Any proof that HR/9 does nothing?

[cardinalbird5]

Re: Any proof that HR/9 does nothing?

Yeah it is an average would be my guess. Never look at the overall. Look at the individual ratings.

[Cozar]

Re: Any proof that HR/9 does nothing?

                    rating           sim                gameplay   
Player Name          HR/9     HR     IP   HR/9     HR     IP   HR/9
-------------------------------------------------------------------
Josh Johnson           80     38  197.0  1.736      7   54.3  1.160
Clayton Kershaw        79     20  205.0  0.878      7   62.0  1.016
C.J. Wilson            78     22  216.0  0.917      6   66.3  0.814
Jaime Garcia           78     24  221.3  0.976      5   53.7  0.839
Tim Lincecum           78     26  224.3  1.043      6   70.7  0.764
Ubaldo Jimenez         78     22  221.0  0.896      6   71.7  0.753
Zack Greinke           77     24  242.0  0.893      5   75.7  0.595
Chris Carpenter        77     24  218.7  0.988      5   67.0  0.672
James McDonald         76     19  190.0  0.900      5   60.3  0.746
Mike Pelfrey           75     31  209.0  1.335     11   59.0  1.678
-------------------------------------------------------------------
TOTAL:                       250 2144.3  1.049     63  640.7  0.885

                    rating           sim                gameplay   
Player Name          HR/9     HR     IP   HR/9     HR     IP   HR/9
-------------------------------------------------------------------
Andrew Brackman        48     21  183.3  1.031      5   58.7  0.767
Jordan Zimmermann      48     28  196.0  1.286      4   59.0  0.610
Tommy Hunter           47     29  208.3  1.253     10   61.0  1.475
Aaron Harang           47     30  201.7  1.339      7   61.7  1.022
Rich Harden            47     26  196.7  1.190      5   52.7  0.854
Armando Galarraga      45     27  192.7  1.261      5   48.7  0.925
Ted Lilly              45     17  219.3  0.698      8   62.3  1.155
Kevin Slowey           45     27  207.3  1.172      6   58.0  0.931
Jake Arrieta           42     35  202.3  1.557     11   60.3  1.641
Trevor Cahill          40     23  219.0  0.945     13   65.3  1.791
-------------------------------------------------------------------
TOTAL:                       263 2026.7  1.168     74  587.7  1.133

I'm going to go out on a limb and wager that the top group of pitchers you posted not only have better HR/9 ratings than the bottom group, they also have much better K/9, BB/9 and H/9 ratings.

Based on Brian's answer, it is those better K/9, BB/9 and H/9 ratings that largely explain why they give up fewer HR/9 in gameplay situations. So while the actual gameplay HR/9 and rated HR/9 might show a correlation, to prove causation you need to edit the pitchers so they all have the same ratings in other categories (and pitch types).

[Rs31632]

Re: Any proof that HR/9 does nothing?

MAYBE ITS ME BUT ONLY ONE THING MATTERS...

THROWING THAT HEAT!!!!!!!...... I LOVE GUYS W/ 95+ VELOCITY

Nate Jones is the man

[nomo17k]

Re: Any proof that HR/9 does nothing?

I'm going to go out on a limb and wager that the top group of pitchers you posted not only have better HR/9 ratings than the bottom group, they also have much better K/9, BB/9 and H/9 ratings.

Based on Brian's answer, it is those better K/9, BB/9 and H/9 ratings that largely explain why they give up fewer HR/9 in gameplay situations. So while the actual gameplay HR/9 and rated HR/9 might show a correlation, to prove causation you need to edit the pitchers so they all have the same ratings in other categories (and pitch types).

What you are saying may also be true (i.e., overall good pitchers tend to allow less HRs), but what I am hoping is that the gameplay engine is actually reproducing low HR/9 performance by simulating how they tend to keep the ball in the park...

It is in general true that ground-ball pitchers tend to allow less HRs in real life. Therefore, if the game is actually simulating those players as ground-ball pitchers by assigning pitches like sinkers/2SFB which tend to induce grounders, AND the game is actually simulating those pitches well enough (in inducing ground balls), then those low HR/9 performances are the result of proper simulation... (rather than an ad-hoc attribute like HR/9). That's what I'm hoping.


Although HR/9 rating isn't the cause of HR/9 in regular gameplay, it's entirely reasonable to assign good HR/9 ratings to those pitchers who allow less HRs because of their pitch repertoire, for example... so that they allow less HRs in both sim and gameplay engines. I just mistook what really is causing the performance difference... it's a kind of mistake people, even professional statisticians, make all the time (though I'm not trying to make an excuse...)...

... it's like saying eating a lot of meat makes you fat.... when what really is making you fat may just be the fact that you are eating too much of anything in general (or even just something else... like soft drinks). causation cannot be inferred from correlation. my excuse ends now...

