Thoughts on OOTP6 and designing a baseball sim

[John Galt]

After reviewing some of the OOTP6 stuff on this board and the OOTP boards, I've been trying to figure out what ratings would form the "perfect" statistical engine for a baseball game. While I'm happy with some of the proposed improvements in OOTP6, I think it still fails in a few regards. Anyway, this is how I think baseball "really" operates:

Hitting

Contact
Power
Patience/Aggressiveness
Eye
Speed
Bunt
Pull (added after dawgfan's post - maybe unnecessary, but true accuracy would require it)

For any given AB:

A BB is more likely with high patience and eye
A HR is more likely with high contact and power
A 3B is more likely with high contact and speed (and maybe just a little power)
A 2B is more likely with high contact and power (and maybe just a little speed)
A 1B is more likely with high contact
A K is more likely with low eye and contact and high patience (aggressive hitters decrease their K's by not reaching deep counts)

A double play is more likely with low speed rating.

A bunt rating decreases the chance of the lead run getting out (or a double play) on a sacrifice and replaces contact and power on a normal hit attempt.

In OOTP6, I think K's and BB's won't be properly computed with the current system (although I don't think it is settled yet). Also, I think "gap power" is a myth. Evidence is this fact is looking a hitter with high SLG and high 2B is a great predictor for increased HR's. Power is power - hitters can't control if the ball really lands in a gap or not.

Also, there should be NO individual L/R split ratings for right-handed hitters. Bill James and others have shown over time, a right handed player's L/R split regresses to the same mean. Players with exaggerated splits only occur because of normal distribution curves and small sample sizes. Every right handed hitter should have the same L/R split.


Pitching

Velocity
Stuff
Command
Endurance
Type (Groundball, Flyball, Knuckleball with differences in extremes)
L/R

For a given AB:

A BB is more likely with low command
A K is more likely with high velocity, command, and stuff
A HR is more likely with low command and high flyball (this is the rating I'm least confident about, but would love to hear input)

Those are the 3 true outcomes of pitching and a DIPS theory. All other outcomes should be hits - based upon hitters ratings and defense (groundball pitchers are more reliant on infield defense). I include "Type" as a reference to "knuckleballers" and "groundball" pitchers because they act as modifiers in a DIPS theory. If a came really wants to include knuckleballers, I think they should have a separate algorithm for when they are on the mound).

Endurance should work on a curve that slowly deteriorates over a game (and eventually collapses) that modifies velocity, stuff, and command.

L/R should be a bias number on a scale leaning one way or another and should effect stuff (not command or velocity).

I think the Poise rating of OOTP6 is a very bad idea. It is easy to argue there are pitchers with more "poise" (Moyer seems to be the popular example), but poise is something you only see in hindsight. When Moyer was 28 and sucked, no one said he had "poise." Now that he is old and pitching well, he has "poise." I don't want to know that a pitcher has "poise" when he is in AAA at the age of 20 - you just can't know. Assuming this rating is going to be included, it should DEFINITELY be hidden (so you don't know a player will have it until they consistently outperform expectations) and should be minimal (Moyer is good because he doesn't have a lot of BB's and HR's - controlling 2 of 3 key outcomes).


Fielding

Range
Arm (I'm not sure if dividing IF, OF, and C makes sense yet)
Error Pct
Decisionmaking (added per dawgfan - small effect, but still important)

I think these are all good in OOTP6


Running

Speed
Instincts (Stealing and Running)

These are good as well, but the running instinct shouldn't have too much of an effect.




I think one of the reasons the discussions on the OOTP boards disturb me is that people want to talk about how the ratings effect statistics and not events. AVG is a stat that adds all sorts of different hits (1B, 2B, 3B, HR) against other events (K's and outs). Approaching the ratings from an aggregate level makes things unnecessarily confusing and often misleading. Instead ratings should effect singular AB's and when those AB's are aggregated they should produce realistic results.

Anyway, these are my ideas for the core of a great statistical baseball engine (and how I think baseball works). I've left out details such as the amounts each rating effects each outcome, but I think the key pieces are here. I think OOTP6 will be a good game, but I think it is a bit off in a few key places.

Edited:

For descriptions of the current ratings to be used in OOTP6, see this thread:

http://www.ootpdevelopments.com/boar...5&pagenumber=1

Edited for dawgfan's good ideas.

[Ben E Lou]

I'll pick this apart a little later, but allow me to say that this post could have the distinction of being John Galt's longest that I agreed with nearly everything contained therein.


(EDIT: Piss-poor grammar and syntax.)

[mauchow]

I will say just one thing on Gap Power...

Do you know right now who's going to lead the league in doubles this year?

You can make a pretty darn good guess I'll bet.

Take Mark Grace for example(during his good playing years)...

Every single year I could have told you Grace was going to be on top of the leaderboards in doubles. Why? Because he has Gap Power. Not homerun Power. Every year Grace was around 8-15 HR's. Yes! There is gap power. Yes, you can control whether or not you're going to hit it into the gap. As a college hitter, I know as soon as I see the ball come out of the pitchers hand if I'm going to be pulling it down the line for a double, or hitting it right back up the middle for a single. I could talk all day about this, and would rather not do that...

[GoSeahawks]

Originally Posted by John Galt

I think the Poise rating of OOTP6 is a very bad idea. It is easy to argue there are pitchers with more "poise" (Moyer seems to be the popular example), but poise is something you only see in hindsight. When Moyer was 28 and sucked, no one said he had "poise." Now that he is old and pitching well, he has "poise." I don't want to know that a pitcher has "poise" when he is in AAA at the age of 20 - you just can't know. Assuming this rating is going to be included, it should DEFINITELY be hidden (so you don't know a player will have it until they consistently outperform expectations) and should be minimal (Moyer is good because he doesn't have a lot of BB's and HR's - controlling 2 of 3 key outcomes).

