Stats-Based CPU Sliders [MLB 12 Version]

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 12 Version]

75 games in the books? Wow, you sir are a machine. Through my first 20 games i was a little worried (very little because of the small sample size) about hits and runs being a tad low. But I got to say, so far games 21-28 are putting any worries to rest that I had. They have been very good offensive games. I'm still gonna do a 50 (at least) game run on Default but I agree with you, I believe this years version is that close without touching a thing. Your stats over 75 games obviously back that up!

I'm on Vita and keep running the games in background while I do other stuff, so that's the not-so-secret secret...

It's good to double check if Vita and PS3 versions both give similar stats, so we should still compare once you guys accumulate enough stats. Then we can say we are likely really on equal footing, haha.

[seanjeezy]

Re: Stats-Based CPU Sliders [MLB 12 Version]

Do you think lowering consistency 1 click would throw everything out of whack? wouldn't it address the low WP and HBP and possibly BB and strike%?

Edit:
Also Nomo, what have you seen in regards to relief appearances and manager hook? are specialists facing the right amount of batters with reliever stamina on default?

[Heroesandvillains]

Re: Stats-Based CPU Sliders [MLB 12 Version]

I'm impressed, Nomo.

I'm in a bit of a rush right now, but I'll be back tomorrow with some of my thoughts. I'm just going to suggest an, "Easy does it." I'd like to see you go to less extremes this year (in an effort to get a finalized set done, sooner versus later).

But I know how much you love to test the sliders, so obviously Tgreer, Floyd and myself can do that while you do whatever it is you've been planning on.

I'm thinking consistency could come down to 4, as well as fouls to 4, and stolen base success could go to 7 (thanks to Ninjoid).

And...I'm already elaborating when I promised I'd pick your brain tomorrow! LOL! I'll be back, but GREAT WORK MAN! You just saved so many of us so much time!!! Thank you, Nomo!

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 12 Version]

Do you think lowering consistency 1 click would throw everything out of whack? wouldn't it address the low WP and HBP and possibly BB and strike%?

I don't think it does, at least not by just one click. I was using Pitcher Consistency 2 (out of 10) last year to bring 2.5 walks to about 3.4 per game per team and that's quite a few clicks down, but it really didn't affect other parts of the game much.

If anything, this year I could add a bit more walks probably by just one or two clicks down, which probably won't bring HBPs and WPs close enough to real life; these need to be almost tripled. But I'd take walks more than WPs since the latter happens less in general.

Also Nomo, what have you seen in regards to relief appearances and manager hook? are specialists facing the right amount of batters with reliever stamina on default?

I'm tracking a couple stats to see how pitchers are used, but not detailed enough to see if the moves really make sense. But the number of pitchers used in games (Np in the spreadsheet), number of pitches starters throws (PC1), number of IPs starters throw on ave (IP Strt) are right on, so as far as the number of hooks it seems fine. I've seen 9 CGs which is about twice more than expected from MLB ave, but this might be luck so I'm not concerned much.

I'm impressed, Nomo.

I'm in a bit of a rush right now, but I'll be back tomorrow with some of my thoughts. I'm just going to suggest an, "Easy does it." I'd like to see you go to less extremes this year (in an effort to get a finalized set done, sooner versus later).

But I know how much you love to test the sliders, so obviously Tgreer, Floyd and myself can do that while you do whatever it is you've been planning on.

I'm thinking consistency could come down to 4, as well as fouls to 4, and stolen base success could go to 7 (thanks to Ninjoid).

And...I'm already elaborating when I promised I'd pick your brain tomorrow! LOL! I'll be back, but GREAT WORK MAN! You just saved so many of us so much time!!! Thank you, Nomo!

Is the BR Steal Ability slider backward this year again? Have to see you guys stats again...

I agree with Consistency 4. I probably end up using Foul 4 if the foul number keeps up, but I actually want to see a bit more swing and misses than what that slider can do.

Actually few appear concerned, but I find GIDPs quite nagging. They are rally killers and as annoying as all these balls hitting pitchers in up the middle... But don't know what the best way really is to reduce them.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 12 Version]

Okay, here's the set I'm gonna go with for the next test. What changed is in italics.

Contact: 4
Power: 5
Timing: 5
Foul Frequency: 5
Solid Hits: 5
Starter Stamina: 5
Reliever Stamina: 5
Pitcher Control: 5
Pitcher Consistency: 4
Strike Frequency: 5
Manager Hook: 5
Pickoffs: 5
Pitch Speed: 10
Fielding Errors: 6
Throwing Errors: 5
Fielder Run Speed: 4
Fielder Reaction: 5
Fielder Arm Strength: 2
BR Speed: 5
BR Steal Ability: 5
BR Steal Frequency: 7
Wind: 5
Injury Frequency: 5

My reasoning for the changes:

By lowering Contact, I hope to reduce batting average while increasing swing & miss which should lead to more Ks.