[cardinalbird5]

Re: Any proof that HR/9 does nothing?

                    rating           sim                gameplay   
Player Name          HR/9     HR     IP   HR/9     HR     IP   HR/9
-------------------------------------------------------------------
Josh Johnson           80     38  197.0  1.736      7   54.3  1.160
Clayton Kershaw        79     20  205.0  0.878      7   62.0  1.016
C.J. Wilson            78     22  216.0  0.917      6   66.3  0.814
Jaime Garcia           78     24  221.3  0.976      5   53.7  0.839
Tim Lincecum           78     26  224.3  1.043      6   70.7  0.764
Ubaldo Jimenez         78     22  221.0  0.896      6   71.7  0.753
Zack Greinke           77     24  242.0  0.893      5   75.7  0.595
Chris Carpenter        77     24  218.7  0.988      5   67.0  0.672
James McDonald         76     19  190.0  0.900      5   60.3  0.746
Mike Pelfrey           75     31  209.0  1.335     11   59.0  1.678
-------------------------------------------------------------------
TOTAL:                       250 2144.3  1.049     63  640.7  0.885

                    rating           sim                gameplay   
Player Name          HR/9     HR     IP   HR/9     HR     IP   HR/9
-------------------------------------------------------------------
Andrew Brackman        48     21  183.3  1.031      5   58.7  0.767
Jordan Zimmermann      48     28  196.0  1.286      4   59.0  0.610
Tommy Hunter           47     29  208.3  1.253     10   61.0  1.475
Aaron Harang           47     30  201.7  1.339      7   61.7  1.022
Rich Harden            47     26  196.7  1.190      5   52.7  0.854
Armando Galarraga      45     27  192.7  1.261      5   48.7  0.925
Ted Lilly              45     17  219.3  0.698      8   62.3  1.155
Kevin Slowey           45     27  207.3  1.172      6   58.0  0.931
Jake Arrieta           42     35  202.3  1.557     11   60.3  1.641
Trevor Cahill          40     23  219.0  0.945     13   65.3  1.791
-------------------------------------------------------------------
TOTAL:                       263 2026.7  1.168     74  587.7  1.133

Now that Brian@SCEA gave a definitive answer I'm not going to argue otherwise, but it's kinda interesting so I post the numbers that I meant to cite anyways. It's a good exercise of when you can get fooled by statistics as well.

You see that according to the numbers above, high HR/9 pitchers gave up less HRs in gameplay overall. That's how I interpreted the attribute... (on the other hand, in simmed games you see even less effect of HR/9 attribute.... at the time I thought it was a bug/game balance issue).

But now that we *know* that HR/9 doesn't have any effect in gameplay... why did I think I saw what I saw (that is, HR/9 attributes have something to do with less HRs given up by high HR/9 pitchers)?

I think the reason could be (a) the sample size/player variation wasn't big enough, so I was just seeing a biased result and/or (b) HR/9 attribute in fact does correlate with those pitchers ability to hold/give up HRs, but what I was seeing was a correlation, not a causation.

I'm actually wishing the reason is (b), because SCEA has a reason to give those guys high HR/9 ratings... because they gave up less HRs in real life. But if the reason for their performance is not due to high HR/9 ratings, then it may be because they are given proper pitch repertoire which leads to less HRs... pitchers with pitches that sink (like sinker) tend to give up more ground balls and less fly balls, so they tend to give up less HRs.

If the reason why they gave up less HRs is indeed due to their pitch repertoire, I think that's a very nice thing.... the pitches in the game is doing what they are suppose to do.

Nomo it also looks like the gameplay results are much more random. Once you take into account that the top half are just better pitchers with better attributes across the board then it makes sense they give up less HR's, but the actual results in gameplay look to vary quite a bit, regardless of their HR/9 rating.

You'd have to make every other attribute like K/9, H/9, BB/9, control, break, velocity, clutch, etc. all the same and then vary HR/9 to see if there is a difference.

[KBLover]

Re: Any proof that HR/9 does nothing?

Do pitchers have a GB/FB rate, or are all pitchers basically 50/50? Same for hitters?

Does this also mean H/9 is basically the only attribute the matters a lot? If you can't get a hit, you can't get a HR - that's pretty much the game's thinking vs a more DIPS-style thought process where BB/K/HR is determined (Ball in play or not) and then, if there's contact, a hit is possible based on velocity and trajectory of the batted ball and the range and reaction of the fielders?

Does it also mean I'm wasting my time trying to get my sliders to produce reasonable game-to-game BABIPs as well as trying to get contact ratings to heavily influence BABIP?

[Bobhead]

Re: Any proof that HR/9 does nothing?

