Stats-Based CPU Sliders [MLB 17 Version]

[nomo17k]

This thread will be about yet another attempt at creating a slider set based on empirical data through running a bunch of CPU vs. CPU games, without any user input at all, just because there should still be a place for misanthropes. There was a similar thread last year.

CPU vs. CPU games master data spreadsheet

The currently recommended "beta" set (a slight modification of the 5/16 & 5/14v2 sets).

Contact: 6
Power: 5
Timing: 4
Foul Frequency: 5
Solid Hit: 6
Starter Stamina: 7
Reliever Stamina: 7
Pitcher Control: 3
Pitcher Consistency: 4
Strike Frequency: 5
Manager Hook: 3
Pick Offs: 5 (untested)
Fastball Pitch Speed: 10
Offspeed Pitch Speed: 10
Fielding Error Infield: 8
Fielding Error Outfield: 5
Throwing Error Infield: 7
Throwing Error Outfield: 8
Fielder Run Speed: 3
Fielder Reaction: 5
Fielder Arm Strength Infield: 5
Fielder Arm Strength Outfield: 5
Baserunner Speed: 5
Baserunner Steal Ability: 7
Baserunner Steal Frequency: 5
Wind: 3
Injury Frequency: 5 (untested)
Trade Frequency: 5 (untested)

Notes on this set:

• Wild pitches are off, and only a patch can fix their inflation (keep fingers crossed)
• Batting average is about .005 down (plays at .250)
• HRs might be slightly inflated, but still within realistic level at 1.20 HRs per game per team (which is close to the current 2017 MLB average).

A slight variation of the beta slider set that would play just as well but with a different flavor:

Pitcher Control: 2
Pitcher Consistency: 5

with everything else the same. With this change, you will see slightly less walks but more hits, but the overall amount of offense should be similar. This set is roughly based on the 5/23 & 5/22 sets.


User Guide

These are adjustment made under the following philosophy for the game to play

• statistically as close as possible to the 2016 MLB averages
• aesthetically pleasing baseball for watch as CPU vs. CPU games

The assumption is that the slider set is used with a roster set with attribute distribution at the MLB level that are similar to the default SCEA roster that ships with the game. An example of such a roster set includes the very first release of OSFM.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 17 Version]

Another place holder for now.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 17 Version]

I've already gotten in the initial 75-game set with default sliders... Typically I've maxed out Pitch Speed sliders just because doing so makes the apparent pitch speed closer to real-life numbers, but I started out with everything at default this year, as maxing out Pitch Speed had (negative) side-effects on steal success rate in the last couple years.

Here are some notable points from all sliders at default, compared to real-life MLB numbers:

• Batting average is low (.238 vs. .255).
• BABIP is a bit low (.286 vs. .300) likely because the game produces more fly balls than ground balls.
• Swing & miss % is a bit higher (23.7% vs. 21.8%). This increases SO% (24.0% vs. 21.1%) and also contributes to the low batting average.
• Hitters do not chase chase as much (O-Swing % 27.3% vs. 30.3%), but this has been pretty constant the last couple years.
• HRs are a bit inflated (5.3 HR/BIP vs 4.6 HR/BIP).
• Doubles might be a little bit down still (6.1 2B/BIP vs. 6.8)... I think hits down the line do slice away in flight, but once they hit the ground they still don't roll away as much, so we s may be still missing the kind of doubles we often see down the line in real life.
• Wild pitches are quite high (0.95 per game vs. 0.37). Hit-by-pitch is also a little high, which might mean that pitchers are tuned to be wilder than in the past.
• Walks are a little down (BB% of 7.7% vs. 8.2%).
• Fielding errors are down (only about 1/3 of real-life numbers).
• There may be a slight tendency for early hook for starters "on average", but relievers appear to throw more innings.

[tgreer]

Re: Stats-Based CPU Sliders [MLB 17 Version]

As an MoM only guy I wait for these every year. Glad you are doing this again Nomo thank you. I ran approx 35 test games before I saw you were back. Main things that have stuck out to me are too many K's and WP's on default along with the low batting avg. Let me know if there are any side tests you need ran. I can get in 8-12/day as I work at home. Looking forward to these as always. My stats are not as in depth as yours (whos are lol). But I do keep up with R,H,E,2B,3B,HR,SB,CS,BB,K,WP,BA AVG

[floydpinkster]

Re: Stats-Based CPU Sliders [MLB 17 Version]

Great to have you back Nomo... I have "Borrowed" your Spreadsheet and have been Inputting stats as you do. If you need any help running games let me know!

[Armor and Sword]

Re: Stats-Based CPU Sliders [MLB 17 Version]

Nomo you always bring great information every year to The Show Nation. Your efforts are very appreciated.


Sent from my iPhone using Operation Sports

[floydpinkster]

Re: Stats-Based CPU Sliders [MLB 17 Version]

So far the worst problem I have seen is the crazy amount of Wild Pitches... The are having a big impact on the games, and by aces. have you see this aswell Nomo?

[Teleo]

Re: Stats-Based CPU Sliders [MLB 17 Version]

Excited for these!

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 17 Version]

Nomo you always bring great information every year to The Show Nation. Your efforts are very appreciated.

Excited for these!

Thanks for your kind words, guys. It's my pleasure if the shared info helps people enjoy the game more.

I'm doing this in part to help improve our overall simulation baseball experience by appreciating how great the game actually plays in many areas already, but also pointing out where it still falls short, to offer some explanation as to why some areas of the game isn't quite right yet...

I believe discussing all these things constructively in the community is a great way to give the developers feedback for bringing even more authenticity to The Show.

