Stats-Based CPU Sliders [MLB 18 Version Season-Mode Memorial Edition]

[BigShow2010]

Re: Stats-Based CPU Sliders [MLB 18 Version Season-Mode Memorial Edition]

Hey nomo. Do those sliders work when simulating or in quick mange mode?

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 18 Version Season-Mode Memorial Edition]

Hey nomo. Do those sliders work when simulating or in quick mange mode?

When you sim or quick manage, the game runs on a different engine, so most sliders (except Stamina, Manager Hook, Injuries) do not work similarly as they do for the actual gameplay.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 18 Version Season-Mode Memorial Edition]

That was weird. I know if I put 6 Power for the CPU, I'd likely be getting bombed out of the park on HoF. So that's, what, like 7 or 8 on All-Star? I know above All-Star gives a bonus to the listed sliders (or at least it used to).

I don't understand what you mean by bonus, but I think sliders mostly work as offset modifiers to the base parameters which you can change by the difficulty level.

I was curious how much we can tune by sliders vs. difficulty level, and when I was experimenting with sliders for my own playing before, I found that (some) sliders have enough dynamic range as to effectively have a bigger effect than a change in difficulty level.

For example, Contact slider at 10 with HoF difficulty would give you a bigger PCI than Contact slider at 5 (default) with All-star. Some people complain about difficulty level not having enough granularity (e.g., some would want to fix the game at a difficulty between All-star and HoF), but slider adjustments can make that happen.

I've also seen that offense droop. It's almost like a wider PCI + lower power = more any contact but that contact is likely weaker because it could catch the edges of the PCI and lower power will make those a lower batted ball velocity. So, unless the higher Contact makes a better aligned PCI, the lower Power will neuter any of those "glancing hits", lowering BABIP and all contact "oomph", so to speak.

Do you think that's what you were seeing, Nomo?

As for Solid Hits - I have no idea. Sometimes it seems like that slider does nothing when I adjust it in vsCPU. I can only describe it to myself as the "carry" slider (how much carry do the hits have). That's about all I see it do.

For simplicity (and also because I think it's a reasonable assumption), my assumption has been that each slider works independently of others. For example, Solid Hits probably only adjusts the launch angle of batted ball, Power only adjust the initial speed of the batted ball of the bat contact, etc. Those make sense in view of what's been said in the slider descriptions as well.

But especially this year, I feel that assumption might not be entirely correct.

What you said about what Power slider did by one click would be roughly similar to what I thought that would do. My thinking of bringing Power back to the default/5 was that if it adjusts only the launch speed of the batted ball, then the right amount of that adjustment would mostly affect the fraction of fly balls/line drives that goes over the fences, without negatively affecting the rest of hit types (because of compensation effects among all different hit types that do not go over the fences). The question was how much... and it was quite drastic.

And now see what is happening with the 4/26 and 4/27 sets... the only change was that I increased Pitcher Control by one in the latter, in hope of keeping the offense in check a bit. That was achieved, but now the HR% literally tanked.

It does sort of makes sense that HRs are reduced... the pitchers are making their pitches more effectively after all. But should it reduce HRs *this* much, by just one click of a slider which affects the aspect of the game indirectly?

Not according to my past experience and intuition adjusting sliders for the game.

I think ball physics and pitching/batting engine has probably changed quite drastically in this regard.

[doctemple]

Re: Stats-Based CPU Sliders [MLB 18 Version Season-Mode Memorial Edition]

It's a kind of difficult question since the reason why I haven't declared any set to be recommended is that none has quite satisfied me in the areas that I think are essential, but if this were the last day before I would get abducted by aliens to some distant planet, I'd say the 4/21 or 4/22 set would be the closest to the target stats.

As I look at then again, they are both pretty good to the extent that I wonder why I didn't make one of them recommended, but I think I wasn't satisfied by the BB% (a bit low) and HR% (a bit high), and also I wanted to explore slowing down Fielder Run Speed for visual aesthetics.

But I think either slider set plays reasonably well.

Hey Nomo, I don’t see your sliders in the vault. Probably a dumb question, but how do I get them?