Do pitchers have a GB/FB rate, or are all pitchers basically 50/50? Same for hitters?

Neither.

Pitchers don't have an assigned GB rate, but every pitch has its own tendencies by nature. Curveballs tend to be hit in the air. Sinking pitches tend to be hit on the ground.

Thus, a pitcher with multiple sinking pitches (sliders, sinkers, etc...) will have a higher GB rate than a pitcher who specializes in his curveball.

Does it also mean I'm wasting my time trying to get my sliders to produce reasonable game-to-game BABIPs as well as trying to get contact ratings to heavily influence BABIP?

Definitely not. I know it's possible because I've done it. You just might be going in the wrong direction or tweaking the wrong sliders. Or maybe you're even calculating BABIP the wrong way. Seems silly but because of how some of The Show's stats are stored and labeled, it's something you might want to double check. I know I've made a few calculation mistakes already this year.

[nomo17k]

Re: Any proof that HR/9 does nothing?

Nomo it also looks like the gameplay results are much more random. Once you take into account that the top half are just better pitchers with better attributes across the board then it makes sense they give up less HR's, but the actual results in gameplay look to vary quite a bit, regardless of their HR/9 rating.

You'd have to make every other attribute like K/9, H/9, BB/9, control, break, velocity, clutch, etc. all the same and then vary HR/9 to see if there is a difference.

This would be true, but I was hoping by grouping several pitchers with very high and very low HR/9 ratings, I could at least see *some* effect from HR/9 alone. Likely I was too optimistic, since as you correctly point out the group with high HR/9 has more quality pitchers than the other group in this case. I should've been more careful with this analysis. But the table probably tells us that HR/9 attribute doesn't have a very drastic effect on HR/9 stats in this game though.

[nomo17k]

Re: Any proof that HR/9 does nothing?

Do pitchers have a GB/FB rate, or are all pitchers basically 50/50? Same for hitters?

Does this also mean H/9 is basically the only attribute the matters a lot? If you can't get a hit, you can't get a HR - that's pretty much the game's thinking vs a more DIPS-style thought process where BB/K/HR is determined (Ball in play or not) and then, if there's contact, a hit is possible based on velocity and trajectory of the batted ball and the range and reaction of the fielders?

Does it also mean I'm wasting my time trying to get my sliders to produce reasonable game-to-game BABIPs as well as trying to get contact ratings to heavily influence BABIP?

Adding to what Bobhead wrote (though I don't necessarily with the curveball interpretation... I tend to be early on low curveballs so I hit them into a lot of groundballs)...

I think Bobhead is correct about the pitch types having intrinsic tendencies to induce certain types of batted ball... I'd also add pitch locations into mix, because PCI tends to be located round the mid level (for the well-rounded strike zone coverage), pitches down in the zone tend to be ground balls, up tend to be fly balls because how balls travel off the bat correlate with pitch location and how PCI is placed relative to it. You can clearly see this tendency in pitcher/batter analysis and apply ground/fly ball filters.

Not sure what you exactly mean by DIPS-style thinking, but the gameplay engine is very heavily physics oriented, rather than dice roll oriented. (The sim engine probably is heavily more dice roll oriented...).

Rather than predetermining the outcomes (e.g., the game making a hitter ground out to shortstop just because it's time for him to do so based on situations), the gameplay engine only derives the initial speed and trajectory of the batted ball from what the game defines as the quality of contact. The quality of contact is largely the composite of user skills (timing and PCI location), Contact rating, the opposing pitcher's H/9 rating. Once that's determined, the standard ball physics takes over and fielders independently decide how to deal with what's going on with the ball depending on the situation. Just as in real life basically.

BBs and Ks in the gameplay, especially the former, are almost entirely up to the user's approach. You can influence Ks by increasing Vision; you can stay alive by fighting off pitches after pitches that you would otherwise miss.

[cardinalbird5]

Re: Any proof that HR/9 does nothing?

I move my PCI around a lot and I try to hit homeruns often as well. Usually the PCI doesn't lie for me. If I am under the ball on early timing...I fly out to 3b/ss/lf. If I am late and over the ball...I ground out to 2b and so on.

I just notice weird things though against most of my opponents who don't like to move their PCI. I've seen them hit oppo grounders sharply on a pitch high and inside where the ball is touching the top of the PCI. I've also seen the opposite. It just seems like anything near the middle of the plate will always give them a cheap hit 1 out of 3 times, regardless of their timing and pci placement. I also see my opponents hit some cheap homeruns where I jam them middle in where the ball is clearly above the PCI and they did not use power swing.......

So I guess my question is....why is this and how big of a role does influencing swings have? I influence almost every pitch, but I don't know how much it helps. I do hit a lot of fly balls though, as I try to.