By lowering Pitcher Consistency, I hope to induce more walks, and slightly lower strike percentage as well. Likely not enough, but this should also increase HBPs and WPs a bit.

I see more throwing errors happening than fielding errors. The overall number of errors seems to be fine, but for now I'll just increase Fielding Errors to see how things go.

Decreasing Fielder Run Speed is largely aesthetic, but this might produce a little more XBHs. I don't wish to increase triples, but I can have more doubles.

Decreasing Fielder Arm Strength is also aesthetic but it's obvious arms are too strong so this is justified. Hopefully this contributes to slightly more XBHs, and reduce GIDPs on bang-bang plays at 1B.

Increasing BR Steal Frequency to increase steal attempts.

[Heroesandvillains]

Re: Stats-Based CPU Sliders [MLB 12 Version]

I don't think it does, at least not by just one click. I was using Pitcher Consistency 2 (out of 10) last year to bring 2.5 walks to about 3.4 per game per team and that's quite a few clicks down, but it really didn't affect other parts of the game much.

If anything, this year I could add a bit more walks probably by just one or two clicks down, which probably won't bring HBPs and WPs close enough to real life; these need to be almost tripled. But I'd take walks more than WPs since the latter happens less in general.

I'm tracking a couple stats to see how pitchers are used, but not detailed enough to see if the moves really make sense. But the number of pitchers used in games (Np in the spreadsheet), number of pitches starters throws (PC1), number of IPs starters throw on ave (IP Strt) are right on, so as far as the number of hooks it seems fine. I've seen 9 CGs which is about twice more than expected from MLB ave, but this might be luck so I'm not concerned much.





I'm impressed, Nomo.

I'm in a bit of a rush right now, but I'll be back tomorrow with some of my thoughts. I'm just going to suggest an, "Easy does it." I'd like to see you go to less extremes this year (in an effort to get a finalized set done, sooner versus later).

But I know how much you love to test the sliders, so obviously Tgreer, Floyd and myself can do that while you do whatever it is you've been planning on.

I'm thinking consistency could come down to 4, as well as fouls to 4, and stolen base success could go to 7 (thanks to Ninjoid).

And...I'm already elaborating when I promised I'd pick your brain tomorrow! LOL! I'll be back, but GREAT WORK MAN! You just saved so many of us so much time!!! Thank you, Nomo!Is the BR Steal Ability slider backward this year again? Have to see you guys stats again...

I agree with Consistency 4. I probably end up using Foul 4 if the foul number keeps up, but I actually want to see a bit more swing and misses than what that slider can do.

Actually few appear concerned, but I find GIDPs quite nagging. They are rally killers and as annoying as all these balls hitting pitchers in up the middle... But don't know what the best way really is to reduce them.

I'm seeing too many CPU GIDP's in my played games too; which perplexes me because (A) 11 had this down (though the ball physics undoubtedly have something to do with it being different now) and (B) as a slider tweaker, I have NO idea how to go about changing this other than possibly increasing Timing/Solid Hits (which are sliders I HATE touching).

I don't think I can disagree with any of the tweaks in the CPU sliders you posted Nomo (other than maybe the drastic change to arm strength...which may actually help CPU runners beat out more DP's; come to think of it! LOL).

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 12 Version]

I'm seeing too many CPU GIDP's in my played games too; which perplexes me because (A) 11 had this down (though the ball physics undoubtedly have something to do with it being different now) and (B) as a slider tweaker, I have NO idea how to go about changing this other than possibly increasing Timing/Solid Hits (which are sliders I HATE touching).

I don't think I can disagree with any of the tweaks in the CPU sliders you posted Nomo (other than maybe the drastic change to arm strength...which may actually help CPU runners beat out more DP's; come to think of it! LOL).

Yeah, GIDPs weren't that much of problem last year, and looking at groundball-flyball ratio, I don't think the game is producing too many grounded outs this year... last year, there was a slight over production of grounders. (But I don't trust the game's identifications of FB/LD/GB too much... sometimes they are questionable.)

I think the main reasons for inflated GIDPs that are different from last year are (1) new ball physics making the batted ball travel faster at times and (2) infielders throw quicker and throws are harder in general.

I lowered Arm Strength drastically to see if (2) is very significant, which may or may not... We cannot do much with (1).

