strikeout totals way too low is franchise

[nomo17k]

Re: strikeout totals way too low is franchise

Hope this gets fixed. Strikeouts are way off.

https://docs.google.com/spreadsheet/...2c&usp=sharing

There if you look in detail H/9, BB, AVE/RISP *might* be slightly inflated compared to last year's MLB stats (when offense continues to be a bit suppressed), but nothing really to worry about.

But definitely something wrong with strikeouts.....

[MrOldboy]

Re: strikeout totals way too low is franchise

Hope this gets fixed. Strikeouts are way off.

https://docs.google.com/spreadsheet/...2c&usp=sharing

There if you look in detail H/9, BB, AVE/RISP *might* be slightly inflated compared to last year's MLB stats (when offense continues to be a bit suppressed), but nothing really to worry about.

But definitely something wrong with strikeouts.....

As a question were the K/9 simmed stats in MLB 12 accurate in your opinion? Not on an individual basis as obviously Chapman and Kimbrel don't work, but the overall average.

The rest of the stats look pretty good and some variation has to be attributed to error with any model, but overall the sim engine looks pretty good. But the K/9 stats stand out big time. In my lab we'd be pretty happy with those correlations of a model to expected results.

[nomo17k]

Re: strikeout totals way too low is franchise

As a question were the K/9 simmed stats in MLB 12 close to accurate?

The rest of the stats look pretty good and some variation has to be attributed to error with any model, but overall the sim engine looks pretty good. But the K/9 stats stand out big time. In my lab we'd be pretty happy with those correlations of a model to expected results.

Yes sim engine has always been tuned very well for the most part, so it stands out that Ks are so off.... it doesn't look like a complicated fix though.

K/9 rating calibration can still be improved, I think... if you are talking about simming exceptionally good strikeout pitchers like Kimbrel. But it can also end up being funky if the game ends up doing K/9 = 99 rating for 16 K/9 in MLB stats... there aren't really many pitchers that good at whiffing, so you might end up with a roster with one 99 (Kimbrel) and the other very decent strikeout pitchers with K/9 ratings in 70s, for example.

[vcu9]

Re: strikeout totals way too low is franchise

Well, I will post some stats soon. I have fixed the K/9 Simulation Issue with my Custom Edits. I have taken SCEA's rating then added my 3-year ratings to that, then divided by 2, then -2 from that rating.

For pitchers, I have taken my 3-year ratings and added 3 rating points to those K/9 and BB/9 ratings. I plan on releasing the rosters with the full edits before the OSFM is out.

I only play one season and I do not need the OSFM. I will import a few top prospects from the player vault. If you are interested in the roster, check the roster forum.

The rest of the stats are very good. I am glad they FINALLY fixed the RISP bug.

I will post stats of the cpu sim engine from 54 games in a rivalry series between the Angels and Athletics soon.

Thanks SCEA.

[MrOldboy]

Re: strikeout totals way too low is franchise

Yes sim engine has always been tuned very well for the most part, so it stands out that Ks are so off.... it doesn't look like a complicated fix though.

K/9 rating calibration can still be improved, I think... if you are talking about simming exceptionally good strikeout pitchers like Kimbrel. But it can also end up being funky if the game ends up doing K/9 = 99 rating for 16 K/9 in MLB stats... there aren't really many pitchers that good at whiffing, so you might end up with a roster with one 99 (Kimbrel) and the other very decent strikeout pitchers with K/9 ratings in 70s, for example.

Just don't have the K/9 scale on a linear scale with the K/9 rating at the very high ratings.

Maybe just scale everything down a tiny bit so the current 0-99 scale might be on a 0-96 scale. And then the jump from 96 to 97 would be huge, like 2 K/9. So you could have 97=12K/9 , 98=14 K/9, and 99=16 K/9. Since those crazy high K pitchers are so rare I think a resolution of 2 K/9 per point is plenty to at least have a few pitchers in that K/9 range. If we could just have a few high K/9 options I'd be plenty happy.

[bautistabomb]

Re: strikeout totals way too low is franchise

Hope this gets fixed. Strikeouts are way off.

https://docs.google.com/spreadsheet/...2c&usp=sharing

There if you look in detail H/9, BB, AVE/RISP *might* be slightly inflated compared to last year's MLB stats (when offense continues to be a bit suppressed), but nothing really to worry about.

But definitely something wrong with strikeouts.....

Good work Nomo. Pretty much everything looks good apart from Ks. Offensive numbers might be slightly inflated, but nothing too extreme. Go back maybe 5 or 6 years and the offensive stats would be right on. K numbers are way off though, no matter what mlb time period you would try and compare them too. Hopefully something that can be fixed easily.

[patti72]

Re: strikeout totals way too low is franchise

I play about 70-80% of my games in franchise mode, if this doesn't get fixed it would ruin the enjoyment for me knowing that my 5 starters would finish in the top 20 in K's. I know it sounds ridiculous, but I wouldn't be able to enjoy it when I know it will be that skewed going in.

[patti72]

Re: strikeout totals way too low is franchise

Also with NOMO's stat sheet, if the K's were where the are supposed to be, overall offense might even out, just with less balls put in play. Just subtle enough to say move runs per game into the 4.3-4.4 range

[vcu9]

Re: strikeout totals way too low is franchise

Thru 67 games, the stats are great.


K/9 BB/9

Angels 7.134 3.194

Athletics 6.91 3.36

[nomo17k]

Re: strikeout totals way too low is franchise

Just don't have the K/9 scale on a linear scale with the K/9 rating at the very high ratings.

Maybe just scale everything down a tiny bit so the current 0-99 scale might be on a 0-96 scale. And then the jump from 96 to 97 would be huge, like 2 K/9. So you could have 97=12K/9 , 98=14 K/9, and 99=16 K/9. Since those crazy high K pitchers are so rare I think a resolution of 2 K/9 per point is plenty to at least have a few pitchers in that K/9 range. If we could just have a few high K/9 options I'd be plenty happy.

Yeah, something along that line could also work..... But that could be a disaster for people (i.e., many guys, including Knight) trying to figure out IRL stats -> The Show rating scales by running a lot of sim games and stuff, hahah........ it will be very time consuming to understand the correlations if the scales are not linear.

[PhillyWhat2011]

Re: strikeout totals way too low is franchise

This can't be fixed by just adjusting the sliders?

[nomo17k]

Re: strikeout totals way too low is franchise

This can't be fixed by just adjusting the sliders?

Nope... ....... and you don't want that sort of ad-hoc fix anyways....

[nascarfreak24ny]

Re: strikeout totals way too low is franchise

Read a great article on strikeouts today, they have gone up drastically in the last few years, something definitely was done wrong for franchise this year.

http://bleacherreport.com/articles/1559656-the-truth-behind-the-rise-of-the-strikeout-in-major-league-baseball

[nomo17k]

Re: strikeout totals way too low is franchise

Read a great article on strikeouts today, they have gone up drastically in the last few years, something definitely was done wrong for franchise this year.

http://bleacherreport.com/articles/1...eague-baseball

Yeah very interesting... thanks for sharing.

(I predict seanjeezy chiming in just to say that the fastball % from fangraphs has come down since the pitch f/x only recently started recognizing different types of fastballs, hahahah.....)

[Lovesports]

Re: strikeout totals way too low is franchise

Can anyone advise what the ideal roster adjustments would be? +10 on k/9 and -10 on pv sound right?