Posted by adlorenz on 1/20/2021 12:19:00 PM (view original):

Posted by 06gsp on 1/20/2021 11:58:00 AM (view original):
There are many issues with open leagues, but I don't think salary cap is one of them. There are a lot of things in need of fixing in sim league baseball that attention could be better spent on. I'd be curious to hear the rationale for why this change specifically.

We are working on some of those as well, but the question was asked because the majority of Open Leagues seem to see the same types of rosters and the salary cap seems to have been in this state for awhile (if ever changed). The adjustment is essentially to mix it up a little and is a simple as changing a number in the code. So it seems like a low effort, somewhat game changing adjustment.

If I am off base with the idea that is completely fine, it is why I asked you all as the players for your opinion.

-Adam

just curious.in addition to potentially changing the salary cap, has there been any thought to a more tiered system of experience for open leagues?

as a fairly new GM (~2400 games), i have had some success, but that's in large part to vets sending me to the forums, decision trees, and such. most new players wont do that. as such in open leagues they get matched with players that have 10's of thousands of games experience, and dozens of championships. what is the result? well they get hammered.

in a competitive hobby market, to me the key is matching players of similar skill levels, whether that's based on wins, games played, championships, etc. these tiers could then be further subdivided into 0-80 / 80-100 / +100 salary caps if wanted.

keeping a new player from getting their doors blown off does 3 things. a) it will not discourage the player from purchasing again, b)provides a fair market for the players to be rewarded, c) if there is a wins goal that leads to a higher league, it will also keep the players engaged and not abandon their teams.

my last few OL's have been very frustrating in that my teams are very good vs the newbies (.700ish); but then against very experienced players that drops to .450 or way less. its a maddening process that when joining leagues, i'm looking more at other GM's records than my own players.

just my 2 cents

Posted by Rock_Harders on 2/10/2021 12:55:00 PM (view original):

Posted by adlorenz on 1/20/2021 12:19:00 PM (view original):

Posted by 06gsp on 1/20/2021 11:58:00 AM (view original):
There are many issues with open leagues, but I don't think salary cap is one of them. There are a lot of things in need of fixing in sim league baseball that attention could be better spent on. I'd be curious to hear the rationale for why this change specifically.

We are working on some of those as well, but the question was asked because the majority of Open Leagues seem to see the same types of rosters and the salary cap seems to have been in this state for awhile (if ever changed). The adjustment is essentially to mix it up a little and is a simple as changing a number in the code. So it seems like a low effort, somewhat game changing adjustment.

If I am off base with the idea that is completely fine, it is why I asked you all as the players for your opinion.

-Adam

just curious.in addition to potentially changing the salary cap, has there been any thought to a more tiered system of experience for open leagues?

as a fairly new GM (~2400 games), i have had some success, but that's in large part to vets sending me to the forums, decision trees, and such. most new players wont do that. as such in open leagues they get matched with players that have 10's of thousands of games experience, and dozens of championships. what is the result? well they get hammered.

in a competitive hobby market, to me the key is matching players of similar skill levels, whether that's based on wins, games played, championships, etc. these tiers could then be further subdivided into 0-80 / 80-100 / +100 salary caps if wanted.

keeping a new player from getting their doors blown off does 3 things. a) it will not discourage the player from purchasing again, b)provides a fair market for the players to be rewarded, c) if there is a wins goal that leads to a higher league, it will also keep the players engaged and not abandon their teams.

my last few OL's have been very frustrating in that my teams are very good vs the newbies (.700ish); but then against very experienced players that drops to .450 or way less. its a maddening process that when joining leagues, i'm looking more at other GM's records than my own players.

just my 2 cents

There's definitely been talk of this, it's been brought up a number of times in the past several pages. I think there are some potential benefits, and you brought a few of them out. I specifically like "c" - to have a wins goal that brings them up to a higher league, and how that might keep people motivated to stick with an underperforming team to squeeze out every win they can. It's annoying to me to play in a league where several owners have clearly mailed it in. It really upsets the competitive balance and distorts the stats.