I think poise is a great addition. I will use a personal example of why I like the idea.

I was a pitcher in college and I believe what seperated me from other pitchers was that I could control my emotions while I was on the mound. I wouldn't get rattled after walking the leadoff batter or giving up a big hit. I could focus up and battle through the inning. Another thing that I was able to do was consistently throw inside even after I hit a batter. In fact one game I hit 7 batters while pitching a one hit shutout.

On the flip side there was this kid on my team with a good arm (87-89mph). He could go five innings shutting down the opposing team. Suddenly he would walk a guy. The next guy would come up and get a hit off of him. After allowing a hit it was game over. The guy would lose it completley. His mind would be out of the game and from there it was ball after ball until the coach would pull him. He is my personal example of someone who had no poise. He couldn't handle failure and I don't think it was in his mental makeup to be a pitcher.

You need players who can bounce back after failure and that's why I think poise could be a great addition. I guess it really depends on how ootp impliments it.

[John Galt]

mauboy, you raise a couple good points and one that I knew was a partial shortcoming to my current description.

First off, the fact that you knew the moment you hit the ball where it was going is not the same thing as knowing before your bat hit the ball. You may be able to hit it with some measure of accurary through timing, but that generally controls pull v. opposite field, not gap v. non-gap.

Second, when I said gap power is a myth I didn't quite describe what I meant right. Mark grace didn't have "gap power" in the sense that he didn't hit "gaps" more often than an OF anymore than most hitters. Rather, he had power, but not home run power. The way I envision power working is in tiers. Players with very high power will hit HR's and 2B's, players with high power will hit slightly less HR's and slightly more), good power will get you more 2B's but not as many HR's (Grace). This was not clear from my first description, but I think that accurately reflects "power" in baseball. The ability to hit a "gap" v. "not a gap" is non-existent. Likewise, having HR power, but not 2B power is non-existent. That is why separating power into 2 categories is a mistake, IMO.

Edited:

I just thought of another way to explain this. Power is roughly a predictor of the total number of 2B's and HR's a hitter will hit (3B's are more speed related). So, contact + power determine the net total of 2B + HR. The percentage of those hits that end up being HR's is determined by power (ie - the doubles that clear the fence). Maybe that makes things clearer?

[John Galt]

Originally Posted by GoSeahawks

I think poise is a great addition. I will use a personal example of why I like the idea.

I was a pitcher in college and I believe what seperated me from other pitchers was that I could control my emotions while I was on the mound. I wouldn't get rattled after walking the leadoff batter or giving up a big hit. I could focus up and battle through the inning. Another thing that I was able to do was consistently throw inside even after I hit a batter. In fact one game I hit 7 batters while pitching a one hit shutout.

On the flip side there was this kid on my team with a good arm (87-89mph). He could go five innings shutting down the opposing team. Suddenly he would walk a guy. The next guy would come up and get a hit off of him. After allowing a hit it was game over. The guy would lose it completley. His mind would be out of the game and from there it was ball after ball until the coach would pull him. He is my personal example of someone who had no poise. He couldn't handle failure and I don't think it was in his mental makeup to be a pitcher.

You need players who can bounce back after failure and that's why I think poise could be a great addition. I guess it really depends on how ootp impliments it.

Even if all this is true (which there isn't statistical evidence for it that I've seen), that is a reason for hidden rating. People who watch a pitcher with "poise" are seeing it because he is doing well, not because he is more in control of his emotions. A pitcher that was immune to pressure but sucked would be said to be unemotional or disinterested. Poise is a merely a hindsight reflection of success beyond expected ability.

Until someone does a study of pitchers identified with high "poise" in the minors and finds they were more successful, I think the rating is a very bad idea. It gives too much information to the player and it may not actually exist.

[dawgfan]

An interesting discussion. I'm not sure I'll ever 'trust' OOTP to get baseball right, at least to my view of how baseball works. I'm waiting patiently for CraigSca's baseball sim, since he seems to follow most closely my vision of how baseball works, the both of us heavily influence by sabermetrics.

I agree with most of your ideas, with the following comments/disagreements:

In general I agree with your hitting breakdown. I think an additional category would have to do with hitting style, i.e. to what extent is the batter a pull-hitter; a higher rating here I think would correlate with a higher number of GIDP, and might increase HR depending on ballpark configuration. You could also argue there's some variation in how level a batter's swing is - ranging from extreme uppercut (resulting in more flyballs and K's) to very level (resulting in more groundballs, fewer K's). This is a modifier to the contact rating.

Originally Posted by John Galt

Also, there should be NO individual L/R split ratings. Bill James and others have shown over time, player's L/R split regresses to the same mean. Players with exaggerated splits only occur because of normal distribution curves and small sample sizes. Every hitter should have the same L/R split.

Actually, this holds true only for RHB. Lefties are subject to greater/more variable L/R splits, the most accepted theory being that LHB face far fewer LHP growing up and as such are more prone to being vulnerable to LHP, whereas RHB face mostly RHP growing up and their eyes adjust to pitches coming from that side of the pitching rubber.

Originally Posted by John Galt

Fielding

Range
Arm (I'm not sure if dividing IF, OF, and C makes sense yet)
Error Pct

I would also add decision-making - does the fielder throw to the right base? How aggressive is he in going for the big play rather than the safe out?

With arm, I think it should be sub-divided between strength and accuracy.

Originally Posted by John Galt

Running

Speed
Instincts (Stealing and Running)

These are good as well, but the running instinct shouldn't have too much of an effect.

I disagree on the instinct comment - there needs to be a distinction between runners who are simply fast, and those that are aggressive and smart. You don't have to be blazingly fast to be an effective baserunner and base-stealer - if you know how to read the pitcher you can get a great jump that overcomes a lack of speed.

