Stats-Based CPU Sliders [MLB 13 Version]

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 13 Version]

Sounds good. I understand the decision to make the CPU throw more strikes. Everyone is a "free swinging Soriano", as Matty V. would say.

It's early but I played around with strike frequency a tiny bit. It might be a little more volatile than your testing with the pitcher control slider last year. Walks up but batting average up as well. I'll let you be the judge.

I'm keeping track of 1st pitch strike % so that might help as well.

[Jimbo614]

Re: Stats-Based CPU Sliders [MLB 13 Version]

Hey Nomo...
I'm going with No DH and using the Polo Grounds, Crosley, Sportsman, Griffith and Shibe in my CPU vs CPU Franchises.
How might that affect my results?
Also, what do you think about carrying over last years sliders?

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 13 Version]

Hey Nomo...
I'm going with No DH and using the Polo Grounds, Crosley, Sportsman, Griffith and Shibe in my CPU vs CPU Franchises.
How might that affect my results?
Also, what do you think about carrying over last years sliders?

I think it would be very interesting to use those funky stadiums!! They probably have very different park factors compared to more "standard" ball parks (I'd expect more cheap HRs down the line, crazy number of triples and inside-the-park HRs in Polo, for example), but that's what makes things interesting... so long as you won't be "testing" numbers against the recent MLB average, I don't see any issues there...

In general, carrying over last year's sliders won't give you similar results, since how sliders affect the game may not be the same from one year to the next...

... However, some portion of the gameplay has been quite stable the past couple years, so I would actually expect the game to play fairly similarly even if you used sliders from MLB 12. But SCEA always make adjustments... for example, last year I say slightly less fielding errors compared to throwing errors. I increased Fielding Error slider by one or two, but this year they seem to be evened out at the default....

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 13 Version]

More than 2/3 done for the initial 75-game set with sliders at default (except Pitch Speed 10), I'm thinking of what to change next.

Comparing to 2010 - 2012 MLB average, the tentative observations are:

(1) walks are quite low

(2) doubles and triples are slightly low.

(3) strikeouts and swing & miss % is slightly high.

(4) fouls are slightly low.

(5) stolen base success rate is slightly low, thought this is one of the harder one to get right since steals are situation dependent (need much larger sample size), and some CSs are definitely coming from failed hit-and-run attempts.

(6) fielding errors are slightly low.

Even with those deviations, I like how the games are playing in general. Very tight games in general, but there are some blowouts here and there.

I still maintain that the only adjustment that should really be made is to adjust for the loss of walks, but I'm hoping to adjust for others as well.

[steviegolfballs]

Re: Stats-Based CPU Sliders [MLB 13 Version]

Hi Nomo! Glad to fid you here again!

Something that might interest you, I also ran games with unadjusted sliders and got generally the same results as you are seeing (I ran 50 games). After that I moved pitcher control to 4 for 30 games and the walk and offensive explosion was unacceptable. For my next 30 games I returned control to 5 and dropped strike frequency to 4. The results were much better! Walks, strikeouts, runs, doubles and home runs were very near average. Hopefully this helps when you finish your initial run. I am on the road right now and the data is back home, I can supply in a few days if you wish.

Steve

[vcu9]

Re: Stats-Based CPU Sliders [MLB 13 Version]

Hey Nomo.

Do you think that Starter Stamina at 5 is too high? It seems like a pitcher with a stamina rating at 90 or above can throw 140 pitches before his energy bar totally disappears. Do you think Starter Stamina at 4 is better? And how about Reliever Stamina at 2? I think that is best I look forward to your input.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 13 Version]

BR Speed Slider Test

I've done this before and it looks like things haven't changed much in how BR Speed slider affects the speed of baserunners.

I recorded the players running to 1B on bunt, starting from the frame when the contact was made and stopping the frame when the runner touched the 1B bag. All RHBs, with varying Speed ratings. The measurements table is here for those curious.

The default (5) seems to reproduce realistic speed, in that the average player (Speed rating of 50) has 4.2 sec from the bat contact to 1B, which is the MLB average.

