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

[nomo17k]

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

I am not blacked out here in Western Canada where the closest teams (Seattle and Denver) are not close at all. No if there is a MLB.TV blackout in Canada it would only be restricted to the GTA. As far as VPN performance,
I use IPVanish on my home computer and my smart TV as well as my cell phone and I do not notice any performance issues. But I don't watch baseball games from my phone and yeah that is concerning if they will be using GPS to determine if you should be blacked out.

I see, that makes sense... For a moment I thought the entire Canada is blacked out for Blue Jays games. That'd be insane, LOL!

[nomo17k]

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

Just a brief update on sliders.

The 4/4 set looks pretty close to what I eventually want to settle in terms of overall feel of how the game plays, but there are a few issues that I potentially want to correct if I can, e.g.,

- decrease both BB and SO
- decrease HBP and WP
- increase doubles

Pitchers are a little more wild than I'd like them to be, and I'd normally want to keep Pitcher Consistency closer to 5 (currently 3), but a comparing a previous 4/2 set which is close to the 4/4 set except Timing and Pitcher Consistency, there appears quite a bit of "jump" in how the game plays... so I'm exploring some slider combination to achieve above without touching Pitcher Consistency.

I think I'm already hitting a sort of "uncanny" (or weirdness) valley in slider parameter space, at which just a click or two of sliders causing the game to play substantially differently. Those who have followed slider threads from previous years probably know this had happened in the last couple years. Basically I feel it's in a "close but no cigar" phase... good thing is, at least it's already pretty close.

[nomo17k]

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

So the last post was for the 4/6 set and until I converge on a set that I can consider "recommended," I'm likely to continue fine tuning as I have been for the set since the last post.

For the sets 4/6-4/8, I've been mostly trying to balance BABIP to be around .300, since that is one of the most useful indicator of if the hit variety is well balanced to produce realistic batting average.

I'm mostly satisfied with per-pitch stats like strike %, though I'd like to keep Pitcher Consistency higher if I could to help reduce HBP and WP. But doing so would probably reduce the amount of walks, so I am making compromise there.

With the new ball physics and the hit variety (which are both easily the best ever seen in The Show), the game doesn't have as much problem producing doubles as it has in the past. Instead, I tend to inflate them a bit with quite a few offense-friendly sliders tipping toward helping solid hit variety. For a better balance between singles and XBHs, I reduced Fielder Reaction in hope of increasing the number of singles.

That's where things are at with the 4/8 set.

[nomo17k]

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

I'd also like to keep this thread open for random discussions about slider and adjustments, because it's my belief that there isn't really a single slider set that nails down everything (even for HUM-free CPU vs. CPU games), so if people want to do their own adjustments, I'd encourage them to do so, and I'd like to offer help where my experience having been doing this for so long can be useful.

I'm not really doing this to take credits or feel good about myself or keep the fruit of labor exclusive to myself, so I'm very open to transfer of knowledge to those who are interested and so inclined.

For example, I really wouldn't have minded someone doing similar thing (like sonsart2 did earlier... the only issue for me was that there is no reason to duplicate the thread title I've using, and I wished he could at least drop me a note of acknowledgement, if he were to use pretty much the same spreadsheet that I've created over years for example).

So if you have any question as to what I am doing or if you think my input would help you adjust your slider set to your liking, please feel free to post your question to this thread.

[dsmith710]

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

Nomo, I’ve been following along most of the time with your spreadsheet as I’m getting ready to start a MoM franchise. I really like what you have going.

Do you think the amount of doubles are correlated with Foul Frequency? It’s easy to assume it affects foul balls, lengthening at-bats, but doesn’t it also impact whether or not those balls down the line stay fair/foul?

[nomo17k]

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

Nomo, I’ve been following along most of the time with your spreadsheet as I’m getting ready to start a MoM franchise. I really like what you have going.

Do you think the amount of doubles are correlated with Foul Frequency? It’s easy to assume it affects foul balls, lengthening at-bats, but doesn’t it also impact whether or not those balls down the line stay fair/foul?

