An analysis of default sliders vs real life

[tinpanalley]

For whatever it's worth, for your consideration, for a "hmm, interesting" if that's all it provides for you, I thought I'd share my sports games method for defining sliders that I've been doing for years...

I do this every year with my sports games (most often NHL, The Show, FIFA, although I've done it with others) and it's been helpful for getting realistic stats. It yielded some interesting observations this year though with MLB 14.

Basic overview of my method - Essentially I leave the sliders at default and simulate a handful of seasons. Usually 4 or 5. I collect the overall team stats and the individual leaders, by which I mean, I take photos of the screen (to not waste too much time). I then compare those stats with the real stats and averages from the previous few real life seasons. Yes, I know there will be certain fluctuations, but this is for an overview, not for perfection. If I wanted to do that, I'd have to go far deeper than just comparing totals and averages.
Once I get a sense of how "off" the simulated stats are compared to what happened in real life, it gives me a sense of how to move the sliders. Honestly, every season, it gets me realistic stats.

What happened this year?
I was using the recently released OSFM V1 Franchise File released by ILLICIT206 off of Knight165's Full Minors set. Great franchise file, amazing detail in prepping each team's MLB rosters. Just amazing work. The simulations, on default sliders, yielded the following results consistently, year after year. In the majority of the statistics the highs and lows were higher and lower (respectively) in the game by about 10% in each direction. But these were the interesting ones...

<!--[if gte mso 9]><xml> <o:OfficeDocumentSettings> <o:AllowPNG/> </o:OfficeDocumentSettings> </xml><![endif]--> TEAM HITTING
SB and CS - low by 10-20 every year
GIDP – too low on every team by 40

TEAM PITCHING
Complete Games – 26-9 (most to least) sim per team, 9-1 real per team.
SHO – 18-3 (most to least) sim, 4-0 real.

TEAM FIELDING
A – 1537-1365 (most to least) sim, 1924-1500 real.

PLAYERS
Essentially, all extra base hits are too low in simulation seasons.

PITCHERS
ERA – game ERAs need to be worse by .5 points per pitcher
HR – game pitchers all need to give up 5-10 more
SHO – game pitchers have too many by about 30%
CG – game pitchers get twice as many across the board as in real life
IP – on average, game pitchers last 2 full innings more per game than they should
WHIP – should be higher in the game by .10

What's the point of this?
To give you a sense of what happened consistently over 5 simulation seasons and to give you a rough (not exact) sense of how the game plays out of the box, with that roster/franchise file, on default, and only allowing the game to calculate results so that you can use that in moving your own sliders around.

I created a slider set for myself that reflects what I think gets these simulated stats more in line (with some margin for error, of course) with real life numbers. I had notes for way more stats but these were the ones with big variations from real life. Overall, on default, the game DOES produce relatively realistic stats. This is just for those of you, perhaps like me, who like to really play with the realism factor.

Hope this was at least interesting, if nothing else.

[kcsam]

Re: An analysis of default sliders vs real life

For whatever it's worth, for your consideration, for a "hmm, interesting" if that's all it provides for you, I thought I'd share my sports games method for defining sliders that I've been doing for years...

I do this every year with my sports games (most often NHL, The Show, FIFA, although I've done it with others) and it's been helpful for getting realistic stats. It yielded some interesting observations this year though with MLB 14.

Basic overview of my method - Essentially I leave the sliders at default and simulate a handful of seasons. Usually 4 or 5. I collect the overall team stats and the individual leaders, by which I mean, I take photos of the screen (to not waste too much time). I then compare those stats with the real stats and averages from the previous few real life seasons. Yes, I know there will be certain fluctuations, but this is for an overview, not for perfection. If I wanted to do that, I'd have to go far deeper than just comparing totals and averages.
Once I get a sense of how "off" the simulated stats are compared to what happened in real life, it gives me a sense of how to move the sliders. Honestly, every season, it gets me realistic stats.

What happened this year?
I was using the recently released OSFM V1 Franchise File released by ILLICIT206 off of Knight165's Full Minors set. Great franchise file, amazing detail in prepping each team's MLB rosters. Just amazing work. The simulations, on default sliders, yielded the following results consistently, year after year. In the majority of the statistics the highs and lows were higher and lower (respectively) in the game by about 10% in each direction. But these were the interesting ones...

TEAM HITTING
SB and CS - low by 10-20 every year
GIDP – too low on every team by 40

TEAM PITCHING
Complete Games – 26-9 (most to least) sim per team, 9-1 real per team.
SHO – 18-3 (most to least) sim, 4-0 real.

