convince anyone. based on your orignal post, this might be what you want or it may not.

what actionjackson alluded to is something i do to avoid these problems and avoid artificialy changing the physics and game environment every year with LTMs to maintain a perceived average that "should" (i disagree) be hit each year... which has no place in a simulation without cause, but is fine for entertainment in any shape. the same reasons the game tells you not to fix the statistical output to hit the totals given is the same reason i sugges to not arbitrarily use auto-calc every year or even at all (it's a checkbox on same page with the reasoning stated in-game)

by simming very long periods of time, you are essentially figuring out hte average talent created by the fictional player creation process. if you experience enough league turnover and use those averages to calculate LTM's, you get a true average of statistical output over time and league turnover. at this point, the only fluctuations in statistical output are caused by randomness in the season and randomness of talent levels entering the league - and any other factors in the game coded to simulate real-life.

So, when you get a few drafts in a row with heavy pitching talent, you will likely experience a pitching heavy league in the near future - amazing thought huh? if batting is stronger, then offense will rise. i think that's the way it should be. if you are using auto-calc and you experience an influx of disproportion (batting or pitching), the LTMs will be changed to rein in whatever is out of whack due to that change in talent.

if you decide that the mound is going to be raised, you might lower some LTM and raise others. that would be a cause to change LTMs from this perspective. it doesn't mean they may never change. they just require a reason for the change.

you can only do this if you calculate a truer avearge of statistcal output relative to LTM/LT. one year or twelve isn't enough. this requires many generations of players. you need enough player turnover in the league to calculate it accurately. not only that, you need to repeat that process. it takes multiple long-run simulations to make all the adjustments necessary to LTMs.

tips:
never use the first ~50 years of data - even if you start out as a fictional leauge. 50yrs out is a safe point to start this type of project. make a backup at that point and start any testing from there. also, i'd start the league 50yrs before the time you intend to start your actual league. this way you can merely put the LTMs in and make use of that backup ASAP. it's better to use the backup or make a quickstart from it. you can delete the 50 years of stats and history if you wish. why? because, if you start a new league and it isn't exactly the same, the LTMs may or may not apply to this new league.

if that interests you, i have an old thread and a spreadsheet that can help you in this endeavor. beyond simple questions, i offer no technical support. if oyu know how to link to files and other basics of spreadsheets, it won't be hard. it will be time-consuming. it's a heck of a lot easier to ignore these things and just click auto-calculate each year.

[Tyler87898]

Thanks for the write up, but I guess I'm still a bit confused on how all the modifiers or whatever work. Guess I will have to do some long-term experimenting.

[bwburke94]

Originally Posted by Tyler87898

Why would extra rounds of draftees help? Wouldn't it also create more better batters?

To clarify: the intent is to neutralize the effect of the strong pre-alteration batters by creating a few extra rookies each year, increasing pitching depth while not affecting batters as much.

[NoOne]

Originally Posted by Tyler87898

Thanks for the write up, but I guess I'm still a bit confused on how all the modifiers or whatever work. Guess I will have to do some long-term experimenting.

it's complicated.. but i'm going to use an extremele simple example:

League Totals and modifier:
HR: 4000 (the value in the game stats& ai settings)
LTM: 1.100

results: 5000 home runs hit during the year.

auto-calc will reduce HR by ~20% (less than).

i could be wrong, but you'd actually use a sigma notation, if you are familiar with that math term. so you reduce 1.100 by .001 and that would subtract from the difference in home runs untill you have deducted enough: 5000- .001/1.100 * 5000, then (new total) - .001/1.099 * (new total), and so on until you reach ~4000. a simple financial calculator has this function built-in.. a graphing calculator or spreadsheet can do it, too. a guesstimate is good enough, really... because you are likely changing more than 1 interrelated LTM at a time.

the larger the difference the further short of proportional... ie 1000home run difference might be 16-18% less, not the full 20% from the quick and messy math above (1000/5000).

but, that's not really how it works in action. the LTMs are interconnected in some ways here and there. e.g. changing doubles will significantly affect triples. changing K's will increase/decrease batting average, which will have its own various effects on things realted to batting average. changing hits/singles mostly affects BABIP in your league... please note some of this could be a bit off, but it should get you thinking in the right way.

auto-calc will use last years results and adjust all of this for you. the only drawback to using auto-calc, in my opinion, is if your current league is currently very high or very low in talent.

imagine a horizontal line with a wave going around it. the area under/above the wave, relative to the line, is equal. the line is the center of it's fluctuation. if you have a league that is very high in talent and click auto-calc, you are raising that center of fluctuation. (vice versa with unusually low talent)

ugh the drawing works in preview... adding dots
.................................................. ..___
..... ___ ............. ___ ............... /.......\___ blah blah blah
__/____\ ____ /____\_____ /_______
.../............\____/.......... \____/

obviously, real life or even a mildly sophisticated simulation will not be so symmetrical and evenly distributed, but when the area above and below equaly, that is the center.

however, you will mask a low-talent league or you may decrease teh output of a high-talent league... since it's reacting to a small sample size, this is not automatic. luck (various random results of a season) could mask somethign from auto-calc etc etc...

my suggestion is to click it many times and average the results... the more you do it, the more accurate the changes relative to talent in your league. at that point you have created a baseline and relative to talent changes and league turnover, it will ebb and flow. if you were high in talent, it will mostly ebb... if you were low in talent it will mostly flow... in the middle and it will be farily even up and down.

forgot to mention: if want to see 10-15 superstars in each draft, do play with PCMs. just realize having too much talent will have its own repercussions. also, it will take an entire league turnover of newly created players to see the total effect on your league (~27+ years?).