This is an interesting possibility, given that the Foul Frequency slider description says that one of the things it does is to "increase fouls on borderline hits that would have landed just fair," and some doubles are hit down the foul lines.

I still doubt that is a major factor in inflated doubles... the reason is that, despite the vastly improved ball physics, I think the ball hit down the line still doesn't roll away from the foul line like we tend to see in real life, and if we concentrate on hits landing near the foul line, I'd still bet they tend to remain singles in the game more often than they'd become doubles in real life.

For the sets I've been experimenting with, doubles are probably easy to come by since I have sliders like Contact and Solid Hit a notch above the default (along with Pitcher Control a couple clicks down from the default) which tend to increase line drives (which shows up as LD% in the spreadsheet).

[Incognito]

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

Interesting game variation on your latest set. A string of below average scoring games, followed by a slug fest where the Yankees score 20 runs w/9 home runs hit, wow.

Based on your stats and most games I've seen, I've also noticed stolen bases are down. I'm not sure if the base runners getting thrown out trying to advance on pitches in the dirt play any significant role in the SB percentage or not.

[nomo17k]

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

Interesting game variation on your latest set. A string of below average scoring games, followed by a slug fest where the Yankees score 20 runs w/9 home runs hit, wow.

Based on your stats and most games I've seen, I've also noticed stolen bases are down. I'm not sure if the base runners getting thrown out trying to advance on pitches in the dirt play any significant role in the SB percentage or not.

I'd like to believe that the variation of mostly very low offense games with occasional offensive spikes isn't a pattern, but that Yankees game is certainly an anomaly... I think the new Yankees Stadium in the game is HR-friendly in the game (as it is in real life), but 9 HRs by the home team is certainly something I don't see often.

Evaluation SB can be tricky, since the CPU manager has a tendency for over-reliance on hit-and-run, which can make computing SB success rate and frequency a little harder to interpret. The game also records SB on successful suicide squeezes... both of which I try to take note of when I can and remember.

I'm trying to see which BR Steal Frequency (6 or 7) is more appropriate right now.

[Guys_WhoGame]

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

I'd like to believe that the variation of mostly very low offense games with occasional offensive spikes isn't a pattern, but that Yankees game is certainly an anomaly... I think the new Yankees Stadium in the game is HR-friendly in the game (as it is in real life), but 9 HRs by the home team is certainly something I don't see often.

Evaluation SB can be tricky, since the CPU manager has a tendency for over-reliance on hit-and-run, which can make computing SB success rate and frequency a little harder to interpret. The game also records SB on successful suicide squeezes... both of which I try to take note of when I can and remember.

I'm trying to see which BR Steal Frequency (6 or 7) is more appropriate right now.

So parks actually have an effect in-game? Do you think that can mess with stats? example: Playing at a pitcher park like PNC park?

[nomo17k]

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

So parks actually have an effect in-game? Do you think that can mess with stats? example: Playing at a pitcher park like PNC park?

There are definitely park effects when you play out the game (the simulation engine doesn't appear to have any notion of ball park though). The most noticeable is the park dimension and winds (though I'm not exactly sure the wind strength and directions are accurately modeled after each real-life stadium location), but things like the altitude also has an effect.

I think the devs also said a couple years ago they started implementing a little more detailed atmospheric effect (like temperature), but that would not be easily noticeable.

It is interesting. A couple years ago when I simulated a whole season all played in CPU vs. CPU games and looked at the Colorado Rockies's home and road stats, for example, they hit for a significantly higher batting average at home in the game.

As for "messing with" the stats, the variety is what I actually want to be counted toward the stats. To avoid heavily biased toward a small number of more peculiar ball parks, many games with different starting pitchers and home teams are rotated, in hope that things average out over time.

[ericjwm]

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

Depends on how the current set goes. (And patches not messing things up...)