The biggest problem is that by having several varieties of "open" leagues is that it makes each of them fill that much more slowly. Is a new player going to sit patiently by for several days (or more) waiting on the league to fill? I think we risk losing people before the season ever starts if they have to wait that long. I could see MAYBE one more option for open leagues specifically designed for beginners, but even that is questionable. I can see some veterans choosing to make new accounts just so they could play in beginner leagues, negating the whole purpose. There's already a champs league, designed for veterans of the site, but it's completely voluntary if a veteran joins that league or the open league. It might even out the balance if people over a certain experience level were automatically put in champs leagues instead of open leagues, but I imagine that would upset some veterans as well. Attracting new people and keeping veterans happy are both important - to which side do we cater? Perhaps if the site grows, there will be room for more leagues.

Posted by Rock_Harders on 2/10/2021 12:55:00 PM (view original):

Posted by adlorenz on 1/20/2021 12:19:00 PM (view original):

Posted by 06gsp on 1/20/2021 11:58:00 AM (view original):
There are many issues with open leagues, but I don't think salary cap is one of them. There are a lot of things in need of fixing in sim league baseball that attention could be better spent on. I'd be curious to hear the rationale for why this change specifically.

We are working on some of those as well, but the question was asked because the majority of Open Leagues seem to see the same types of rosters and the salary cap seems to have been in this state for awhile (if ever changed). The adjustment is essentially to mix it up a little and is a simple as changing a number in the code. So it seems like a low effort, somewhat game changing adjustment.

If I am off base with the idea that is completely fine, it is why I asked you all as the players for your opinion.

-Adam

just curious.in addition to potentially changing the salary cap, has there been any thought to a more tiered system of experience for open leagues?

as a fairly new GM (~2400 games), i have had some success, but that's in large part to vets sending me to the forums, decision trees, and such. most new players wont do that. as such in open leagues they get matched with players that have 10's of thousands of games experience, and dozens of championships. what is the result? well they get hammered.

in a competitive hobby market, to me the key is matching players of similar skill levels, whether that's based on wins, games played, championships, etc. these tiers could then be further subdivided into 0-80 / 80-100 / +100 salary caps if wanted.

keeping a new player from getting their doors blown off does 3 things. a) it will not discourage the player from purchasing again, b)provides a fair market for the players to be rewarded, c) if there is a wins goal that leads to a higher league, it will also keep the players engaged and not abandon their teams.

my last few OL's have been very frustrating in that my teams are very good vs the newbies (.700ish); but then against very experienced players that drops to .450 or way less. its a maddening process that when joining leagues, i'm looking more at other GM's records than my own players.

just my 2 cents

I addressed this earlier in this thread: https://www.whatifsports.com/forums/Posts.aspx?topicID=525261&threadID=12024285#l_12024285 and have quoted the post in full below because I think it helps explain some of the thought processes.

I think people forget what makes OL great. It's how fast they fill. You can draft a team and have it start the next day. This is why people play OLs more than anything else. Tiered OLs won't work because even a relatively low-threshold entirely changes this dynamic. For example, the threshold for Champs Leagues is just 5 WS appearances, not titles. An owner with 0 titles can play in Champs leagues, plus they have increased rewards, but even at their fastest it takes 3-4 weeks to fill a champs league, and there was a gap of 8 years between champs leagues filling at one point. Making a multi-tiered OL system would only compound this problem. People want their teams to play, not sit in leagues waiting weeks, months, or years to fill, as would be the case if this already existing tiering is expanded on in any fashion.

Next, we all want new owners to stick around, and I think it's somewhat unfair to attack owners (especially people like dn8779) as being predatory. There's a lot that goes into each OL and they're all unique. An identical team could win 120 in one league and 85 in the next. The biggest advantage over new owners is knowledge of sim workings, not IDing "cookie" players. There's a bunch of ways to win. I had a team with a winning record with an $8m pitching staff, made the playoffs with a 2003 Mariners twist team but only using other Mariner seasons, and won ~90 games with single season teams from 2019 and 2020. This has nothing to do with cookies or preying on new owners, but strictly from understanding how the sim works better than new owners do.

Sure, some players, like '02 Bernhard or '19 Marte, show up on a bunch of teams, but this just means you have an idea of what your opponents are using and gives you an advantage in trying to beat them. These guys aren't invincible or unbeatable. Drafting them doesn't guarantee a 100 win season and a WS title. Even if a new owner drafted them, they still aren't as likely to succeed.