[Tyler87898]

So, to be clear, league totals and modifiers both do not change from year to year automatically?

Also, in your example, why would the result be 5000 homeruns? What exactly does the LTM do?

[Tyler87898]

So I'm doing this test fictional MLB league. I'm in the year 2022 (after starting in 2015) right now. The league total for homeruns is 5900 (default) and the modifier is still 1.000, but this season there were only 4102 homeruns. How can this happen? I would understand some variance in the numbers, but not this much.

[Tyler87898]

Also, how can I see the league (not just sub league) offensive/pitching totals from year to year?

[actionjackson]

Originally Posted by Tyler87898

Also, how can I see the league (not just sub league) offensive/pitching totals from year to year?

You would have to track those in a spreadsheet for yourself by going to Your League > Statistics > Team Statistics and copying down the stats you wish to keep track of.

[NoOne]

CSV exports! you can export that data and link a spreadsheet to the files. then as you export each successive time, all you do it open up the spreadsheet and it will ask you to update the data.

then you can do whatever you want to them. note that they are split by conference/league. so AL/NL would need to be added together if you wanted a total. if oyu have dh/non-dh i'd recommend keeping the data separate, and totals only for the fun of it. those sets of data are two different animals and shouldn't be mingled.

i have a thread with one that is set up for "data dumps," which is a different file and location. you can use that to start or make your own. i have the dh in both leagues so it's combined.

back to LTMs

just because it says 5900 and 1.000 doesn't mean it's going to hit that number. think of the 1.000 related to actual yearly totals, not the totals in the settings. so if the results are 20% too low, try 1.180 (a little less than 20% increase of that value, which happens to be 1 this time, if it was 1.200 already, then a little less than 1.440)

the LTM = the result. so if the result is X% different, change the LTM by roughly that amount.

the league total value just tells you how to calculate the % difference. (5900-4300) / 4300 (negative values mean reduction, ofc. if you do the math a different way, just use common sense. it's the same absolute value). maybe 1.300 to 1.350 ish? that's a steep incline to start. i might even go 1.200 to start out unless it is the only LTM that changes.

in this situation, again i'd sugges hitting autocalc first. your home runs could be off for reasons outside of the LTM specifically for home runs.... hope that makes sense. a ~40% increast is quite large... i'd be wary of that value.

in my experience some of them do not scale 1:1. i find tiers at certain points where large jumps occur from .. e.g. from 1.243 to 1.244. just .001 difference can on occasion be more than you expect.

Edit: 5900 hr is going to be a 4.80-5.20 give or take ERA league - best guess. it will be up there in runs scored. if you want numbers similar to what hte editor predicts or recent mlb years, load up a defualt MLB league and screenshot the stats & ai settings page. copy those League Totals to your league. run auto-calc.

any new league will increase in talent the first 20 or so years. auto-calcing during this time would be wise, if you plan to use it. so, that kinda throws a wrench into what's going on... after 20-30 years it will be in it's typical equillibrium of ebb and flow around a particular average league talent level. it may be low-4000 home runs in 2022, but by 2030 it may look much larger. (this is why auto-calcing often during these years is a good idea)

since this is a new league you are testing, talent will be increasing each year like clockwork for a bit. this is because the minor leagues are filled with players that are not created in the same way as amateurs in your future drafts. there is not as much depth in talent. so, HR LTM 1.000 in 2022 won't be the same as HR LTM 1.000 in 2030. (moreso than usual - if you autocalc, it is based on current talent in the league and changes the LTMs so that the current talent will on average perform to the league totals)

nothing will change automatically:

autocalc will run every year, only if you check the box. you can also run it manually, if you don't want it done all the tiem, but still would like to make use of the function.

i would suggest never changing the League Totals. if you do, change them the first year based on what you you want to center your league stats around. check out baseball-reference.com. you can see the totals of recent years, steroid years etc etc. don't blindly pick these values if you do not use defaults for fictional or mlb leagues.

if you pick them out yourself, make sure they equate to the slash line you are looking for etc... if you want .400 slugging and have 4700homeruns, your going to need a really low 2b/3b or low battingn average. if it isn't in good proporiton, i don't think it will work well. this could explain my experience with large jumps from a .001 increase in LTM.

someonen puts AB at 100,000 and that makes it very easy to set rates of 2b/3b/hr ... everything is intuitive. gotta love the decimal system! i forget who to give credit to, but it was not me. i have not tried this yet, but i intend to in ootp2017.