[NolanRyansSnowmonkey]

Re: Stats-Based CPU Sliders [MLB 18 Version Season-Mode Memorial Edition]

Hey Nomo, I don’t see your sliders in the vault. Probably a dumb question, but how do I get them?

I'll answer, just in case he isn't around and you're trying to get some games in. On the Original Post, there is a link to a spreadsheet. Near the top of that spreadsheet there are different dates listed across the top. Choose the date you want, and the sliders for that set are at the top of the sheet. He mentioned 4/21 or 4/22 are closest for now.

[KBLover]

Re: Stats-Based CPU Sliders [MLB 18 Version Season-Mode Memorial Edition]

I don't understand what you mean by bonus, but I think sliders mostly work as offset modifiers to the base parameters which you can change by the difficulty level.

The last time I read anything about that (years back), difficulty was explained sort of like this:

All-Star = All sliders are as listed for both Human and CPU.

HoF = Negative "offset" to the listed Human sliders, positive "offset" (a.k.a what I called "bonus") to the CPU sliders.

So a 6 Contact and 5 Control for both human and CPU would be:

6 Contact, 5 Control for Human, 6 Contact, 5 Control for CPU on All-Star

5 Contact, 4 Control for Human, 7 Contact, 6 Control for CPU on HoF

Something along those lines. I was wondering if this was still (or ever was) true.

For example, Contact slider at 10 with HoF difficulty would give you a bigger PCI than Contact slider at 5 (default) with All-star. Some people complain about difficulty level not having enough granularity (e.g., some would want to fix the game at a difficulty between All-star and HoF), but slider adjustments can make that happen.

Completely agree. I use difficulty level more for CPU strategy. As you said, sliders can deal with other gameplay aspects.

For simplicity (and also because I think it's a reasonable assumption), my assumption has been that each slider works independently of others. For example, Solid Hits probably only adjusts the launch angle of batted ball, Power only adjust the initial speed of the batted ball of the bat contact, etc. Those make sense in view of what's been said in the slider descriptions as well.

But especially this year, I feel that assumption might not be entirely correct.

What you said about what Power slider did by one click would be roughly similar to what I thought that would do. My thinking of bringing Power back to the default/5 was that if it adjusts only the launch speed of the batted ball, then the right amount of that adjustment would mostly affect the fraction of fly balls/line drives that goes over the fences, without negatively affecting the rest of hit types (because of compensation effects among all different hit types that do not go over the fences). The question was how much... and it was quite drastic.

And now see what is happening with the 4/26 and 4/27 sets... the only change was that I increased Pitcher Control by one in the latter, in hope of keeping the offense in check a bit. That was achieved, but now the HR% literally tanked.

It does sort of makes sense that HRs are reduced... the pitchers are making their pitches more effectively after all. But should it reduce HRs *this* much, by just one click of a slider which affects the aspect of the game indirectly?

Not according to my past experience and intuition adjusting sliders for the game.

I think ball physics and pitching/batting engine has probably changed quite drastically in this regard.

Yeah, it's hard to figure out. From dickering, this is my working assumption so far:

Power = Batted ball velocity with some influence to batted ball type. So power might not just make harder line drives, it might make more line drives (which would also be harder on average. Power might be taking on some of the "old" Solid Hits role of bringing trajectories to a better spot...perhaps better based on batted ball type (grounders get the best ground ball trajectory more often, etc.)

Solid Hits = modifier to contact quality (Weak, Okay, Good) with some influence to batted ball trajectory (mostly helping avoid the worst trajectories). So if the batter hits a ball at the very bottom edge of his PCI, Solid Hits might get him a "normal" grounder instead of a chopper. And then Power would influence how fast that grounder traveled given the "Weak" contact.

And if Control also impacts the Weak/Okay/Good label, that might be why the HR dropped off so badly. More flyballs lost their "ideal" trajectory and then lost their "best" velocity, so they didn't carry. If they still found grass, they could be hits but HR? Nope.

From the perspective of a non-CPUvCPU guy, this seems to be how my mind is working with it right now. I don't know how much sense it makes, honestly.

