[Neags23]

Originally Posted by RonCo

I can't find many pitchers who actually survive well into their 40s in my baseline leaue, either. So, if you wanted to BZ MArkus and have him tone down the Stuff falloff at age 38, that would probably be good.

RonCo, I haven't had a chance to thumb through your report, but this statement made me go find my results.

This is something I posted in the beta forums regarding K's for pitchers. It's the same methodology Tango uses with adjacent pitching years (1 being the peak year). The only difference we have here is that for OOTP, I only used pitchers with greater than 60 IP (and I use the stuff rating while Tango uses the real thing). The n variable is the number of players tested over a 100 year test league (first 20 years being skipped).

From all of my findings based on Tango tests, the stuff falloff (league-wide, of course... I'm sure outliers abound) is actually not steep enough.

Those tests do, obviously, stop at age 40, but if you use Tango as a baseline, from about age 31, stuff ratings don't fall fast enough (but aren't too bad).

Not saying I'm right and you're wrong, I'm just curious as to why we may have opposite results.

[RonCo]

Interesting work, Neags. I'm not sure of all your methodology at its detail, but the thought sounds right enough. Not sure you really need to drop the < 60 IP players, you're really just trying to measure human beings, and the more the merrier. I'm running the same scripts I ran in beta last year. And I'll say that every test league I run shows pitcher's careers on the whole following the path that says K's fade quickly after age 37 when on default settings.

I've been doing more work on some AI settings that is looking positive.

[Neags23]

I forget why we dropped the guys with less than 60 IP. Injury Log's suggestion and guidance led me to do so. I'm sure there was a good reason, even if I can't remember (the beta process really takes a toll on your brain).


What baseline are you using for comparisons that show stuff is falling too fast?

[RonCo]

Originally Posted by SkyDog

That would require me reading that 30-page treatise you wrote. Not likely to happen.

Good results so far with 35 Ratings, 60 Current, 5 Previous. In the cases I've checked so far, the older guys are pitching maybe a half season with horrid stats, then either getting released, retiring, or demoted.

I've had better overall results with 60 Ratings, 30 Current, 8 Previous, 2 2-yrs ago. Both are improvements, though.

[RonCo]

Originally Posted by Neags23

I forget why we dropped the guys with less than 60 IP. Injury Log's suggestion and guidance led me to do so. I'm sure there was a good reason, even if I can't remember (the beta process really takes a toll on your brain).


What baseline are you using for comparisons that show stuff is falling too fast?

My expected baseline is Tango's--same as yours.

Then for each test league I create a ratings chain for each player in each ability. Then I average those chains for all major leaguers and for all players in the universe. I've presented those two groups. I tried to describe that in my paper, though I probably screwed it up somewhere.

I'm finding it hard to get many HoF pitchers, BTW. And when I pull up the stats page in game of pitchers with long careers, I don't generally need to have the ages next to the stats line to determine when the pitcher turned 38. So everything I see in every league I run says that pitcher's K-rates fall off too hard after age 38.

[RonCo]

Originally Posted by Neags23

I forget why we dropped the guys with less than 60 IP. Injury Log's suggestion and guidance led me to do so. I'm sure there was a good reason, even if I can't remember

I assume it's to attempt to model Tango's selection criteria, but doing that misses the entire point of the exercise. IL and I have gone around on this about a hundred times, though, and I really have zero energy left for doing it again. He's still a beta, and I'm not.

[SkyDog]

Originally Posted by RonCo

I've had better overall results with 60 Ratings, 30 Current, 8 Previous, 2 2-yrs ago. Both are improvements, though.

Scouting on or off? I'd like to try playing with scouts on, which is why I fear going that high on ratings.

[RonCo]

Scouting off on my test league. Though from what little I see scouts are pretty accurate. I would be surprised if there was too much difference. I'm no scouting expert, though.

[RonCo]

I still see a pitcher's Ks falling too fast after 37-38, but at either of these AI settings, GMs aren't letting most of them pitch out into their 44-45 yo seasons. The 60/30/8/2 numbers are a little better, though...I think. Improves average career length another tenth of a year or so, too.

[SkyDog]

Originally Posted by RonCo

Scouting off on my test league. Though from what little I see scouts are pretty accurate. I would be surprised if there was too much difference. I'm no scouting expert, though.

Well, I'll run a 100-year test overnight to see how 60-30-8-2 looks.