[Tyler87898]

Originally Posted by NoOne

CSV exports! you can export that data and link a spreadsheet to the files. then as you export each successive time, all you do it open up the spreadsheet and it will ask you to update the data.

then you can do whatever you want to them. note that they are split by conference/league. so AL/NL would need to be added together if you wanted a total. if oyu have dh/non-dh i'd recommend keeping the data separate, and totals only for the fun of it. those sets of data are two different animals and shouldn't be mingled.

i have a thread with one that is set up for "data dumps," which is a different file and location. you can use that to start or make your own. i have the dh in both leagues so it's combined.

back to LTMs

just because it says 5900 and 1.000 doesn't mean it's going to hit that number. think of the 1.000 realted to results, not hte modifier. so if the results i 20% too low, try 1.180 (a little less than 20% increase of that value, which happens to be 1 this time)

the LTM = the result. so if the result is X% different, change the LTM by roughly that amount.

the league total value just tells you how to calculate the % difference. (5900-4300) / 4300. maybe 1.300 to 1.350 ish? that's a steep incline to start. i might even go 1.200 to start out unless it is the only LTM that changes.

in this situation, again i'd sugges hitting autocalc first. your home runs could be off for reasons outside of the LTM specifically for home runs.... hope that makes sense. a ~40% increast is quite large... i'd be wary of that value.

in my experience some of them do not scale 1:1. i find teirs at certain points wear large jumps occur from .. e.g. from 1.243 to 1.244. just .001 difference can on occasion be more than you expect.

Ok this is starting to make sense, however I have another situation in my test league. All the modifiers are still 1.000 but I decreased the league total homeruns to 4900 (from 5900) but the actual homerun numbers INCREASED. How?

Also, you said the league total values are just for calculating the % difference so that you can set the modifiers? So they don't really affect results at all (though they seemingly did in my test league)?


edit: just saw your edit. So if I hit autocalc in my actual league, it will change the modifiers in a way that will get numbers closer to what the league totals are set to?

[NoOne]

Originally Posted by Tyler87898

Ok this is starting to make sense, however I have another situation in my test league. All the modifiers are still 1.000 but I decreased the league total homeruns to 4900 (from 5900) but the actual homerun numbers INCREASED. How?

Also, you said the league total values are just for calculating the % difference so that you can set the modifiers? So they don't really affect results at all (though they seemingly did in my test league)?


edit: just saw your edit. So if I hit autocalc in my actual league, it will change the modifiers in a way that will get numbers closer to what the league totals are set to?

i may have editted after you quoted, which may or may not clear things up.

the manual actually explains why that happens. it has to do with how the league totals are used. that's all outside of what we see. so from our point of view, it's better not to mess with those totals as a means to manipulate results. it's better to set them, then leave them alone.

the LTMs are just more intuitive to use when manipulatiing results.

so, they are used for more than just calculating a percentage for us to use when changing LTMs. i didn't mean they weren't used in the game engine. but their actual values aren't neccessarily important, either... with LTMs you can get what you want, regardless of the league total. if it says 5900 and you want 4900, you should just change the LTM. what you want and the value for league totals do not have to be the same... i can only assume things run smoother if the totals are in proper proportion with each other relative to what results you want to see it fluctuate aorund.... personally, i do set htem to start, and i make sure their values are in proportion to AB for the ERA and Slash values i want to see etc etc. just using the default for fictional or real-world mlb league is good enough... probably better if you don't take the time to calculate it all out in the right proportion.

yes, hitting autocalc will assess talent in league and calibrate the results to that talent. if i recall it runs like 3 seasons worth of calculations and poops out some LTMs that should match talent in league to that type of statistical output.

it's not meant to be perfect. i think it's meant to increase volatility from year to year. i think someone from ootp represented that way before in manual or maybe in forms.... i could be totally wrong on that too, to be clear. even 3 years of data has alot of error associated with it, so don't expect exactly a 4900 home run average for any short window of time.

ignorance is bliss for me on that... it was a see no evil, hear no evil moment... i'd rather not know that's required for a normal amount of fluctuation to occur... really bad sign. la-la-la can't hear la-la-la

edit: 4900 is still going to be a mid 4.00's ERA type leage, at least.... my best guess. 4.5-4.9 maybe 4.3-4.7 range year-to-year.

[Tyler87898]

So I see in another one of your threads that you aren't a fan of using autocalc every year, but it sounds like it would solve my problems, right?

edit: Maybe I should do autocalc each year, but first set all the modifiers to 1.000 before I hit the button.

[Tyler87898]

Also, I think I found a bug when selecting league totals from certain years for 2010 and 2011.

I noticed that when I try to set the league totals to the totals of the 2010 season, they are the same as 2009, and that the ones for the 2011 season are actually for 2010.