Basic order of operations question

[ahunter5]

Hey all - I know this has been answered, but I'll be damned if I can find it via search.


What's the order that is used to determine pitcher/hitter outcome?


Does it look for strikeout or walk first?


BABIP, then Power, then Gap? Meaning, the system determines that there will be a hit of some kind, then whether it's a HR, then, if not a HR, whether it will be a double or triple?

[Syd Thrift]

I'm pretty sure RonCo looked into this a few years ago and IIRC it does look for the 3 true outcomes first before it determines where a ball in play goes, what chances it has of getting caught, etc. By my own play I *think* Ks are determined before walks because lowering K rates seems to cause walk rates to go up on their own, but I'm not 100% positive. The 3 True Outcomes thing also applies to HRs: it does the K/BB/HR checks first and then if the result of the at-bat is neither of those 3, it goes into other possible results.

The big thing to remember of course (which isn't what you asked but I wanted to be clear) is that the game is outcome-based, not physics-based. The advanced stats are fun and a great way to be able to play without ratings turned on but they're also determined after the fact and don't "cause" the outcomes either.

[Todd R]

Originally Posted by Syd Thrift

The big thing to remember of course (which isn't what you asked but I wanted to be clear) is that the game is outcome-based, not physics-based. The advanced stats are fun and a great way to be able to play without ratings turned on but they're also determined after the fact and don't "cause" the outcomes either.

That is widely misunderstood. If OOTP were food the ingredient list would say "Advanced Stats (Artificial flavor. No nutritional value.)"

[locuspc]

Good god man. Give it a rest.

[cwhitman]

Originally Posted by Syd Thrift

I'm pretty sure RonCo looked into this a few years ago and IIRC it does look for the 3 true outcomes first before it determines where a ball in play goes, what chances it has of getting caught, etc. By my own play I *think* Ks are determined before walks because lowering K rates seems to cause walk rates to go up on their own, but I'm not 100% positive. The 3 True Outcomes thing also applies to HRs: it does the K/BB/HR checks first and then if the result of the at-bat is neither of those 3, it goes into other possible results.

The big thing to remember of course (which isn't what you asked but I wanted to be clear) is that the game is outcome-based, not physics-based. The advanced stats are fun and a great way to be able to play without ratings turned on but they're also determined after the fact and don't "cause" the outcomes either.

i've been under the impression from others that it went BB, then K/HR, but I've seen some weird things in the last year that make me less confident that that's the case, like last year feeling like the decision of GB vs FB was impacting homer rates, so that those pitchers with the same HRA but a higher FB rate allowed more homers. that seems to be gone this year though.

[kq76]

I never understood why people think the game must figure out K before BB, or BB before K, or whatever. Or why it is even important.

What seems to make sense to me would be each batter-pitcher matchup would have a different set of K, BB, HR, BIP %s, the game would come up with a random # (say a decimal between 0 to 1) and that # would correspond to one of the %s (so if say the 4 were equally likely (25%) and the # came up .94, then it would fall between .75 and 1 which would be the BIP%) and if necessary it would then run another call (is it a hit or an out) and another, etc (is it a single or a double, etc). I understand my example is probably incredibly reductive, but that's just how I imagine it.

Is there a reason why it must be figuring out one stat at a time, instead of figuring it out in groups? And why is it important? Educate me on this, please.

[x McLovin x]

Originally Posted by Syd Thrift

I'm pretty sure RonCo looked into this a few years ago and IIRC it does look for the 3 true outcomes first before it determines where a ball in play goes, what chances it has of getting caught, etc. By my own play I *think* Ks are determined before walks because lowering K rates seems to cause walk rates to go up on their own, but I'm not 100% positive. The 3 True Outcomes thing also applies to HRs: it does the K/BB/HR checks first and then if the result of the at-bat is neither of those 3, it goes into other possible results.

The big thing to remember of course (which isn't what you asked but I wanted to be clear) is that the game is outcome-based, not physics-based. The advanced stats are fun and a great way to be able to play without ratings turned on but they're also determined after the fact and don't "cause" the outcomes either.

Originally Posted by Todd R

That is widely misunderstood. If OOTP were food the ingredient list would say "Advanced Stats (Artificial flavor. No nutritional value.)"