[RonCo]

On the other hand, I think the MLB HoF consists of about 40% pitchers. My test leagues are routinely hitter-dominated...this last one is 52 hitters, 14 pitchers, or 21% pitchers. Some of this could be the criteria, I suppose (which is default), but I'm guessing that not enough pitchers are playing long careers because they lose their Ks too quickly. Just guessing, of course, but usually when things line up like this there's fire with the smoke.

[ashantewarrier]

RonCo, your effort here is tremendous. Are these settings appropriate for historical fictional sims (historical simming with era-based settings, fictional players)?

[Eugene Church]

RonCo, thanks for sharing your great research with the OOTPB Community.

You really helped to make the game better and more realistic with your studies on the Beta Team.

[RonCo]

Originally Posted by ashantewarrier

RonCo, your effort here is tremendous. Are these settings appropriate for historical fictional sims (historical simming with era-based settings, fictional players)?

Thanks.

I've not done any testing with historicals on these settings, but it seems logical to me that they would help any league that wants proper age peaks. If I get a chance next weekend I may play with historical/fictionals a little just to see.

[RonCo]

Originally Posted by Eugene Church

RonCo, thanks for sharing your great research with the OOTPB Community.

You really helped to make the game better and more realistic with your studies on the Beta Team.

Thanks, Eugene! I wish the beta thing would have worked out this pass through, but it was better this way.

[NomarHits400]

Originally Posted by RonCo

But beyond the GM AI issues, pitchers in OOTP are just overall not well-designed. Very few pitchers should be able to stay in the league if they can't maintain a K/9 ratio over 4.5. But OOTP makes Control and Movement too good across a career (again, we've been able to see this on the Tango charts for quite some time), so you've always been able to find quite a few good pitchers who can't strike anyone out in OOTP leagues. A few is great, mind you...but...well, you get the idea. The problem here is not that Ks drop too fast, but that HRs don't drop fast enough and BB are just too good across the spectrum.

Here is a chart with ERA vs. K/9 plotted in one of my test leagues (with aging/development adjustments). You can see a lot of pitchers with these gosh-awful K/9 ratios are pretty successful when measured by ERA. 43 of the 73 pitchers involved in this chart are 30 yo or younger. Of those 43, 20 have ERA under 5.00.

Fascinating work, RonCo- and I agree with your findings- but I don't know that a pitcher with a sub 5.00 ERA can be considered a find. I guess my point is, how many of those <5.00 were also below 4.50- because it looks like if 4.50 were the criteria that the 20 might half itself again. I just don't think a guy with 4.85 ERA is all that useful- but I certainly agree that OOTP has more of these guys floating around than in real life.

It certainly seems like a big part of this problem is the AI's proclivity to continue to give these guys innings, which can be marginally improved with the AI settings.

[GBS]

So RonCo, I just started a new fictional league. Do you suggest that I try the modifiers mentioned at the top of this thread, to add realism of course?

[RonCo]

Sub-5.00 ERAs are not really "finds", but if the league average ERA is 4.70-4.90 or so (semi-modern-era AL numbers, and structured like my test league was), a 4.85 is not a terrible pitcher, and can be considered successful, especially if he's a starter.

[NomarHits400]

Originally Posted by RonCo

I've been fiddling with aging and development modifiers for fictional leagues. I've pretty much settled on:

Batter Aging: 2.00
Batter Dev: .5
Pitcher Aging: 1.75
Pitcher Dev: .5

If I get some time I'll post a bunch of supporting data, but I warn that I'm really busy during the week.

I am a little confused by this. Setting the aging modifiers as above will cause players to age must faster, corect? This in turn, of course- will cause ratings to deteriorate quicker as well, correct? I assume that this is what accounts for the markedly increased attrition rates of a pitcher's stuff and therefore K rates (among other things)?

It also seems that with development speeds halved, that the AI may be forced to play some of these God-awful older players in the absence of good young talent coming up through the ranks as fast as normal. This seems problematic on the face of it to me.

Are any of my assumptions above logical? I assume these aggressive settings were put in place because players were peaking too soon and/or playing too long...Is that correct?

[NomarHits400]

Originally Posted by RonCo

Sub-5.00 ERAs are not really "finds", but if the league average ERA is 4.70-4.90 or so (semi-modern-era AL numbers, and structured like my test league was), a 4.85 is not a terrible pitcher, and can be considered successful, especially if he's a starter.

Point taken...I understand- I am comforted a bit that at least MOST of these guys are barely replacement level, though.