[NolanRyansSnowmonkey]

Re: Stats-Based CPU Sliders [MLB 18 Version Season-Mode Memorial Edition]

I'm running the 4/25 set out of curiosity and although I'm only halfway through the second game, there have been 28 runs scored, 39 hits, 12 doubles, and 6 home runs. I know two games are nothing, but offense looks very potent. The next 23 games may very well be duds though. Even in the 4/25 there were a few high offense games so this means nothing as of now.

[ericjwm]

Re: Stats-Based CPU Sliders [MLB 18 Version Season-Mode Memorial Edition]

Gotta say, some really interesting discoveries you guys have made lately.
One thing I'm extremely aware of when testing CPU vs CPU is how often the batters pull the ball for hits or outs. Real MLB is very pull heavy these days. I notice with certain slider settings the batters tend to go the other way too often. I've tested Nomo's current sliders and they're playing out fairly well so far. But I'm still experimenting with things to get more pulled ground outs and fly outs... a lot seem to go opposite field still.

I don't especially like fooling with the timing slider because there is not much noticeable discrepancy between clicks lower or higher.

Does anyone have any comments on this or agree/disagree? Like I mentioned it's one of those things I'm obsessive about for CPU vs CPU

[ericjwm]

Re: Stats-Based CPU Sliders [MLB 18 Version Season-Mode Memorial Edition]

I think I prefer strike frequency at 6.. as even the best of pitchers allow too many free passes in odd situations when set on 5... unless you prefer to raise pitch control. Essentially they seem to all of a sudden lose their "stuff" and go from blowing the ball right by batters to curiously pitching around a not so threatening batter when strike frequency is lower.


The pitcher not trying to field the ball on solid hits up the middle and getting hit by the pitch is still a major problem.

[forme95]

Re: Stats-Based CPU Sliders [MLB 18 Version Season-Mode Memorial Edition]

Gotta say, some really interesting discoveries you guys have made lately.
One thing I'm extremely aware of when testing CPU vs CPU is how often the batters pull the ball for hits or outs. Real MLB is very pull heavy these days. I notice with certain slider settings the batters tend to go the other way too often. I've tested Nomo's current sliders and they're playing out fairly well so far. But I'm still experimenting with things to get more pulled ground outs and fly outs... a lot seem to go opposite field still.

I don't especially like fooling with the timing slider because there is not much noticeable discrepancy between clicks lower or higher.

Does anyone have any comments on this or agree/disagree? Like I mentioned it's one of those things I'm obsessive about for CPU vs CPU

I have to agree with you. I hate messing with timing. Even at default I tend to see a lot of hits go oppo this year. More then average. I actually wanted to go up a tick on timing, but I am so against moving it.

Nomo's 17 set was great and loved the outcomes and gameplay, but again with timing at 4 it was defeating the purpose of shifting as a lot of hits where coming opposite the shift allowing a lot of extra hits.

[KBLover]

Re: Stats-Based CPU Sliders [MLB 18 Version Season-Mode Memorial Edition]

I have to agree with you. I hate messing with timing. Even at default I tend to see a lot of hits go oppo this year. More then average. I actually wanted to go up a tick on timing, but I am so against moving it.

I think that's less the Timing slider and more that it's just too easy for hitters to go "against their tendencies" in this game.

High, low, doesn't matter for me. I pitch inside with offspeed and they just push it the other way anyway because they are late.

In fact, I think if they were on-time, it would get pulled more because on-time on a pitch in should be a pull-ish ball, which is into the shift. Even if it's up the middle, that's where 2B/SS is playing.

So I'm thinking maybe I should experiment with higher than 5 if anything. But Timing doesn't always dictate if a hitter is early/late. Sometimes it seems the "timing error" is more random than anything while shifts only work if the hitter is slave to their tendency.

[doctemple]

Re: Stats-Based CPU Sliders [MLB 18 Version Season-Mode Memorial Edition]

I'll answer, just in case he isn't around and you're trying to get some games in. On the Original Post, there is a link to a spreadsheet. Near the top of that spreadsheet there are different dates listed across the top. Choose the date you want, and the sliders for that set are at the top of the sheet. He mentioned 4/21 or 4/22 are closest for now.

Thank you so much for “pinch hitting”!!!