I’m aware of this although I have a specific question:
Suppose for example there is a LHB with extreme pulled-flyball tendencies who gets traded from a team which plays in a relatively neutral ballpark to the Yankees. Would it just be park factors that would account for the expected increase in home runs? Or does the fact that this specific player has that batted ball profile also magnify the increase?

Regardless of design, it would seem very illogical (and disappointing) for the game to not consider both.

[ahunter5]

Originally Posted by Syd Thrift

I'm pretty sure RonCo looked into this a few years ago and IIRC it does look for the 3 true outcomes first before it determines where a ball in play goes, what chances it has of getting caught, etc. By my own play I *think* Ks are determined before walks because lowering K rates seems to cause walk rates to go up on their own, but I'm not 100% positive. The 3 True Outcomes thing also applies to HRs: it does the K/BB/HR checks first and then if the result of the at-bat is neither of those 3, it goes into other possible results.

The big thing to remember of course (which isn't what you asked but I wanted to be clear) is that the game is outcome-based, not physics-based. The advanced stats are fun and a great way to be able to play without ratings turned on but they're also determined after the fact and don't "cause" the outcomes either.

Thanks Syd - I appreciate this. I knew I had seen this somewhere in the past, but could not find it.

[ahunter5]

Originally Posted by kq76

I never understood why people think the game must figure out K before BB, or BB before K, or whatever. Or why it is even important.

What seems to make sense to me would be each batter-pitcher matchup would have a different set of K, BB, HR, BIP %s, the game would come up with a random # (say a decimal between 0 to 1) and that # would correspond to one of the %s (so if say the 4 were equally likely (25%) and the # came up .94, then it would fall between .75 and 1 which would be the BIP%) and if necessary it would then run another call (is it a hit or an out) and another, etc (is it a single or a double, etc). I understand my example is probably incredibly reductive, but that's just how I imagine it.

Is there a reason why it must be figuring out one stat at a time, instead of figuring it out in groups? And why is it important? Educate me on this, please.

I am not great at this game, and I am far from a mathematician or statistician. But in my head, it should matter just to assign importance to different stats.


For example, if you knew that a player had a rating that caused him to strike out 50% of the time that the game checked for strikeouts, you'd probably find it valuable to know when that check was made in the process. If it was first, that means he's striking out 50% of his plate appearances. If it checked walks first, and his eye rating gave the player a walk rate of 20%, then he's only going to strike out 40% of his plate appearances.


If you're suggesting that the game figures out the likelihood for all outcomes at the same time, sort of like a Strat-o-Matic card, that would be interesting. I hadn't read that take anywhere before.

[Reed]

Originally Posted by x McLovin x

I’m aware of this although I have a specific question:
Suppose for example there is a LHB with extreme pulled-flyball tendencies who gets traded from a team which plays in a relatively neutral ballpark to the Yankees. Would it just be park factors that would account for the expected increase in home runs? Or does the fact that this specific player has that batted ball profile also magnify the increase?

Regardless of design, it would seem very illogical (and disappointing) for the game to not consider both.

I have not seen it spelled out but almost positive it will be magnified to some extent. If playing at Yankee Stadium, you want lefty batters or players that hit primarily down right field line to take advantage of short fence.

[Todd R]

Originally Posted by x McLovin x

I’m aware of this although I have a specific question:
Suppose for example there is a LHB with extreme pulled-flyball tendencies who gets traded from a team which plays in a relatively neutral ballpark to the Yankees. Would it just be park factors that would account for the expected increase in home runs? Or does the fact that this specific player has that batted ball profile also magnify the increase?

Regardless of design, it would seem very illogical (and disappointing) for the game to not consider both.

As Syd posted the game is outcome based not physics based. Pulled FB is physics. Batted ball profiles for real players is not based on data. It is created by the game and is not always the same for the same player.

This can be confirmed by importing a historical player several times and comparing the batted ball profile assigned to the different imports. It will vary. It's cosmetic. Players perform the same regardless of batted ball profile.

I wouldn't call this disappointing. Since real data is not available allowing created information to affect output would result in less realistic players.