[FBPro]

I think someone needs to make a baseball sim where there are a bunch of ratings, but no names or titles on them and you can name them yourself.

Suckiness, firepower, sleepy, wicked-fastballness, donkey, etc, etc.......

[AgustusM]

l/r split example

JT Snow career LvR OPS

vL .647
vR .822

there are countless others, this is just the first one that comes to mind

[John Galt]

Originally Posted by dawgfan

An interesting discussion. I'm not sure I'll ever 'trust' OOTP to get baseball right, at least to my view of how baseball works. I'm waiting patiently for CraigSca's baseball sim, since he seems to follow most closely my vision of how baseball works, the both of us heavily influence by sabermetrics.

I agree with most of your ideas, with the following comments/disagreements:

In general I agree with your hitting breakdown. I think an additional category would have to do with hitting style, i.e. to what extent is the batter a pull-hitter; a higher rating here I think would correlate with a higher number of GIDP, and might increase HR depending on ballpark configuration. You could also argue there's some variation in how level a batter's swing is - ranging from extreme uppercut (resulting in more flyballs and K's) to very level (resulting in more groundballs, fewer K's). This is a modifier to the contact rating.



Actually, this holds true only for RHB. Lefties are subject to greater/more variable L/R splits, the most accepted theory being that LHB face far fewer LHP growing up and as such are more prone to being vulnerable to LHP, whereas RHB face mostly RHP growing up and their eyes adjust to pitches coming from that side of the pitching rubber.



I would also add decision-making - does the fielder throw to the right base? How aggressive is he in going for the big play rather than the safe out?

With arm, I think it should be sub-divided between strength and accuracy.



I disagree on the instint comment - there needs to be a distinction between runners who are simply fast, and those that are aggressive and smart. You don't have to be blazingly fast to be an effective baserunner and base-stealer - if you know how to read the pitcher you can get a great jump that overcomes a lack of speed.

All very interesting.

First, I think groundball/flyball for hitters could be included and I omitted at the last minute because I wasn't sure evidence of it really existed independent of power. In other words, is there evidence that shows anything but a direct correlation between powers and flyballs? Is so, it should be added.

Second, Pull ratings are nice (especially for evaluating which defensive player matters), but I omitted them because I was trying to get the core stats. However, I'm more than happy to add them to reflect some important, but small effects.

Third, I forgot about lefties having a L/R split - I will change it - I just get so tired of games with exagerrated splits and other players that magically hit both sides well.

Decisionmaking is probably a good idea, but its effect is probably so small. Even the Manny's of the world only truly lose a few outs a year on stupidity.

I thought about dividing arm accuracy and strength, but then decided it really doesn't matter for statistical accuracy. I presume the arm rating is the aggregation of both - in any given play, it doesn't matter too much whether the throw was too weak or off target. I know I can think of examples where an off target throw is worse, but I just don't know if it is worth adding a rating that matters so little.

Last, I'm not denying "instincts" - I'm just saying there is a tendency to overvalue good running skills when they really only come into play every now and then (how many close plays are there a game where a runner really made is own decision?).

Going back to edit the original post . . .

[GoSeahawks]

Originally Posted by John Galt

Even if all this is true (which there isn't statistical evidence for it that I've seen), that is a reason for hidden rating. People who watch a pitcher with "poise" are seeing it because he is doing well, not because he is more in control of his emotions. A pitcher that was immune to pressure but sucked would be said to be unemotional or disinterested. Poise is a merely a hindsight reflection of success beyond expected ability.

Until someone does a study of pitchers identified with high "poise" in the minors and finds they were more successful, I think the rating is a very bad idea. It gives too much information to the player and it may not actually exist.

I don't believe that it is a hidden rating. If you're around baseball enough you know who has it and who doesn't. There are players who have great poise even though they are pitching terribly.

I have this great tape of a College World Series game from a couple years back where this pitcher is shutting down one of the top teams in the country. The first big hit he allowed was a monster homerun. From there his style of pitching completely changed. Instead of pitching aggressively and running the ball inside he started playing it safe with fastballs and off speed on the outside part of the plate. Consequently the team started pounding him until he gave up the lead.

It's not just in college pitchers you see this with. It happens in the pros as well. Look at Rick Ankiel for example. Anyway, I gotta head off to work.

[John Galt]

Originally Posted by AgustusM

l/r split example

JT Snow career LvR OPS

vL .647
vR .822

there are countless others, this is just the first one that comes to mind

Edited to reflect that Left handed hitters do have split. As for right handed hitters, yes players have different statistical splits, but they are indicative of a normal distribution curve. They don't actually have different splits, they all regress to the mean over time.

[John Galt]

Originally Posted by GoSeahawks

I don't believe that it is a hidden rating. If you're around baseball enough you know who has it and who doesn't. There are players who have great poise even though they are pitching terribly.

I have this great tape of a College World Series game from a couple years back where this pitcher is shutting down one of the top teams in the country. The first big hit he allowed was a monster homerun. From there his style of pitching completely changed. Instead of pitching aggressively and running the ball inside he started playing it safe with fastballs and off speed on the outside part of the plate. Consequently the team started pounding him until he gave up the lead.

It's not just in college pitchers you see this with. It happens in the pros as well. Look at Rick Ankiel for example. Anyway, I gotta head off to work.

I'm sorry if I'm harsh, but this is one of my pet peaves. Being around baseball, creates all sorts of things that you "know" - and a great many of them are wrong or not helpful. Clutch hitters are the classic example.

A pitcher will poor poise is really just a pitcher playing poorly who happens to show his emotions. A pitcher with good poise is really just a pitcher playing well who looks "fierce" or "focused" on the mound. Similarly, does Randy Johnson lack "poise" because he is emotional on the mound even though he is rock solid?

Poise is what you see in hindsight - there is no evidence it really exists. If you knew a pitcher had "poise" early on in his career, you wouldn't see so many busted pitchers and bloated contracts for underachievers.