I've also thought about tinkering with Solid Hit to change groundball/flyball ratio, but if my experiments with the slider last year give us any hints, it didn't change GIDPs much. I actually wouldn't mind increasing line drives (which is a bit lower according to above stats), but it can go either way... Increasing hard hit balls with higher Solid Hit could end up increasing DPs by making ball travel faster. Increasing grounder by lower Solid Hit could also end up increasing DPs by increasing grounders...

Will see what drastic reduction in Arm Strength does in this aspect first. We still have Fielder Reaction and BR Speed to use if necessary to adjust this part of the game.

[Heroesandvillains]

Re: Stats-Based CPU Sliders [MLB 12 Version]

I actually was wondering if you could report your findings in the bugs/glitches thread. Here's why I think this stuff can be patched:

Regarding GIDP's; I'm pretty positive that the CPU uses a "swing influence" of sorts, in a similar fashion that you or I would by pressing up on the RS. I personally think the issue is that the CPU is not trying to hit the ball in the air often enough with runners on 1st and less than two outs. I HONESTLY believe this can be fixed by SCEA without altering gameplay too much. What do you think? The CPU needs to influence fly ball in GIDP scenarios more often.

Regarding WP's and HBP's; specifically in terms of WP, I think SCEA can also patch this but it's cause to me isn't as clear as the above. I *think* the issue is that the ball doesn't get far enough away from the catcher on pitches in the dirt. Again, I think that it also could be patched. In the games I'm watching, I'm trying to envision what would happen if the runner decided to go and I think the runners play it properly. Meaning, I think they'd get thrown out if they decided to try to advance (more often than not). So I don't *think* it's a baserunning aggressiveness issue.

As far as HBP's go? I'm indifferent to it. They were so out of hand in this series a few years back that I'd rather SCEA leave it alone than risk over doing it. I trust them but they were crazy in the 10 version.

Anyway, I think you should report your findings and link the 75 game test as evidence in the bugs/glitches thread.

[steviegolfballs]

Re: Stats-Based CPU Sliders [MLB 12 Version]

Is the BR Steal Ability slider backward this year again? Have to see you guys stats again...


I'm starting to think it might be, I am testing with your latest post but I bumped the success rate up to a 7 and in the 10 games played I've had 11 steals and 10 thrown out. It's not a 50 game sample but it's 20 team games and a trend is forming.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 12 Version]

Is the BR Steal Ability slider backward this year again? Have to see you guys stats again...


I'm starting to think it might be, I am testing with your latest post but I bumped the success rate up to a 7 and in the 10 games played I've had 11 steals and 10 thrown out. It's not a 50 game sample but it's 20 team games and a trend is forming.

Yeah, this is actually a good thing to test so that people won't have to keep guessing for a while . The uncertainty for the sense is the reason why I didn't move it either way this time.

But steal can be tricky as well. I'm on 11 games with the new set, didn't change BR Steal Ability but what I have right now is 0.64 SB, 0.23 CS, which is pretty darn close to MLB ave (0.64 and 0.25). These things come in bunches with teams with good speed, so you want to keep BR sliders fixed for a while till you accumulate enough data.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 12 Version]

I actually was wondering if you could report your findings in the bugs/glitches thread. Here's why I think this stuff can be patched:

Regarding GIDP's; I'm pretty positive that the CPU uses a "swing influence" of sorts, in a similar fashion that you or I would by pressing up on the RS. I personally think the issue is that the CPU is not trying to hit the ball in the air often enough with runners on 1st and less than two outs. I HONESTLY believe this can be fixed by SCEA without altering gameplay too much. What do you think? The CPU needs to influence fly ball in GIDP scenarios more often.

Regarding WP's and HBP's; specifically in terms of WP, I think SCEA can also patch this but it's cause to me isn't as clear as the above. I *think* the issue is that the ball doesn't get far enough away from the catcher on pitches in the dirt. Again, I think that it also could be patched. In the games I'm watching, I'm trying to envision what would happen if the runner decided to go and I think the runners play it properly. Meaning, I think they'd get thrown out if they decided to try to advance (more often than not). So I don't *think* it's a baserunning aggressiveness issue.

As far as HBP's go? I'm indifferent to it. They were so out of hand in this series a few years back that I'd rather SCEA leave it alone than risk over doing it. I trust them but they were crazy in the 10 version.

Anyway, I think you should report your findings and link the 75 game test as evidence in the bugs/glitches thread.

Thing is these are really neither bugs or glitches, right... it's how you tune the game balance and certain things may be deliberately done for the rest of the game to play well.

For example, I think walks are down not because it's a hard thing to produce in game, but it necessarily gives user less control on pitch command and people may not like the lack of control.

I want to see if what I'm doing helps GIDPs before going "hey dev" route... it's probably true certain things can be improved in a patch (like WPs last year).