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 18 Version Season-Mode Memorial Edition]

Hey Nomo, I don’t see your sliders in the vault. Probably a dumb question, but how do I get them?

I'll answer, just in case he isn't around and you're trying to get some games in. On the Original Post, there is a link to a spreadsheet. Near the top of that spreadsheet there are different dates listed across the top. Choose the date you want, and the sliders for that set are at the top of the sheet. He mentioned 4/21 or 4/22 are closest for now.

Thanks for helping out. I usually have access to the forum with the smartphone, but I tend to write only on my computer.

I don't always post the exact slider set in this forum, since they are all accessible in the spreadsheet.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 18 Version Season-Mode Memorial Edition]

The last time I read anything about that (years back), difficulty was explained sort of like this:

All-Star = All sliders are as listed for both Human and CPU.

HoF = Negative "offset" to the listed Human sliders, positive "offset" (a.k.a what I called "bonus") to the CPU sliders.

So a 6 Contact and 5 Control for both human and CPU would be:

6 Contact, 5 Control for Human, 6 Contact, 5 Control for CPU on All-Star

5 Contact, 4 Control for Human, 7 Contact, 6 Control for CPU on HoF

Something along those lines. I was wondering if this was still (or ever was) true.

...

Hmm, if there has been an explicit revelation like this from the developers, I have totally missed it. If this was posted, then it probably predates me, and it is certainly a possibility, given that the devs chimed in much more often before I joined the forum (kinda sad).

Even if the sliders used to work the way you described, I actually doubt that they actually work exactly like this right now.

One experiment I did (I think last year) was to find a slider combination (along the line of how you described) to see if I could find any combination that can exactly match difficulty level by purely adjusting sliders. I only used PCI size, but, for example, I think I couldn't exactly find, say, HoF with some positive Contact offset to match the PCI size on All-star with sliders at default.

If sliders worked the way you describe, I should've been able to exactly match the PCI size by some combination. But they certainly come close in certain combination.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 18 Version Season-Mode Memorial Edition]

...

And if Control also impacts the Weak/Okay/Good label, that might be why the HR dropped off so badly. More flyballs lost their "ideal" trajectory and then lost their "best" velocity, so they didn't carry. If they still found grass, they could be hits but HR? Nope.

...

I actually still doubt Control works as a direct modifier to batter's PCI. Since it influences the overall offensive output, it feels that it has a direct influence, but I think the correlation is spurious. Of course I cannot make this claim with any certainty, but it doesn't fit at all with what B Ma has said about how the Control/Consistency/Strike Frequency sliders work... and I still think his explanation and implementation make sense mostly, so breaking away from it appears very unreasonable thing to do.

I still think this conceptualization is correct:



but I think there is certain quirkiness in the balance between what kind of pitches CPU pitchers throw and how CPU batters place the PCI on them.

What I mean is kinda hard to describe... but I think one thing about PCI placement is that CPU has a particular tendency in how it places its PCI on pitches, and that's quite different from the user. This is easily seen when you use Zone hitting, with which some of us tend to generate a lot more variety of hit types (even more pop-ups and swinging bunts than CPU does). The user shows a lot more freedom in how he places PCI, but CPU seems to have a lot less ways of placing PCI.

And I suspect I'm seeing the quirkiness caused by the CPU tendency of PCI placement in how hit types varies "jumpy" rather than smoothly when a minor slider change is introduced. This year, the quirkiness appears to be especially magnified in those sudden jumps in offensive outputs by a very minor adjustment in one slider at a time... and my suspicion is that the devs didn't have as much time smoothing things out given how much the PCI mechanism changed for this years game.

Basically on surface the PCI appears the same as in the past, but given the vast increase in hit types, the hitting engine now has to have a lot more ways to generate different kinds of hits. The increase is mostly in the poor hit category (the past games did fine in producing solid hits mostly), the part of PCI which used to generate solid hit now needs to be "assigned" to generate hits of a little poorer quality, etc.

That's a lot of fine tuning involved, but what if the PCI placement of hitting AI did not receive appropriate adjustments accordingly?

I think that's the kind of quirkiness that I am seeing here... at least partially.