As for Rick Ankiel, I just don't think he fits in this discussion. If he does, he proves my point. Everyone thought he was a top pitching prospect with tremendous "poise" and then he got the yips. I think that is an extreme case and irrelevant, but if anything it shows why "poise" in foresight is crazy.

[dawgfan]

I dunno JG - as much as I believe in sabermetrics and am skeptical of things like clutch performance, I think there's something to this idea of 'poise' on the mound.

The diffiuclt thing about proving this is that it's not a simple concept that can easily be teased out of normal statistics. You'd have to look hard at splits like how a pitcher did with runners on base, runners in scoring position, and ideally you'd also tease out the game situation - what was the score, what inning, etc.

I agree that what we might think of as 'poise' might be demonstrated in different ways by different people. Simply showing emotion on the mound is misleading - obviously a guy like Randy Johnson thrives by pitching on emotion, riding that adrenaline. That said, early in his career he appeared to not do so well in this category - when he started walking guys, or his fielders committed errors behind him, he seemed more likely to unravel.

The Jamie Moyer example is a good one I think - he may not have had great poise as a 28 year old, but I think this is a skill that can be learned later in your career as you observe others that possess it. Moyer always cites Jimmy Key as a guy he really tried to emulate, both in terms of pitching style and how he battled.

Moyer doesn't have the greatest stuff in the world, but he's aggressive with it, doesn't get easily rattled, and doesn't shy away from pitching the way he has to in order to be successful. Contrast that with Freddy Garcia - here's a guy with phenomenal stuff and decent command. But when his command is a little shaky and he feels the ump is shrinking his strike zone and things aren't going well, he's more likely to give up a big inning.

What we're really talking about here is to what extent do conscious, emotional elements factor into performance above and beyond simple random variation. I think if someone wanted to study these kinds of things, you could pore through the stats and find some meaningful info - in addition to the numbers on how a pitcher does with runners on, runners on in scoring position (plus game situation factors) you could also look at how often a pitcher gives up a "big" inning (3 or more runs? 4 or more runs?) and then compare all these numbers to the average pitcher. Looking at them over time you might get some meaningful info. Of course, you have to account for the possibility that guys can change over time in this area.

In terms of game mechanics, I think where this 'poise' idea is most likely to rear itself is in terms of command. I don't have a huge problem with the idea that there might be a hidden factor that influences command above and beyond simple chance. When you play a game like Strat-O-Matic and all the results for Jamie Moyer are based on chance with probabilities derived from his previous season's performance, what that game mechanic is saying is that variation in performance is due only to chance. I'm not sure I buy that - I think there is an emotional component that affects performance.

[John Galt]

Fair points, dawgfan, but let me offer this in reply:

"Poise" should be presumed not to exist unless statistical evidence exists otherwise. Too often, things people think they see in baseball become conventional wisdom. Finally, someone comes along and disproves the nonsense. Yet, people still cling to the ridiculous beliefs. Whether "poise" is one of the ridiculous assumptions is still an open question, but I'm inclined to put the burden on those offering the theory to provide more than anecdotal evidence for it.

Now, if poise is going to be part of a game, I'm still determined to believe it should only be a hidden stat. Making it visible just doesn't match the way we observe it in real life. Tons of pitchers are said to "know how to pitch" or have "poise" in the minors, but most of them wash out. Most recently, I think of pitchers like Mario Ramos and Ed Yarnell who haven't done anything yet. All the examples of pitchers with high "poise" are really just veterans who have become consistent. Now if you want to make different growth curves for player development, I'm 100% behind that. Then you can explain why pitchers like Schilling and Moyer acquired "poise" later in their careers. I still think "poise" is a poor way to describe things, but a hidden rating that is used to model development is ok I guess. Allowing the rating to be seen with foresight just seems unrealistic.

Also, "poise" in a game if it has any effect is always bad for balance. I know I'll always draft a player with high "poise" and if it really has an effect I'll get artificial benefits that just aren't mirrored in the real world.

Now, you can say all you want about players like Garcia (it could just be that he is hungover all the time as some have suggested), but my point is that "poise" is really just a substitute for a variety of different things: control (not all of the high "poise" pitchers have good control), consistency (or at least perceived consistency since Maddux and Moyer have actually had trouble with that lately), emotional stability, and long careers. Does that really mean "poise" exists and should be a rating? For now, I'm not convinced.

BTW, 2 questions related to Garcia - is there actual evidence that he is more vulnerable to a big inning after an "incident" (controlling for normal endurance issues)? And how does he really differ from someone like Pedro in terms of poise given Pedro's well documented late inning struggles last year (which can be explained by endurance issues)?

[The_herd]

Originally Posted by John Galt

BTW, 2 questions related to Garcia - is there actual evidence that he is more vulnerable to a big inning after an "incident" (controlling for normal endurance issues)? And how does he really differ from someone like Pedro in terms of poise given Pedro's well documented late inning struggles last year (which can be explained by endurance issues)?

My personal belief on this is it's mostly related to endurance, but focus is tied in as well. As players tire, some are able to keep their mind focused on the task at hand and aren't as vulnerable to the big inning as someone that lacks focus. Pedro has always been a player whose mind will tend to wander after giving up a big hit or not getting a call. I really think fatigue magnifies this. When the player is losing his "stuff" as he tires and the mind begins to wander, he's more likely to give up a few hits.

[dawgfan]

Originally Posted by John Galt

BTW, 2 questions related to Garcia - is there actual evidence that he is more vulnerable to a big inning after an "incident" (controlling for normal endurance issues)? And how does he really differ from someone like Pedro in terms of poise given Pedro's well documented late inning struggles last year (which can be explained by endurance issues)?