Dynamic pricing doesn't change this balance either, owners who understand the sim better will be able to find "bargains" or "value" and still build teams that succeed just as frequently, but it keeps things fresh. New owners will still have the knowledge gap issue with dynamic pricing in place. That knowledge gap would increase if new owners were limited to leagues only with other new owners. Most OLs I'm in have frequent dialogues in the league boards from veteran owners offering tips and advice to newer owners. This feedback is way more valuable to newer owners than them waiting a week or more for the league to fill and then winning 85-90 games (and there will still be owners in there that lose 100+).

The change to recommended IP/PA and the addition of links to the FAQ/Advice for Newbies will go much further than any of the above proposed changes.

The only proposed change that might have an impact on retaining new owners, or at least helping close the knowledge gap faster, would be removing fake names for AAA (and the normalization to current season). Everything else proposed here doesn't even address the root issue, just a perception problem, or would create bigger problems (leagues never filling and owners dropping because they never get to see their teams play) than the ones we already have.

i have had some success, but that's in large part to vets sending me to the forums, decision trees, and such.

You credit the owners who have led you here and giving you advice for part of your success, but creating a tiered system that would separate new owners from veteran owners would prevent that from happening.

There is absolutely more that this site can do to help new owners (many have been posted over the 8 pages of this thread), but I do not believe separate leagues for new owners and veteran owners falls into that category.

I, like most here, became a better owner from learning from those who were better than me. This isn’t a “if I had to suffer then you have to suffer” take, but a “learn from those around you” take.

You mention that you focus more on how many wins/championships other owners have rather than your own team. Don’t.

focus on what their teams are doing differently than yours. It’s what I do. If my team falls short of my expectations, I look at the other teams and try to decide if I fell short because they drafted better than me, or if maybe I just had bad luck.

Posted by chargingryno on 2/10/2021 6:04:00 PM (view original):

i have had some success, but that's in large part to vets sending me to the forums, decision trees, and such.

You credit the owners who have led you here and giving you advice for part of your success, but creating a tiered system that would separate new owners from veteran owners would prevent that from happening.

There is absolutely more that this site can do to help new owners (many have been posted over the 8 pages of this thread), but I do not believe separate leagues for new owners and veteran owners falls into that category.

I, like most here, became a better owner from learning from those who were better than me. This isn’t a “if I had to suffer then you have to suffer” take, but a “learn from those around you” take.

You mention that you focus more on how many wins/championships other owners have rather than your own team. Don’t.

focus on what their teams are doing differently than yours. It’s what I do. If my team falls short of my expectations, I look at the other teams and try to decide if I fell short because they drafted better than me, or if maybe I just had bad luck.

+1

I started off just using raw BA and raw ERA drafting players.

Then after hearing some good owners talk about + and # stats, I incorporated those into my searches. I got a little better, but never had a top team in the league. Mostly used BA+ and WHIP+. That’s when I frustratingly asked why my teams weren’t doing better.

If I was in a tiered league, I would’ve never found out why my teams didn’t work. Probably would’ve gotten a lot of unhelpful answers based off superstition or theory.

Instead I got answers that opened up a whole new realm of possibility for my teams.

Leading the league in BA, but mediocre in runs scored? I drafted a team severely lacking speed, causing me to lead the league in GIDP by far.

Why did you draft this Jim Bottomley when that Jim Bottomley had a much higher BA? This Jim Bottomley hits way more XBHs, causing more runners to score from 1st.

Why is this pitcher’s ERA so much better than mine? Well mine has a HR/9+ of 89 (or whatever that one Marichal season’s is cuz that used to be a go-to for me) and isn’t in a ballpark to help prevent him giving up bombs.

For all my rants that no doubt annoyed many owners, that is how I got good at the game, and I have many thanks for those owners that explained why I still sucked at the game.

I'm writing this as someone with no coding knowledge but with decades of experience in this sim as well as other sports simulation games. I voted for the $80M cap because I think it makes a good entry point for new owners and for teams to be slightly better than RL without all of the statistical changes that occur with all-star teams at higher caps.

Overall, I think there are six major changes that need to happen in the sim. That said, teams that have essentially all-star rosters shouldn't be expected to have stats comparable to RL for the reasons stated previously. But, the drivers of strong performance in the sim should mirror RL. It's the distortions in strategy from what works IRL that cause many of the issues noted in this thread and elsewhere.