TEAM FIELDING
A – 1537-1365 (most to least) sim, 1924-1500 real.

PLAYERS
Essentially, all extra base hits are too low in simulation seasons.

PITCHERS
ERA – game ERAs need to be worse by .5 points per pitcher
HR – game pitchers all need to give up 5-10 more
SHO – game pitchers have too many by about 30%
CG – game pitchers get twice as many across the board as in real life
IP – on average, game pitchers last 2 full innings more per game than they should
WHIP – should be higher in the game by .10

What's the point of this?
To give you a sense of what happened consistently over 5 simulation seasons and to give you a rough (not exact) sense of how the game plays out of the box, with that roster/franchise file, on default, and only allowing the game to calculate results so that you can use that in moving your own sliders around.

I created a slider set for myself that reflects what I think gets these simulated stats more in line (with some margin for error, of course) with real life numbers. I had notes for way more stats but these were the ones with big variations from real life. Overall, on default, the game DOES produce relatively realistic stats. This is just for those of you, perhaps like me, who like to really play with the realism factor.

Hope this was at least interesting, if nothing else.

This was very interesting to say the least. Have you run season simmulations again now that you have altered the sliders based off your results from the out of the box default sliders? I'd love to see those and in turn see which sliders you are currently using going forward. I'm very much in favor of playing with the realism factor as you are. Thanks.

Update - As far as I know, when you sim games the sliders are not taken into consideration. So I am not sure that altering the sliders would have any affect on just simming seasons. Are you seeing different results? Did you run any cpu vs cpu games and track the numbers you explained above?

[itbeme23]

Re: An analysis of default sliders vs real life

This was very interesting to say the least. Have you run season simmulations again now that you have altered the sliders based off your results from the out of the box default sliders? I'd love to see those and in turn see which sliders you are currently using going forward. I'm very much in favor of playing with the realism factor as you are. Thanks.

Update - As far as I know, when you sim games the sliders are not taken into consideration. So I am not sure that altering the sliders would have any affect on just simming seasons. Are you seeing different results? Did you run any cpu vs cpu games and track the numbers you explained above?

I'm curious about this as well, because I would like to have a differnt slider set for simmed games as opposed to the games I actually play in my Dodgers franchise (I use 30 team control). The rule of thumb (as far as I know) when using a slider set specifically for simmed games is to:

Lower Pitcher Stamina - to achieve a realistic amount of innings pitched for starters, and to avoid the overuse of relievers.

Raise Injury Slider - I know that Nomo2K gets pretty heavy into this in his CPU vs. CPU thread, but simply put, there are not a realistic amount of injuries, severity, or DL times when simming games with the Injury slider at default. Nearly every thread I have read has said as such. Usually, that slider needs to be somewhere in the 8-10 range in order to produce realistic results. I personally use 8 for my actual games, and I think I'm going to stick with that for simmed games as well. I think Nomo suggests setting the slider at 9 or 10 for the most realistic results for simmed games, but personally, I'm a little weary of setting the slider that high.

I barely started my Dodgers franchise a couple days ago with the OSFM rosters, and I was using my slider set for both games played and sim games. I'm thinking that can't be possible, as I would like to have realistic simmed stats at the end of the season. So I'll probably revert back to default sliders with the above mentioned adjustments for simmed games.

If anybody has a difinitive answer on whether or not sliders actually affect simmed stats, I'm sure that would be a big help.

[tinpanalley]

Re: An analysis of default sliders vs real life

Honestly, it never even occurred to me that it wouldn't. I've always seen results that I thought were based on changing sliders. Maybe it was all in my head? I'm sure it works in other sports games.

[stealyerface]

Re: An analysis of default sliders vs real life

So, in theory, you could look at the slider sets available, load those sliders up, and do the SIMs that you do, and see if a particular set garnered the close-to-life results versus the default?

When I make my sliders, I too play (not sim) a handful of games to see the trends, stats, and overall feel of the game as it plays out in real life. Then, I adjust the game to match the timing, the look, and the pace of the real thing.

The trouble starts when you realize that the fielder arm strength on default is too high. So you fix that. But then, you realize that with the arm strength down, the runners are beating out throws in the infield they never would. Then you adjust the runner speed down, but start seeing troubles with extra base hits, so you need to bring the fielder speed and reaction time down... and the beat goes on and on and on.

So finally, you end up with a slider set where all the cogs and working gears in the machine start to work together, and then you get the visual results you want. That is a ball deep in the hole at short stop results in a runner being out by a step if the SS has a great arm and the runner is fairly quick. If the SS has a sub par arm, and the runner has +speed, he beats it out, or the play goes either way. If A.J. is running to first, the SS can have a fair arm, and still make the play with a pre-loaded throw, and get him by the step and a half he would be out with in real life.