Purely anecdotal evidence. It would be worthwhile to check the available splits for him to see if there's any evidence backing up this assertion. As to the 2nd question, I'm thinking of big-inning blow-ups Freddy would have early in the game that couldn't easily be explained by fatigue.

One thing to remember about your point on statistical evidence - in general I agree with your approach, but given that in order to gain confidence in statistical results you often need a large sample size, often much bigger than one season will yield. Where this can be sticky is in looking at something like this idea of poise - maybe something happened to a guy in an offseason that knocked his emotional state off-kilter (divorce? child? drinking/substance abuse issues?) or in the other direction, he resolved a big emotional issue. Assuming that such factors can affect performance, you'd then have to know personal histories of players and account for those situations when evaluating their career to try and find evidence to back this idea of poise affecting performance.

I guess what I'm saying is, just because we may have a hard time proving that such a factor exists, I also think you have to go to a lot of work to prove it doesn't exist.

When you hear former players discuss concepts like this where they are convinced it's a real factor, and they paint a plausible picture, I think there's some burden of proof on statisticians to prove them wrong rather than the other way around.

[John Galt]

Fair enough. I just keep thinking back to all the things players still "know" even when statistics say otherwise. Clutch hitting is the most famous example, but other things like players doing better in contract years, pitchers having a ton of control beyond the 3 true outcomes, and catchers substantially improving pitching staffs are all "truths" that have been tested and discredited. Yet, players cling to them. Being close to the game gives you a lot of insight, but it also blinds you because you are just too close. Until people start seeing "poise" without success and predicting "poise" successfully, I'm going to continue to be skeptical. Observation in hindsight is just too damn easy.

[The_herd]

To touch on poise a bit, I think a poise type rating should definately be in the game, although hidden. As Dawgfan stated earlier, Jamie Moyer and Jimmy Key are good examples of this. Having it represented as a hidden rating really is the way to go, though. That prevents people from simply filling the last few spots on the team with pitchers with a high poise rating that they know gives them a certain advantage.

A possible arguement against this, at least in Moyer's case, is if you look at his G/F ratio. He's a pitcher whose G/F ratio was well above 1.0 until moving to Seattle. Since arriving there, his numbers have been right at, or below 1.0. You can argue that it's learning to pitch, taking advantage of your surroundings, poise, whatever. What it comes down to is, there was a change in his pitching style and the fact that it happened when he moved into a "pitcher's ballpark" coincided with it.

I do think there is a certain quality about some pitchers that we can only describe as poise, however, I don't think it has as large of an impact on the game as many believe.

[Crapshoot]

Originally Posted by John Galt

Also, there should be NO individual L/R split ratings for right-handed hitters. Bill James and others have shown over time, a right handed player's L/R split regresses to the same mean. Players with exaggerated splits only occur because of normal distribution curves and small sample sizes. Every right handed hitter should have the same L/R split.

It wasn't James- it was a fairly recent discovery actually, that the right handed hitter tends to hit about 9% better against lefties as a general rule.
James actually pointed out the idea that splits are real way back when.
I agree with the idea in theory as a statheadz- but in practice, there are a lot of the general (even OOTP) buying public who would panic at this idea. Also, you still have to deal with all the lefties, where the distribution is a lot more varied.

I think the Poise rating of OOTP6 is a very bad idea. It is easy to argue there are pitchers with more "poise" (Moyer seems to be the popular example), but poise is something you only see in hindsight. When Moyer was 28 and sucked, no one said he had "poise." Now that he is old and pitching well, he has "poise." I don't want to know that a pitcher has "poise" when he is in AAA at the age of 20 - you just can't know. Assuming this rating is going to be included, it should DEFINITELY be hidden (so you don't know a player will have it until they consistently outperform expectations) and should be minimal (Moyer is good because he doesn't have a lot of BB's and HR's - controlling 2 of 3 key outcomes).

Agree absolutely- Poise smacks of "clutch " hitting and "veteran" experience and all that other bs that has little place in a game. I think there is merit to the idea that a pitcher with good stuff doesn't always make it, but it may better be reflected in an intelligence rating of sorts- a pitcher that works harder to spot his opponent's weakness and recognizes his own will probably do better. The problem is that modelling this is absurd at best- as such, I would rather it not be included at all.

[Crapshoot]

btw, Jaime Moyer keeps getting cited as an example of "poise" - Moyer's improvement in performance can almost wholey be attributed to a lowering of his BB/9 rate. In 1997,98,99,01, and 02, he was in the top 6 in the AL in BB/9 - the highest being 2.05 in 97. More so, Moyer has been consistently helped by a damn good defense (Ken Griffey Jr followed by Mike Cameron in CF- two of the greatest defensive CF In their prime). I decided to a do a quick and dirty study on it.

						
	IP	BB/9	K/9	K/BB	HR/9	ERA+
1986	87.3	4.33	4.64	1.07	1.03	80
1987	202	4.32	5.66	1.31	1.25	84
1988	201	2.46	5.42	2.20	0.90	104
1989	76	3.91	5.21	1.33	1.18	82
1990	102.3	3.43	5.10	1.49	0.53	84
1991	31.3	4.60	5.75	1.25	1.44	65
1993	152	2.25	5.33	2.37	0.65	129
1994	149	2.30	5.26	2.29	1.39	105
1995	115.7	2.33	5.06	2.17	1.40	94
1996	160.7	2.58	4.42	1.72	1.29	127
1997	188.7	2.05	5.39	2.63	1.00	117
1998	234.3	1.61	6.07	3.76	0.88	132
1999	228	1.89	5.41	2.85	0.91	130
2000	154	3.10	5.73	1.85	1.29	83
2001	209	1.89	5.12	2.70	1.03	127
2002	230.7	1.95	5.73	2.94	1.09	123
2003	215	2.76	5.40	1.95	0.80	136

Correlation to ERA+
BB/9 -0.841953974
K/9 0.018316508
K/BB 0.758572988
HR/9 -0.415848121
Moyer's success is pretty clearly most correlated with his BB/9 rate, and to a lesser degree with his K/BB ratio - not his "poise". He learnt to better control his pitches, and how to better use them- if you have a 65mph changeup, a 83MPH fastball is plenty fast. Years such as the last one are more likely to be exceptions, and one should expect a Moyer decline, aided by the loss of Cameron - if his K/BB ratio is under 2 again, odds are that his ERA will be over 4.