1) Normalization across eras, particularly for pitchers, needs to be addressed. 2018 Aaron Nola (a seldom-used season) has the most comparable OAV+, ERA+, and HR/9+ to cookie 1902 Bill Bernhard with at least 200 IP/162 and is comparably priced per inning. Bernhard is a cookie because Bernhard shuts off a lot of power and Nola doesn't. Yes, fielders will make more errors behind Bernhard but the tradeoff in SLG is likely more significant. This is why all-time great pitchers like Grove, Seaver, Spahn, Palmer, Carlton, and Marichal are rarely seen on rosters outside of progressive leagues. Either price Bernhard and the other dead ball cookies appropriately to reflect their ability to shut down power or fix the normalization such that modern pitchers are as valuable as the deadballers.

2) Good offensive strategy IRL should equal good offensive strategy in the sim. Great modern offensive players with relatively low 1B/100# such as Joe Morgan, Mike Schmidt, Mel Ott, Reggie Jackson, and Barry Bonds tend to underperform significantly in the sim because of #1 above and the impact of their 1B/100# numbers. Since 1920, power and OBP have been the drivers of strong offensive performance IRL. Yet, successful sim teams are almost never built on this strategy.

3) The first step in the decision tree for the interaction between pitchers and hitters should be whether the ball is put in play. As has been demonstrated IRL, there is is less variability in BABIP between pitchers. But, the pitchers who strike out more hitters tend to be more valuable (all else being equal) because they put less pressure on their defense.

4) Range factor is far too important to success in the sim. Defensive range IRL is extremely situation dependent. One of the best examples of this is Richie Ashburn's range factor playing behind Robin Roberts, who pitched a ton of innings and gave up a lot of fly ball outs while playing home games in Connie Mack Stadium, which had a huge center field. However, the sim treats Richie Ashburn as having more impact defensively than Willie Mays. The strategy of drafting A+++ range fielders and playing them in parks that shut off HR power has no basis for success IRL because those players tend not to have strong power stats. Contrary to what one might think when looking at the sim, 1885 Roger Connor isn't the greatest season ever by a first baseman.

5) Building on the point about Connor, I would suggest getting rid of the 1885-1892 players because the mound wasn't even at 60 feet 6 inches from home plate. Baseball before 1903 was already very different than what we are used to seeing today. The 1885-1892 period would be much less recognizable and houses many of the distortions that can cause issues with the sim. I realize that this probably has very little chance of happening.

6) Lastly, fatigue needs to be addressed. At higher caps, an owner can save a pitcher and use all of his IP over a relatively small portion of the season to distort a pennant race. For example, there is zero chance that 2016 Clayton Kershaw could start eight consecutive games, throw 75-100 pitches every time and still be at 100% at the start of every game. Yet, I saw this happen in a league. At lower caps, this can happen with pitchers like Milacki, Smyly, B. Adams, McDowell, and Verlander. New owners won't expect this and will be frustrated that such ridiculous strategies can work.

One of the reasons deadball pitchers do so well is because they have the advantage on HRs allowed, BUT there are not penalized for allowing more triples than modern day pitchers. THEY SHOULD BE ALLOWING MORE TRIPLES.

The modern day triples hitters get a huge bump in triples because of normalization Conversely, the deadball hitters never get close to their real life triples totals, again due to normalization. Similarly, deadball HR hitters get a bump in HRs while modern day HR hitters always underachieve.

Here's the thing. If you used modern day HR hitters with modern day pitchers, in a low cap, these HR hitters would actually hit HRs. BUT, doing the same thing with deadball triples hitters DOES NOT WORK. That's because deadball pitchers DO NOT GIVE UP MORE TRIPLES than modern day pitchers.

If you ran a 80M draft league using just players from 1911 (a big year for triples), you will find that no player will come close to their triples totals (I know, because I have done this).

Posted by schwarze on 2/12/2021 9:17:00 AM (view original):
One of the reasons deadball pitchers do so well is because they have the advantage on HRs allowed, BUT there are not penalized for allowing more triples than modern day pitchers. THEY SHOULD BE ALLOWING MORE TRIPLES.

The modern day triples hitters get a huge bump in triples because of normalization Conversely, the deadball hitters never get close to their real life triples totals, again due to normalization. Similarly, deadball HR hitters get a bump in HRs while modern day HR hitters always underachieve.

Here's the thing. If you used modern day HR hitters with modern day pitchers, in a low cap, these HR hitters would actually hit HRs. BUT, doing the same thing with deadball triples hitters DOES NOT WORK. That's because deadball pitchers DO NOT GIVE UP MORE TRIPLES than modern day pitchers.