Patch 1.07 messed things up for both playing and manage. I had the solution for 1.06. It was hitting: HOF & pitching: veteran with power at 8. It worked and now it's gone... I'm pissed because the system automatically downloads & installs even if you're playing another game.
So start over from 1.00? IDK
They're going to keep patching

[ericjwm]

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

I'm sorry Nomo not trying to invade your thread. You are a very passionate person about realism, as am I.. Just giving some input about the patches. Hint: batter power in 1.07

[Houston]

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

Hey Nomo, great job as always. I have a recommendation that maybe you could consider.

Instead of lowering reaction have you considered maybe lowering fielder speed? they have always been on default way to fast and they at times look in warp speed.

I myself pay close attention on balls played and how the fielders are running. I get why some slider gurus have it high or on default to cut down hits etc, but I much prefer the fielders running at a realistic rate. I guess visually is what I am talking about that there legs should be moving as they should in real life and at 5 they get going so fast that it look cartoonist. I have always used 0 or 1 and it makes me feel like the player is actually playing in a big area or big park when you do this.

If you half to use 3 I would much rather see 3 than 5. 3 would still cut down on hits but not look as silly fast as 5 but for me 1 looks the best.

[nomo17k]

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

Patch 1.07 messed things up for both playing and manage. I had the solution for 1.06. It was hitting: HOF & pitching: veteran with power at 8. It worked and now it's gone... I'm pissed because the system automatically downloads & installs even if you're playing another game.
So start over from 1.00? IDK
They're going to keep patching

I'm sorry Nomo not trying to invade your thread. You are a very passionate person about realism, as am I.. Just giving some input about the patches. Hint: batter power in 1.07

I've been keeping track of when patches are applied in the spreadsheet, but I don't think I have seen noticeable changes in how the game plays before and after each patch (not yet, at least).

Though inexplicit changes being made to the game via patch can be very frustrating (like the one happened with Power last year), I'd give the game a little longer time to decide whether a patch introduces any changes. It is very easy to fool oneself with spurious correlations when a patch gets released.

[nomo17k]

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

Hey Nomo, great job as always. I have a recommendation that maybe you could consider.

Instead of lowering reaction have you considered maybe lowering fielder speed? they have always been on default way to fast and they at times look in warp speed.

I myself pay close attention on balls played and how the fielders are running. I get why some slider gurus have it high or on default to cut down hits etc, but I much prefer the fielders running at a realistic rate. I guess visually is what I am talking about that there legs should be moving as they should in real life and at 5 they get going so fast that it look cartoonist. I have always used 0 or 1 and it makes me feel like the player is actually playing in a big area or big park when you do this.

If you half to use 3 I would much rather see 3 than 5. 3 would still cut down on hits but not look as silly fast as 5 but for me 1 looks the best.

The reason why I lowered Fielder Reaction is for hopefully increasing singles that go through infielders, since the ratio of XBHs to singles has tilted consistently toward the former. Since I already have enough XBHs, I did not want to potentially increase them by lowering outfielder speed.

In the past iterations of the game, I agree that outfielder speed was rather a big issue when they ran in the vast greenery of outfield as if there is no concept of weight whatsoever. But I think a couple iterations ago, their speed was lowered quite a bit.

The broadcast cam can fool our eyes due to the peculiar perspective it uses to keep the action happening at a distance closer... effectively making outfielders appear to move quicker than they actually do.

Therefore the "eye test" I usually perform on outfield speed is to watch vault movies at the end of a game, zooming out far and above the stadium, almost looking straight down and watch how fast fielders move.

Do I think outfielders move too fast in the current game when I do that? I am actually not sure if they are moving very realistically, but I don't necessarily think they are unphysically quick either.

What do you think when you do this eye test? I think the root issue with outfielders in the game is not their intrinsic quickness, but they don't have a proper concept of how fast they could/should move approaching certain types of hit. Real-life outfielders don't always move at a constant speed, but in the game, they mostly do run at a constant speed (except when AI somehow decides they can jog... which can lead to frustration sometimes when you are controlling the fielder yourself).

All that said, I prefer lower outfielder speed if that's what I can make happen. But realistic stats are a higher priority, so until the need arises I am not quite touching Fielder Run Speed slider.