I'm not sure if the game captures the real dynamic range of MLB baserunners. Does the fastest RHB still needs 4.0 sec to reach 1B? A table like this one seems to say this is reasonable, but I often hear a report saying some LHBs are capable of reaching 1B within below 4 second.

[vcu9]

Re: Stats-Based CPU Sliders [MLB 13 Version]

BR Speed Slider Test

I've done this before and it looks like things haven't changed much in how BR Speed slider affects the speed of baserunners.

I recorded the players running to 1B on bunt, starting from the frame when the contact was made and stopping the frame when the runner touched the 1B bag. All RHBs, with varying Speed ratings. The measurements table is here for those curious.

The default (5) seems to reproduce realistic speed, in that the average player (Speed rating of 50) has 4.2 sec from the bat contact to 1B, which is the MLB average.

I'm not sure if the game captures the real dynamic range of MLB baserunners. Does the fastest RHB still needs 4.0 sec to reach 1B? A table like this one seems to say this is reasonable, but I often hear a report saying some LHBs are capable of reaching 1B within below 4 second.

Great Fact Nomo!!

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 13 Version]

Hi Nomo! Glad to fid you here again!

Something that might interest you, I also ran games with unadjusted sliders and got generally the same results as you are seeing (I ran 50 games). After that I moved pitcher control to 4 for 30 games and the walk and offensive explosion was unacceptable. For my next 30 games I returned control to 5 and dropped strike frequency to 4. The results were much better! Walks, strikeouts, runs, doubles and home runs were very near average. Hopefully this helps when you finish your initial run. I am on the road right now and the data is back home, I can supply in a few days if you wish.

Steve

Yeah I'm following all the other related threads and seeing what people are getting. I do run a lot of games myself but some stats are very hard to nail down, so I'm always curious what others are seeing. Please feel free to post here or in other threads... I'm all for sharing useful info.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 13 Version]

Hey Nomo.

Do you think that Starter Stamina at 5 is too high? It seems like a pitcher with a stamina rating at 90 or above can throw 140 pitches before his energy bar totally disappears. Do you think Starter Stamina at 4 is better? And how about Reliever Stamina at 2? I think that is best I look forward to your input.

When CPU leaves a pitcher even when he's over 120 - 130 for the pitch count, it's likely the hard-coded manager decision is at work... typically this happens when CPU is pitching with a huge lead. I don't think there's any way to override this with sliders. It's one of those things we need to hope that AI improvement will take care of at some point.

Personally I haven't found any reason to change stamina or manager hook related sliders. Starters pitch, on average, 5.87 inning in game (5.95 in MLB), and on average 3.96 pitchers are used (3.90 pitchers in MLB). The numbers that I'm tracking seem pretty decent to me.

[ralphieboy11]

Re: Stats-Based CPU Sliders [MLB 13 Version]

Hi Nomo! Glad to fid you here again!

Something that might interest you, I also ran games with unadjusted sliders and got generally the same results as you are seeing (I ran 50 games). After that I moved pitcher control to 4 for 30 games and the walk and offensive explosion was unacceptable. For my next 30 games I returned control to 5 and dropped strike frequency to 4. The results were much better! Walks, strikeouts, runs, doubles and home runs were very near average. Hopefully this helps when you finish your initial run. I am on the road right now and the data is back home, I can supply in a few days if you wish.

Steve

That's interesting. It really takes quite a few games to getting a good working average it seems. I really didn't see the same thing with pitcher control. Through 34 games with it at 4, I didn't see enough of a change for me. Still too many strikes, not enough walks. Batting average was at .254 and walks were only 2 a game.

While I kept those stats and may go back to that later, I tried pitcher strike frequency at 3 and foul frequency at 6 for a few games. Through 36 games, batting average was at .265, walks at 3.38 although I had a few extra innings games that probably bumped that up. XBHs were looking great at this setting though compared to default.

I do think I'm leaning toward strike frequency being the main ingredient to getting more acceptable walks and pitches per AB. 3 might be a bit extreme, but 4 might be the ticket for me as well.