If you ran a 80M draft league using just players from 1911 (a big year for triples), you will find that no player will come close to their triples totals (I know, because I have done this).

+1

The deadball pitchers get the benefit of not allowing HRs, but them also get the benefit of giving up fewer XBH (2B & 3B) due to those hits being impacted by ERA, which was lower in the deadball era primarily because of the errors committed behind them. They gave up more 2B/3B on a rate basis than pitchers allow today. Along with the increase in errors, the increase in 2B/3B should compensate for the slugging difference gained/lost with HR allowed rates.

But this is inverted in the sim, so deadball pitchers benefit on both sides. As has been posted previously in here, the current TWISL is running testing the value of deadball vs modern pitchers and through the first week (around 20 games), this same effect is being borne out. At a similar $/IP, the OAV/WHIP when factored for errors allowed are almost identical, but the XBH/HR allowance swings the value in favor of deadball pitchers because of allowing fewer of these types of hits they allow fewer runs overall. The error rate closes the gap slightly, but there remains a gap roughly proportional to the XBH/HR rate. I have more details in the thread tracking the league in the Theme League forums.

Correcting the XBH rates would go a long way toward correctly bridging value across eras as well as bringing more realism. Until then, at least, raw HR/9 and raw ERA should weigh a little higher in the salary formula.

low strikeout rates for deadballers is a problem too, I've had Rube Waddell in a progressive for 4 years starting in 1901 and his K results are as follows. (The league has a randomness element of which season you get which is why I've had his real life 1903 season twice)

1901: 64 Ks in 285 IP (real life 168 Ks in 243.2 IP)
1902: 152 Ks in 461 IP (real life 302 Ks in 324 IP)
1903: 118 Ks in 468 IP (real life 349 Ks in 384 IP)
1904: 146 Ks in 443 IP (real life 302 Ks in 324 IP)

the advantage to this is I've been able to use him for more innings because of the lack of Ks, but he's giving up way more unearned runs than he should because the fielding in this era is so bad. his K rate is about a third of what it should be over a huge sample size, and I haven't been using him when fatigued. makes me a bit skeptical of the decision tree to see someone so far off real life stats in a progressive.

The strikeout problem is real...here's a post with some more data - drawn largely from the same league 06gsp is talking about:
Strikeouts

I haven't tested the deadball strikeout issue, but I think that is more on the hitter side. They (mostly) kept track of pitcher strikeouts back then, but rarely kept track of hitter strikeouts. So, while the pitcher values are correct, the hitter values are significantly understated. I would guess if we ran the deadball hitters against modern pitchers they would also put up absurdly low K #s. I would also guess if we ran the deadball pitchers against modern hitters, we'd see numbers closely in-line with their K# numbers. In some cases, the hitter Ks were entirely estimated (similar to CS rates). I'd love to test this further, but I'd bet that the hitters are where this issue lies. I'd guess if that proves out, admin could utilize the pitcher strikeout records per year and based on recorded hitter rates, assign the difference in Ks back to those hitters for more realistic K numbers from those early hitter seasons.

There are some favorite ways to win, and many bargain players that are used, but as many have said, there are different ways to win - and win with players that aren't the same all the time. Don't get me wrong, I have an unhealthy obsession with Gavvy Cravath, but I also like to find different ways to win to be challenged. I have a team that is 22-16 in an open league, and it was made exclusively from player seasons that had 15 or less usages at the time. I didn't expect '71 Gary Nolan to be 7-0 with a 2.73 era in 10 starts. And I'm definitely overpaying for 1921 Austin McHenry, but these are the things that you get to discover in open leagues. Sometimes I like to put together a team that I hope will get lucky and win 100 games instead of 80, and sometimes I like knowing that open leagues can give seasoned drafters a chance to do something different. And the newer players benefit from the vets who do so. Just my 2 cents.

Did the same thing when I played my 1 OL two years ago. Although I only did that for hitters cuz I’m too lazy to find pitchers.

I actually have the best 1910 John Knight in the performance history

Posted by d_rock97 on 2/12/2021 4:31:00 PM (view original):
Did the same thing when I played my 1 OL two years ago. Although I only did that for hitters cuz I’m too lazy to find pitchers.

I actually have the best 1910 John Knight in the performance history

I used Frank Baker’s 1910 recently, at the time only had two uses (50rc/73rc respectively). He ended my season with 122rc, and is now one of my favorite 3B’s to use