As an MoM only guy I wait for these every year. Glad you are doing this again Nomo thank you. I ran approx 35 test games before I saw you were back. Main things that have stuck out to me are too many K's and WP's on default along with the low batting avg. Let me know if there are any side tests you need ran. I can get in 8-12/day as I work at home. Looking forward to these as always. My stats are not as in depth as yours (whos are lol). But I do keep up with R,H,E,2B,3B,HR,SB,CS,BB,K,WP,BA AVG

Nice to see you again, tgreer. I value MoM very much... I strongly think good AI should always be able to play realistically against another AI, and in baseball game that tends to be in MoM. Hope this year's game plays well in the mode.

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 17 Version]

Great to have you back Nomo... I have "Borrowed" your Spreadsheet and have been Inputting stats as you do. If you need any help running games let me know!

Feel free to use the spreadsheet anyway you want. I don't gain anything by keeping anything secret here.

So far the worst problem I have seen is the crazy amount of Wild Pitches... The are having a big impact on the games, and by aces. have you see this aswell Nomo?

Yeah, I think those monster HRs is fairly an issue people are already talking about loudly, but I do agree that rampant wild pitches appears likely to be an even bigger issue... I sayy bigger since I don't know if Pitcher Consistency slider is sensitive enough to curb them down enough to a realistic level.

I'm currently using Pitcher Consistency 7 to hopefully see a reduced WPs, but so far it doesn't seem to be doing much. I wonder the slider is effective enough to affect WPs this year...

[nomo17k]

Re: Stats-Based CPU Sliders [MLB 17 Version]

Just a quick update of what I'm testing now away from the default sliders, the 4/4 tab in the Google sheet:

• Contact: 6 - This is primarily for making CPU swing & miss less, and also increase batting average a bit.
• Power: 4 - A rather conservative nerfing of power to reduce HRs a bit.
• Starter Stamina: 7 - Trying to see if giving starting pitchers a little more stamina helps reducing quick hooks
• Pitcher Control: 3 - Making pitchers a little more hittable by giving them less command of their pitches
• Pitcher Consistency: 7 - I want to reduce wild pitches by making pitchers throw less errant pitches.
• Manager Hook: 3 - To help starters stay longer in the game. This unfortunately has a side effect of CPU managers leaving relievers a bit too long though
• Pitch Speed: 10 - For more true-to-life pitch speed
• Fielding Errors In: 0 - I tested at 8 briefly (25 games or so in 4/3 tab) and surprisingly only saw a tiny amount of errors... just checking if this slider is somehow reversed or not.
• Fielding Errors Out: 6 - Just a tiny adjustment for more realistic value
• Throwing Errors Out: 8 - Just a tiny adjustment for more realistic value
• Fielder Run Speed: 4 - Trying to increase doubles a little
• BR Steal Ability: 10 - This adjustment is in conjunction with the max pitch speed, which makes stealing harder.
• Wind: 0 - I don't want to typically do this since I like having variety, but I'm totally turning off wind effects to see what effect it has on the HR numbers. In the past, stronger winds tended to increase HRs.

[miamikb2001]

Re: Stats-Based CPU Sliders [MLB 17 Version]

Loving that your back at Nomo.Definately agree WP are an issue no game killer but it'd be nice to get it lower for sure.Been working set and getting good results but i will definately do some testing with your's to see how they compare.


Sent from my iPhone using Tapatalk

[floydpinkster]

Re: Stats-Based CPU Sliders [MLB 17 Version]

I notice that a lot of the WP's are just balls in the dirt in front of the catcher or balls that don't roll very far and in real life a base runner would not run try to advance on... or be gunned down if they tried but, in this game, the runners are attempting to advance on almost all balls blocked in front of or around the catcher and are easily successful.

I know It would probably have effects on other aspects of the game but would base runner speed affect this problem area of the game???
Lower base runner spread my make the runner less apt to attempt to advance on these types of close to the catcher situation perhaps???

[tgreer]

Re: Stats-Based CPU Sliders [MLB 17 Version]

Looking back at the first set I ran through I used your error sliders from last year.

Fld Errors In: 9
Fld Errors Out: 3
Thr Errors In: 5
Thr Errors Out :8

23 Games in committed 24 Errors (0.52) not bad. I did not check the replay vault on all of these but I did on a bunch. Most were of the infield ground ball variety which is good. Anyway just wanted to throw this your way if you hadn't checked the fielding sliders from last year on this years game.

[statdude]

Re: Stats-Based CPU Sliders [MLB 17 Version]

Hey Nomo, I used your sliders last year in my MoM franchise for all 162 games. I put together a spreadsheet with all the stats that my players had over the season if you want it. No idea how relevant it would be, but I'd thought I'd offer it.

One thing that I noticed throughout the entire season in 2016 was an over reliance on Fastballs if a pitcher had it as their "X" pitch. This problem was compound even further when they would have a 4SFB as their "X" and a 2SFB as their "O". This would lead to pitchers essentially not wanting to throw their breaking balls and off speed pitches. I noticed this the most with Sonny Gray where he would not want to throw his CB or SL. Gray would throw his FB (4SFB and 2SFB) over 75% of all pitches.

I may have negated this effect a bit by swapping their "X" and "O" pitches to make sure a breaking pitch was their "X" pitch. I didn't do it for Gray, but I did for Kluber who had a similar problem. I made his "X" pitch a Slurve and "O" pitch a 4SFB. This led to Kluber relying on his Slurve more often, but not too much. There were games where he threw more Slurves than 4SFB, but Kluber made it up by throwing 2SFB as well. Essentially, it upped Kluber's Slurve up to real life usage %. It did not do it for every breaking ball and off speed pitch, but it did curb the high usage % of FBs.

Just some thoughts.