[EZB]

1. Is it a walk?
2. Is it a HR?
3. Is it a K?
4. Is it a hit?
5. Is it a 2B/3B?
6. Is it a 3B?


Thus, walking a lot reduces everything else.

[Syd Thrift]

Originally Posted by x McLovin x

I’m aware of this although I have a specific question:
Suppose for example there is a LHB with extreme pulled-flyball tendencies who gets traded from a team which plays in a relatively neutral ballpark to the Yankees. Would it just be park factors that would account for the expected increase in home runs? Or does the fact that this specific player has that batted ball profile also magnify the increase?

Regardless of design, it would seem very illogical (and disappointing) for the game to not consider both.

It does not, no. As noted, the game is outcome, not physics based. We do not have data on HR rates by pull rate. There's *some* data from the 2000s or so onward on batted ball data where you could probably extrapolate this but the game simulates all of baseball history, not just the part that we have that kind of advanced data on.

To answer the other question... some parks would favor dead pull hitters over others and we just plain don't have the stats to say which without doing a lot of supposition. The short right field at old Yankee Stadium seems pretty obvious but... wouldn't the Polo Grounds favor the kind of dead pull hitter who hits HRs right down the line but severely punish a "standard" pull hitter who hits them into the alleys? How much? We literally don't have that kind of batted ball data for the Polo Grounds because the vast majority of the games played there, we simply do not have it and cannot collect it.

The other issue here of course is that switching to a physics based system would be a *massive* lift, perhaps a lift that this particular game will never take. It would require completely rewriting the engine for the modern era, taking all that time to work everything out so that everything still lines up, and so on. It'd probably be years before the game matched the same levels of statistical fidelity it does now. And again, that's just the modern era. With historical ball... yeah, look out.

I'm not saying there couldn't ever be a physics based OOTP like game. Someone could probably make one that's really, really good and that advances this kind of thing by quite a bit. OOTP isn't that game and probably never will be.

[OutS|der]

Originally Posted by EZB

1. Is it a walk?
2. Is it a HR?
3. Is it a K?
4. Is it a hit?
5. Is it a 2B/3B?
6. Is it a 3B?


Thus, walking a lot reduces everything else.

Logical order to me would be

4, then 1/3 and then 2/5/6.

Or even better 1. Is it an out. 2. determine how they got on base.

[kq76]

Originally Posted by ahunter5

For example, if you knew that a player had a rating that caused him to strike out 50% of the time that the game checked for strikeouts, you'd probably find it valuable to know when that check was made in the process. If it was first, that means he's striking out 50% of his plate appearances. If it checked walks first, and his eye rating gave the player a walk rate of 20%, then he's only going to strike out 40% of his plate appearances.

I'm not getting why you think there'd necessarily be a difference. I mean, yeah, I understand with your %s you're getting different results, but that doesn't need to be the case.

Let's say it does work the way you suggest, it checks for walks first. After the 20% for walks, you've got 80% of the PAs left. If you still had the player getting 50% Ks, just up the K% of the remainder to 62.5%. 80*62.5% still equals 50. Whether you do it my way or your way, it's still possible to get the same end stats. The %s just change along the way is all.

Maybe it does work your way (and to be clear, you're certainly not alone, many others have suggested it). It just seems odd to me that it would. In my mind it just makes the math unnecessarily complicated.

[Reed]

Originally Posted by ahunter5

For example, if you knew that a player had a rating that caused him to strike out 50% of the time that the game checked for strikeouts, you'd probably find it valuable to know when that check was made in the process. If it was first, that means he's striking out 50% of his plate appearances. If it checked walks first, and his eye rating gave the player a walk rate of 20%, then he's only going to strike out 40% of his plate appearances. .

This is not how statistics work. Kq76 is right.

Let’s follow your logic and triples is the very last thing it checks after walks, strikeouts, singles, etc.. if player didn’t get a walk, strike out, singles etc and the only thing left is triples, what then is his chances of getting a triple? By your logic he has about a 1% chance of getting a triple when it should be at this stage a 100% chance that he gets a triple. You need to factor in when something doesn’t happen (like a walk), then the chances for one of the other outcomes increases. This is assuming there is a strict order but I have no idea if that is the case.