So when folks look at the sliders and think, "Well, this guy has no idea what he is doing with the runner speed down so low", that is because all the other parts of the machine have been adjusted to make that decision make sense. My theory is that if the game mimics the "look" of the game, meaning hits that should be hits, running catches that should be made, and extra base hits having a risk reward that plays out like a real game, the stats should be pretty close, because the game's speed and flow are choreographed like the actual game.

The one thing I love is when a play happens, and I say to myself, "That is EXACTLY how it would have played out in real life". Ball hit on the ground to short, step throw, out by a step. Ball hit to second with a man at third, but not playing the defense in, throw to home is late by a slide. Ball hit to the triangle at the Fens, but Napoli running, stand up double while Middlebrooks might stretch for the triple, and Victorino has an easy triple.

The cadence of the plays are what are most important to me, so that the timing of runner and fielder and batted ball play out like the game's real speed. And I believe once that concoction of sliders are tuned to the timing of the game, the Stats should play out accordingly.

~syf

[nomo17k]

Re: An analysis of default sliders vs real life

...

Raise Injury Slider - I know that Nomo2K gets pretty heavy into this in his CPU vs. CPU thread, but simply put, there are not a realistic amount of injuries, severity, or DL times when simming games with the Injury slider at default. Nearly every thread I have read has said as such. Usually, that slider needs to be somewhere in the 8-10 range in order to produce realistic results. I personally use 8 for my actual games, and I think I'm going to stick with that for simmed games as well. I think Nomo suggests setting the slider at 9 or 10 for the most realistic results for simmed games, but personally, I'm a little weary of setting the slider that high.

...

I don't touch the Injury Frequency slider for my CPU vs. CPU set just because it's not realistic for me to get in enough non-sim games to nail things down to a satisfactory level.

I think BrianU (or somebody else) has compiled an extensive list of injuries/DL info the last year or so, so I would use that sort of information if I were to make any changes (at all).

Honestly, it never even occurred to me that it wouldn't. I've always seen results that I thought were based on changing sliders. Maybe it was all in my head? I'm sure it works in other sports games.

For The Show, sim and non-sim games are run on different game engines. Only a few sliders affect both modes (Manager Hook, Stamina-related, and Injury Frequency). That's because the gameplay actually simulates processes (physics, etc.) but the sim mostly only simulates results.

[tinpanalley]

Re: An analysis of default sliders vs real life

For The Show, sim and non-sim games are run on different game engines. Only a few sliders affect both modes (Manager Hook, Stamina-related, and Injury Frequency). That's because the gameplay actually simulates processes (physics, etc.) but the sim mostly only simulates results.

Well, then in that case, my test was pretty useless I guess. I know it works in Madden and NHL and FIFA. Maybe it's just EA games. Sorry to waste anyone's time.

[stealyerface]

Re: An analysis of default sliders vs real life

I believe that if you loaded up a slider set, played one game with them, then simulated your season to the end of the year, you would see more home runs with the Power Slider set to ten than if it were set to zero.

Is this not true?

~syf

[Heroesandvillains]

Re: An analysis of default sliders vs real life

I believe that if you loaded up a slider set, played one game with them, then simulated your season to the end of the year, you would see more home runs with the Power Slider set to ten than if it were set to zero.

Is this not true?

~syf

No.

The power slider has no impact during simmed games.

[ParisB]

Re: An analysis of default sliders vs real life

You might also see less homeruns if you increase Power.

Sliders aren't so absolute. Statistics vary for each player and each team, both up and down, game to game, week to week, season to season.

[the_riot]

Re: An analysis of default sliders vs real life

Only a few sliders affect both modes (Manager Hook, Stamina-related, and Injury Frequency).

Wow. I had no idea any sliders affected sim results. I have manager hook lowered a couple clicks because it seems like the cpu manager has a short leash, but I may have to try and put it back to default due to this revelation.

[dpower15]

Re: An analysis of default sliders vs real life

So which sliders actually influence simulated games? It has always bothered me that when I play I need starter stamina around 7 to get realistic pitch counts without getting lit up, but never thought that slider was the reason for the abundance of complete games and shutouts that I see as i progress through a season.

So for simmed games if I turned the stamina down to say, 3-4, and brought manager hook up a few clicks, that should reduce the number of complete games and shutouts I see over a season?


Sent from my iPad using Tapatalk