[jts88]

Re: Stats-Based CPU Sliders [MLB 13 Version]

Hey guys, just wanted to throw in my two cents. I'm a long time player of the Show and have slowly converted over to full MoM. I've played roughly 50 games with very few adjustments. I noticed the sliders seem to me to be more sensitive this year than in years past.
I'm not one for tracking data, so this isn't fact. Just observation.
I really struggled out of the box. I was getting way too much offense for 10 games but things seemed to straighten out after that.
These are some adjustments i made.
Default except for the following
Power - 4. May bump back to 5
Strike Freq - 4. I tried 3, too many deep counts
Pitch Control - started at 6, 5 seems perfect.
Fieder Reaction - 6. Too many solid hits getting into the gaps. I tried lowering solid hits, but that caused weaker gameplay all around.
Fielder Arm - 4. Increases 1st to 3rd and 2nd to home for baserunners. I was worried about too many infield hits, but reaction increase seems to counter that.
Anyway, just wanted to share. Thank you, guys, for all you do. This game and your hard work gives an old guy like me a lot of enjoyment on my spare time. Looking forward to your findings and to Knight and Co. rosters.

[rdklein]

Re: Stats-Based CPU Sliders [MLB 13 Version]

First, Nomo... I just want to give you a big thanks.

I have been following your posts for three years, and I think I have only commented once on how much I appreciate the information you are willing to share. I'm past due.... seems I'm one of the many who reads, yet doesn't post. I love Operation Sports, and this is the first place I go for sliders whether it be for NBA 2K, Madden, or NCAA... however, I think I've made maybe 10 posts in the last 3 years. I don't post my sliders because, well, I think most have their own preferences.

However, you, sir... make a great baseball game SO much better. There are some FANTASTIC posters in this forum, but nobody in these forums breaks a game down quite like you.

If there is someone from a sports gaming company who is reading my post? You need to hire Nomo17k for "Quality Control - sp. statistical anomolies". (EA hasn't had a person at this position for the past 10 years. Hire Nomo EA!)

My 2 cents:

[quote=jts88;2044823980]Strike Freq - 4. I tied 3, too many deep counts
Pitch Control - started at 6, 5 seems perfect.

this.

I know you are probably next going to focus on Ball/Stike ratio in the next round of tests. I tried 3 also, but I experience what jts88 experienced. Then I went with Control 4, Consistency 4, Strike Freq 4 : still too many walks. (had 10 the first game I simmed).

Currently, I have PControl = 5, PConsist = 4 and Strike Freq = 4.

After 10 games, I am only slightly higher in walks compare to mlb avg. I don't know how these affect the other stats... I work an average of 50 hours a week, I'm married, so i just don't have time to track it. But, I hope my info helps with your next test.

I love the spreadsheet... and for the guys out there who just read and don't always post their graditude...

Thanks Nomo. You're a pro.

[floydpinkster]

Re: Stats-Based CPU Sliders [MLB 13 Version]

rdklein is right... We talked a bit last year but i should also give props and appretiation to you nomo... I had 12 months of great cpu vs cpu fun thanks to your Show 12 sliders. Looking forward to what you come up with this year. I know sliders can't fix the game or make it perfect, but your sliders get The Show as close as can be!!!
Thanks Bud

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 13 Version]

Thanks for the kind words guys.... I think we are just lucky that baseball is a type of sport in which stats are one of more fascinating aspect of the game, and because of that we have more data and tools to work with them to understand the game better....

I haven't developed an appreciation for other sports like football (though a coworker of mine is an avid football fan and have been preaching me as to how strategically interesting the game really is.....), so cannot say for other games but I'm pretty sure they try what they can to make their games better.... but the game doesn't necessarily have to be a very good simulation to be fun, you know what I mean? I'd imagine it would be harder to create a sim game for football or basketball...

We are just lucky a capable dev team has been able to use the available data about baseball and distill it into a nice video game... all I've been doing is just to show how good the gameplay really is as a simulation, and not so much exposing holes and deficiencies... we don't have a lot of ways to *fix* the game anyways.