[Crapshoot]

Edit, that looks awful- just look at the results below. Any help from anyone who knows how to better post tables in here ?

[The_herd]

Originally Posted by Aadik

Edit, that looks awful- just look at the results below. Any help from anyone who knows how to better post tables in here ?

Put it in notepad, then paste into here and use the [code] tags.

Edit-that should help.

[Crapshoot]

Thanks...

[The_herd]

Throw in the Groundball/Flyball ratio onto that chart, and consider that he moved to Seattle in '96, and you have a pretty good case that Moyer's success has more to do with learning how to locate his pitches, and less to do with Poise.

[CraigSca]

I hate getting into conversations late...but this was movie night with the wife, so I'll offer a few random musings...

Gap Power vs. Homerun Power. I think there are two things at work here - 1) strength of the hitter. Not necessarily Mark McGwire or Jason Giambi strength, but "athlete" strength. In the 80's I can remember a lot of shortstops just being truly awful when it came to power output - your Chris Speiers, Pepe Frias, Kiko Garcia, etc. I don't have the numbers off the top of my head, but my assumption is they were all small guys and couldn't match the athletic output of a guy like Soriano (who's obviously not a huge weightlifter but is obviously an athlete). 2) the degree of uppercut of the swing. I think this is the piece that "magically" decides whether a guy is a high doubles hitter as opposed to a homerun hitter. I can't take credit for this opinion - I culled this mainly from "The Physics of Baseball" - but it makes a lot of sense. High uppercut hitters hit more homeruns, but "generally" have lower batting averages than a guy with a more level swing. Think about it - if the guy with the uppercut swings early, his swing is up higher, and he pulls a groundball. If he's slow on the swing, he hits a flyball to the opposite field. If he hits it dead-on, strength and the trajectory coming off the bat increase his chance for a homerun. Now, let's think about the guy with the level swing - if he swings early, he pulls a line drive. If he swings late he hits a line drive to the opposite field. If he hits the ball dead on - he should hit a linedrive to centerfield. All the while, the fact that he has a more level swing will decrease the trajectory, giving him "gap" power. This is admittedly simplistic, but it gets the idea across.

Poise - I'm in the same boat as a lot of people and feel this may fall into the same bag as clutch hitting. I'm relatively close to New York, so I heard all last year about Jeff Weaver and his inability to cope with mistakes made in the field behind him. As with a lot of things in the baseball field, it does make a lot of sense to think a pitcher can be effected by such things and have his performance take a hit accordingly. However, I'm also of the mindset that while that sounds believeable, I'm really not sure that it should be introduced into a game engine until such a thing can be proven and quantified.

Finally, while there's been no official announcement regarding this (it's not as if this will register on the AP newswire :lol), Markus has asked me to come aboard and work with him on OOTP. My role with Out of the Park is to aid Markus in making OOTP as statistically accurate as possible. I can't take credit for the switch to DIPs, but I can say I helped in fine-tuning the DIPS implementation as well as provided input regarding the role of defense and some other features currently being added to the game. Markus is an incredible programmer and he's been nothing but a pleasure to work with thus far. I think we've both been extremely open with how we each went about things with our respective programs. Hopefully, I can allow him to focus on the things he enjoys and help make OOTP the most statistically accurate baseball game on the market.

-Craig

[ISiddiqui]

Woo!... good news Craig... so if it isn't 100% statistically accurate we blame you?

[Karim]

The fact that Craig is helping with OOTP6 gives me high hopes for this release. Not that Markus couldn't have done a good job on his own but Craig has always had great ideas for making a statistically accurate baseball sim.

[ISiddiqui]

Well the best news is that Markus is looking for the best people he can to help with the statistical accuracy of the game .

[CraigSca]

Originally Posted by ISiddiqui

Woo!... good news Craig... so if it isn't 100% statistically accurate we blame you?

Haha - well, we've only been working together a short while, and the great majority of that time Markus has been working on the next release. He's asked me to verify a couple things, and then also put me to task on some new things for the game, which I've done. I've also made my recommendations regarding a few things, but ultimately this is his baby, and he has every right to do anything he wants with it. Again, thus far he has been extremely amenable to my suggestions and I have to give him a lot of credit for that.

One thing that has been interesting is the sharing of information from a programmer's point of view. It seems there are a lot of things that we were doing exactly the same (one in particular) though we had no contact as each of us were working on our own projects. However, there also some things we were doing radically different - I think this is where both of us can mutually benefit.

Just like the rest of us, Markus wants the game to constantly improve and he really does want this game to be as accurate as possible. I think the next version will be a noticeable improvement in this regard (heck, I can't wait to try the new DIPS system out myself!) and this improvement will continue in the future.

-Craig

[lynchjm24]

Originally Posted by mauboy1

Yes, you can control whether or not you're going to hit it into the gap. As a college hitter, I know as soon as I see the ball come out of the pitchers hand if I'm going to be pulling it down the line for a double, or hitting it right back up the middle for a single. I could talk all day about this, and would rather not do that...

What an utter pile of crap.

[lynchjm24]

So I've learned in this thread that:
Jaime Moyer has 'poise'
Career ERA+ 109
Pedro Martinez doesn't have poise
Career ERA+ 174

[Ksyrup]

I tend to agree that poise is the equivalent to clutch hitting. It seems to me that poise is being equated, at least in the real-life examples being thrown around, with above-average pitchers who don't have overpowering stuff. Moyer, Key, Maddux - these are all guys who've learned an alternate, yet successful, way to pitch rather than just overpowering hitters.