But if have enough evidence to support our request, it wouldn't hurt to report them in that thread I guess.

[Heroesandvillains]

Re: Stats-Based CPU Sliders [MLB 12 Version]

Thanks Nomo. Good points. I realize they're not exactly bugs, but maybe we can add this to the feedback threads when they're up then?

Here's my thinking:

We can all agree that SCEA's programming is deep. And generally, the CPU has the coding to access similar things that the User does. This probably relates to contact/normal/power swinging, the API logic, etc.

Now, I'm wondering if this applies to swing influence also? Specifically, situational hitting. The CPU knows when to sac bunt; hit and run, etc. I have to wonder if they're programmed to recognize GIDP situations?

I believe that if this logic or coding is in place; if an "up on the right stick" or flyball influence was tinkered with in GIDP situations, via a patch, that this "issue" could be tampered down to an MLB acceptable spectrum. And personally, I think my playing experience (when I actually play instead of CPU vs. CPU) would be better too.

Does anyone know if situational groundball/flyball influence is utilized by the CPU? Or specifically, in sac fly/advance runners/GIDP situations? I'd have to assume so, because I can do it as often or as little as I'd like.

The CPU is programmed to recognize my Guess Pitch and Guess Location habits too. And reversley, they recognize my patterns and Guess Pitch and Location on me. So I don't think I'm being dreamy if SCEA programmed the CPU to use Swing Influence too.

I may have to PM Knight or Nemesis (or even Ramone). If so, this can be patched and I think even "players" would appreciate the fine tuning. Especially if it can be patched for 12.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 12 Version]

Thanks Nomo. Good points. I realize they're not exactly bugs, but maybe we can add this to the feedback threads when they're up then?

Here's my thinking:

We can all agree that SCEA's programming is deep. And generally, the CPU has the coding to access similar things that the User does. This probably relates to contact/normal/power swinging, the API logic, etc.

Now, I'm wondering if this applies to swing influence also? Specifically, situational hitting. The CPU knows when to sac bunt; hit and run, etc. I have to wonder if they're programmed to recognize GIDP situations?

I believe that if this logic or coding is in place; if an "up on the right stick" or flyball influence was tinkered with in GIDP situations, via a patch, that this "issue" could be tampered down to an MLB acceptable spectrum. And personally, I think my playing experience (when I actually play instead of CPU vs. CPU) would be better too.

Does anyone know if situational groundball/flyball influence is utilized by the CPU? Or specifically, in sac fly/advance runners/GIDP situations? I'd have to assume so, because I can do it as often or as little as I'd like.

The CPU is programmed to recognize my Guess Pitch and Guess Location habits too. And reversley, they recognize my patterns and Guess Pitch and Location on me. So I don't think I'm being dreamy if SCEA programmed the CPU to use Swing Influence too.

I may have to PM Knight or Nemesis (or even Ramone). If so, this can be patched and I think even "players" would appreciate the fine tuning. Especially if it can be patched for 12.

I don't know exactly how things are coded internally, so I wouldn't speculate on that exactly... but it's like true that CPU has situational awareness, which means with a runner on 1B, he would try to avoid inducing a GIDP. How exactly? I don't know, but *if* how that's done is really the culprit, then...

Are anyone else tracking GIDPs though? It's a bit inflated on 75 game testing above, but they aren't crazy inflated... I have to keep watching this, but so far with 12 games in with my new set, it's 0.96 per game per team now (0.75 MLB ave). So it has come down a bit.

[Heroesandvillains]

Re: Stats-Based CPU Sliders [MLB 12 Version]

I'm going to go back through my games and write down the the DP's from the ones I remember watching. Otherwise, I'm going to continue tracking then from here on out.

I agree. It can't just be your data.

Ugh...wish I wrote the matchups down.

But I can assure you I'm seeing it too. In played games and CPU games. When you said "rally killer," that was spot on. It's not insane, like you were saying, but it can be a bit deflating. I'm excited by your newer results, though.

Hopefully everyone is tracking them too.

[steviegolfballs]

Re: Stats-Based CPU Sliders [MLB 12 Version]

I don't know exactly how things are coded internally, so I wouldn't speculate on that exactly... but it's like true that CPU has situational awareness, which means with a runner on 1B, he would try to avoid inducing a GIDP. How exactly? I don't know, but *if* how that's done is really the culprit,

Are anyone else tracking GIDPs though? It's a bit inflated on 75 game testing above, but they aren't crazy inflated... I have to keep watching this, but so far with 12 games in with my new set, it's 0.96 per game per team now (0.75 MLB ave). So it has come down a bit.

I'm tracking them and I am getting roughly the same results. About 1.0 per game per team. Not outlandish but above average.