Frank Tanana is the ultimate example of why poise doesn't exist. He was an overpowering pitcher who learned to pitch "with poise" for the last half of his career. I just don't buy that. If a guy can't (or no longer can) pitch like Randy Johson, but is otherwise successful, it isn't because of poise. It's because there's more than one way to be a successful pitcher, and that particular pitcher has figured out what works for him.

[John Galt]

Originally Posted by CraigSca

Gap Power vs. Homerun Power. I think there are two things at work here - 1) strength of the hitter. Not necessarily Mark McGwire or Jason Giambi strength, but "athlete" strength. In the 80's I can remember a lot of shortstops just being truly awful when it came to power output - your Chris Speiers, Pepe Frias, Kiko Garcia, etc. I don't have the numbers off the top of my head, but my assumption is they were all small guys and couldn't match the athletic output of a guy like Soriano (who's obviously not a huge weightlifter but is obviously an athlete). 2) the degree of uppercut of the swing. I think this is the piece that "magically" decides whether a guy is a high doubles hitter as opposed to a homerun hitter. I can't take credit for this opinion - I culled this mainly from "The Physics of Baseball" - but it makes a lot of sense. High uppercut hitters hit more homeruns, but "generally" have lower batting averages than a guy with a more level swing. Think about it - if the guy with the uppercut swings early, his swing is up higher, and he pulls a groundball. If he's slow on the swing, he hits a flyball to the opposite field. If he hits it dead-on, strength and the trajectory coming off the bat increase his chance for a homerun. Now, let's think about the guy with the level swing - if he swings early, he pulls a line drive. If he swings late he hits a line drive to the opposite field. If he hits the ball dead on - he should hit a linedrive to centerfield. All the while, the fact that he has a more level swing will decrease the trajectory, giving him "gap" power. This is admittedly simplistic, but it gets the idea across.

First off, it is great to hear you have joined the OOTP team.

Now, to the issue of gap power. The physics of baseball is a great resource and explains how and why the balls fly off the bat. What it doesn't do is explain why certain events happen more than others and how certain events interrelate.

On the gap power debate, I think the physics can be misleading. If a hitter is a hypothetical pure flyball hitter, he is going to be a crappy player unless he has Bonds power because flyballs stay up too long. Likewise, a hypothetical pure groundball player will be a crappy player in almost all cases.

What matters then are the cases in between. While there is a range of groundball to flyball ratios among major leaguers, there is also a degree of self selection in whether you become a groundball or flyball hitter. Barry Bonds was never going to be a groundball hitter because it would waste his power. I can't find the stats (CAN SOMEBODY FIND HITTER G/F RATIOS?), but I'm willing to bet (based on my memory) that all the top flyball hitters are all HR hitters (which is consistent with what your saying as well).

I'm not denying at all that G/F ratios effect the doubles v. HR distribution. My point (after clarification) is that a tiered power rating should be able to capture the importance of G/F ratios. That is, hitters select whether to become groundball or flyball hitters based on their power. I see no problem with a G/F rating, but integrating it into the overall statistical engine is a tricky proposition and I think a tiered power rating would serve just as well.

Now, just to clarify, when I'm saying "gap power" is a myth, I'm saying the ability to hit into a "gap" v. in the range of an OF is mythical, not that some players don't find the gaps more (because their balls don't go over the fence - not because they can avoid OF's better). My problem with a "gap power" rating is the same problem I had with the "triples" rating. Yes, some players get more triples, but it isn't because they know how to hit the right spots in the OF, it was because of other factors. "Gap power" seems to be combining the 2B and 3B ratings of old. I think their should just be one power rating and you can make it tiered (as I've proposed) or you can add a G/F ratio. However, the current system seems to describe an attribute that just doesn't exist.

[CraigSca]

John,

I agree wholeheartedly. I wasn't speaking specifically to the GB/FB all issue, I was mainly speaking to the issue of uppercut vs. level swing. These may be inter-related at some point, but for this discussion I'm going to keep them separate. What I was mainly trying to show is that a guy like Wade Boggs didn't necessarily have "gap" power, he just hit the tar out of the ball and because his swing was so level, the ball never had the proper trajectory to go over the fence (an therefore got the label without necessarily having the attribute).

And...regarding correlation between FB/GB ratio and "homerun power" - Yes, I would venture to say there's a correlation, but it's not as clearcut as you would think (this is from memory, I'd have to look at the ratios, too). From what I remember, there were some extreme FB hitters who didn't hit homeruns and extreme GB hitters who did hit their share. Pair that with the discussion above and it becomes clear that a player can have an EXTREME uppercut, and perhaps have a tendency to be easily fooled by offspeed pitches. Therefore, when he does get his timing right, he knocks the ball out of the park. When he's early (since the word is out he's easy to fool via the changeup) ,he swings high on the ball, and hits harmless ground balls to the left side (if he's a RHB). So...in essence you can get a "GB" hitter who hits homeruns.

Really, I think we all agree that gap power may not necessarily be the quantifiable attribute that's going to be the be-all, end-all. I'm not sure how much leeway Markus has regarding this because he's coming up on the end of the development cycle on V6, but I'll certainly discuss it with him.

-Craig

[John Galt]

Thanks for the reply.

I would be interested to see the numbers correlating GB/FB with HR because I can't seem to find them anywhere. That would help ease my concerns on some of the issues here.

As for gap power, I think it is important at the very least to consider a different name for the rating. I think GB/FB would be a better title. Ideally the system would work as you described. I'm still not sure if my simplified "power" rating wouldn't be better (a lot of that depends on the correlation between GB/FB and HR's). And another concern I have if the rating stays is that you get High Power/Low Gap Power players. I can understand the opposite case, but High Power should translate to a decent measure of whatever you want to call Gap Power as well.

Either way, I'm hoping OOTP6 will be a great game. I'm just more demanding with baseball than any other sports sim and since the initial Baseball Mogul I haven't been able to enjoy a baseball sim for very long (and Baseball Mogul only because it was so groundbreaking, not because it was accurate).

[oykib]

Neyer had an article on poise or clutch ability a few years back that wasn't particularly 'scientific', butit was the best explanation that I've seen.

All maor leaguers-- or just about-- have poise. That's how they were able to deal with all the setbacks that they've had in their careers. Almost no one breezes through all the way to the major league level.

The example above about the college teammate doesn't work when we are talking about Major League pitchers who have gone though high school, college, and various levels of minor league competition against the top players available. They've all been rocked plenty of times and have learned to deal with it.

Now, we've all seen players at the Major League level lose it even though they had great stuff. Those guys quickly decline and fall from their rosters. Poise isn't some rating that should be 1-10 with six or more being some kind of bonus. Poise is a prerequisite.

Maybe that's why the prospect in your sytem tears up the lower minors nd implodes in AAA. Or maybe that's a guy like Hensley Meulens or someone who perfomed well below the MLEs that he put up in the minors. That guy lacks poise. But you can't lack poise and be even a contributing mamber of a big league team.

[Ksyrup]

FYI, from Marcus today:


While spending the weekend in bed with a damned flu, I had some time to think about poise and what it was supposed to do, and what it is/may be in real life.

Actually it was intended as a rating that slightly influences all numbers that a pitcher has control of, so K's, BB's and HR's... the intend was to being able to create a wider variety of pitcher types. However, this would have been a completely random rating when importing historical seasons. And how should it work in the player development? Is this really a talent you can develop? Also, I didn't like the term 'poise' from the start, since it is easy to misunderstand and you can interpret all sorts of stuff into it.

So, with all the discussion going on, and with some further thinking, I came to the conclusion that this is nonsense. So, the rating was scratched.

Here is the final list of pitcher ratings:
Stuff (heavily influences Ks)
Movement (Influences HRs)
Control (Influences BBs, and slightly Ks and HRs)
Velocity (Slightly influences Ks)

Endurance, Hold Runners, HBP, Balks and WPs.

Cheers,
Markus

[John Galt]

Yeah!!!!

I hope he rethinks a few of the other ideas, but I think this is great news.

I still don't know what "stuff" and "movement" are really about - aren't they really the same, and if not, how is "stuff" different than "velocity" (especially for fastball pitchers).

[oykib]

I'm hoping that this version of OOTP is the one that does it for me. I keep buying it. But it keeps just missing with me. I don't know what it is. As everyone on this board knows, I love baseball. It's my favorite sport by a fair margin. On consoles I play it more than all the other sports combined. But I haven't had a PC baseball game that I've really played to death since Earl Weaver.

Actually, I've played 100s of seasons of BM. But that's more of a diversion than a game.

[oykib]

Dola--

Maybe the thing that's missing is success cycles. That's seems to be the final piece of a dynamic trading and free agency model that I feel is necessary to really enjoy a baseball sim.

[John Galt]

Originally Posted by oykib

I'm hoping that this version of OOTP is the one that does it for me. I keep buying it. But it keeps just missing with me. I don't know what it is. As everyone on this board knows, I love baseball. It's my favorite sport by a fair margin. On consoles I play it more than all the other sports combined. But I haven't had a PC baseball game that I've really played to death since Earl Weaver.

Actually, I've played 100s of seasons of BM. But that's more of a diversion than a game.

I can think of my progression of baseball sims like this:

Earl Weaver => Tony Larussa II => Baseball Mogul (the original) => ?

I've had so many other games (FPS baseball, other Larussa versions, OOTP, WSB, etc.), but haven't had a satisfying game since Baseball Mogul. And that seems even more bizarre to me because BM didn't even have L/R splits. It just did so many other things well and was so easy to play, that I loved it. Now, my expectations are higher, and I'm hoping this OOTP version will be close to meeting them.

[FBPro]

BM was a nice "first gen" game for 1997, too bad it stayed in 1997.

[CraigSca]

Originally Posted by John Galt

I still don't know what "stuff" and "movement" are really about - aren't they really the same, and if not, how is "stuff" different than "velocity" (especially for fastball pitchers).

I always considered "stuff" and "velocity" as the same the as well, but it may just be a naming issue here.

FB used velocity, movement, command, control as the main attributes for a pitcher. The only difference between FB and OOTP is "stuff vs. command". Slightly different, but they make effect the same things. I considered control to be the ability to get it over the plate (and thus highly influence walks). Command was the ability to "place" your pitches where you wanted them (stay on the corners as opposed to grooving it right down the heart of the plate). Anyway, I've been tackling other things, but it's great to hear he's removing poise and that the DIPS system is performing as intended .

[Huckleberry]

Keep us updated, Craig. As of right now, DMB has the best statistical baseball simulator on the market, and I don't consider it all that close. If they're never going to release a career mode, which looks probable, then I'd love to see OOTP make theirs as good of a simulator.

As for the rest of this thread, I would state my opinions but they would be John Galt regurgitations, and that's just disgusting.

[Raven]

bump.

[Gastric ReFlux]

Terrific discussion, and I'm of the group that thinks OOTP will be making a huge step in the right direction.

I'm also of the anti-social group that has found the injury-generation engine sadly lacks the functionality to let gamers create Death Leagues.

[Ksyrup]

Originally Posted by Gastric ReFlux

Terrific discussion, and I'm of the group that thinks OOTP will be making a huge step in the right direction.

I'm also of the anti-social group that has found the injury-generation engine sadly lacks the functionality to let gamers create Death Leagues.

You take the good, you take the bad, you take them all and their you have...

OOTP.