DIPS Engine ?

[Crapshoot]

I present you to the case of Michael Cullum. Mike's good- real good. Looking at K/BB, he's at the gold standard of the best seasons all time - as in Pedro's 1999-2000 seasons.

Overall G GS W L SV ERA IP HA R ER BB K CG CG% SHO QS QS%
18 18 10 4 0 3.83 127.0 122 63 54 7 105 5 27.8 0 10 55.6

Mike is prone to the HR - he's given up 17 HR - or averaging 1.3 per 9 IP. That being said, his BAPIP is .312 - the league average is set to .295. He has behind him a defense that is above average(6 or higher) at every position except 3rd, and an OF which often goes 9-10-9. Without Cato installed sadly, the only fielding stats I can offer is that Milwaukee leads the league in fielding percentage, and has committed the second fewest errors in the league. I'm not inclined enough to calculate range factors for the league yet...

That being said, a player with a K/BB ratio of 15! has an ERA of 3.83 (not even in the top 15 in the league) - though admittedly sample size issues exist.

League ERA in 2005: 4.69

Admittedly, this is not exactly a huge drawn out arguement - but it seems amazing to me that a player with a K/BB ratio of 15 has an ERA that while good, is hardly reflective of his dominance.

[TonyJ]

Just copying and pasting from another board:

I have a reliever in my current solo league who is rated four stars, 10/10/6 in talent (talent only, of course) and has put up some pretty damned dominating numbers in some categories, and awful numbers in others. It makes no sense. To wit, his most recent season:

126 IP, 115 H, 8 BB, 170 (!!)K. Just looking at that you would think that he was a demi-god reliever right? You would be wrong as he also allowed 20 doubles and 27 home runs which added up to 4.14 ERA. Now, really, similar to the "Dave Kingman eat your heart out" player in FJ's recent thread, this is just not at all a possibility. No pitcher that dominant with control that good is going to allow that many home runs. I mean, that's .6 BB and 12.1 K/9 which is better than Randy Johnson. Can you imagine RJ of the past few seasons allowing 45 homers in a season? No, because a pitcher with dominant stuff and great control doesn't give up a lot of home runs.

Damnit, this game can be so frustrating.

[Joshv02]

Originally Posted by Aadik

I present you to the case of Michael Cullum. Mike's good- real good. Looking at K/BB, he's at the gold standard of the best seasons all time - as in Pedro's 1999-2000 seasons.

Overall G GS W L SV ERA IP HA R ER BB K CG CG% SHO QS QS%
18 18 10 4 0 3.83 127.0 122 63 54 7 105 5 27.8 0 10 55.6

Mike is prone to the HR - he's given up 17 HR - or averaging 1.3 per 9 IP. That being said, his BAPIP is .312 - the league average is set to .295. He has behind him a defense that is above average(6 or higher) at every position except 3rd, and an OF which often goes 9-10-9. Without Cato installed sadly, the only fielding stats I can offer is that Milwaukee leads the league in fielding percentage, and has committed the second fewest errors in the league. I'm not inclined enough to calculate range factors for the league yet...

That being said, a player with a K/BB ratio of 15! has an ERA of 3.83 (not even in the top 15 in the league) - though admittedly sample size issues exist.

League ERA in 2005: 4.69

Admittedly, this is not exactly a huge drawn out arguement - but it seems amazing to me that a player with a K/BB ratio of 15 has an ERA that while good, is hardly reflective of his dominance.

His FIP era is 3.45. An ERA of 3.83 is not unreasonable, even given the small sample size, etc.

EDIT: that assumes, of course, that 3.2 is the right additional. But, ya get my point, anyway.

EDIT2: ahh, I get the point that TonyJ makes. Yeah, I don’t have any good idea if pitchers with good SO rates also have good HR rates. I always just assumed it made sense to be that way. I guess I can see it could go either way.

[sebastian0622]

I wonder: How much do K/9 rates and BB/9 rates correlate to HR/9 rates? Is there evidence that HR/9 is truly independent of them?

[TonyJ]

Originally Posted by Joshv02

ahh, I get the point that TonyJ makes. Yeah, I don’t have any good idea if pitchers with good SO rates also have good HR rates. I always just assumed it made sense to be that way. I guess I can see it could go either way.

Well it's not just the strikout rate, but also the walk rate. Basically, not only is he striking out well over a batter an inning, he also doesn't walk anybody. That means two things: that he throws strikes and that he is tough to hit. I have a hard time imagining such a pitcher having such an awful home run rate.

[Joshv02]

Originally Posted by TonyJ

Well it's not just the strikout rate, but also the walk rate. Basically, not only is he striking out well over a batter an inning, he also doesn't walk anybody. That means two things: that he throws strikes and that he is tough to hit. I have a hard time imagining such a pitcher having such an awful home run rate.

Yeah, but it could also mean that he throws the ball over the plate (i.e., strikes) and he throws hard -- which, I could imagine, leads to high HR rates. I haven't looked it up, I will now, but I am thinking of a BJ Ryan.

[Joshv02]

Originally Posted by Joshv02

Yeah, but it could also mean that he throws the ball over the plate (i.e., strikes) and he throws hard -- which, I could imagine, leads to high HR rates. I haven't looked it up, I will now, but I am thinking of a BJ Ryan.

And, I'd be wrong -- the high HR rate I rememered was when he also had a high BB rate. When his BB rate went down, his HR rate did, too.

[TonyJ]

Originally Posted by Joshv02

Yeah, but it could also mean that he throws the ball over the plate (i.e., strikes) and he throws hard -- which, I could imagine, leads to high HR rates. I haven't looked it up, I will now, but I am thinking of a BJ Ryan.

Looks like Ryan strikes out batters at a comparable rate, but he doesn't have a very good walk rate at all. The best comp I could find for the pitcher above was Randy Johnson, and Johnson will give up an occasional home run since if you manage to hit his fastball squarely it is going to go a long way, but nothing close to my reliever.

Edit- You got to it first

[andymac]

Originally Posted by Joshv02

And, I'd be wrong -- the high HR rate I rememered was when he also had a high BB rate. When his BB rate went down, his HR rate did, too.

Yea, it would seem to make sense that if someone is accurate in the area of throwing ball & strikes, they should be accurate within the strike zone also. Meaning this guy doesn't walk anyone so he should have the ability to hit his spots within the strike zone, he's not just chucking it down the middle.

[TonyJ]

Originally Posted by andymac

Yea, it would seem to make sense that if someone is accurate in the area of throwing ball & strikes, they should be accurate within the strike zone also. Meaning this guy doesn't walk anyone so he should have the ability to hit his spots within the strike zone, he's not just chucking it down the middle.

Yes, bingo exactly on the money. He has something like Greg Maddux control and Randy Johnson k rate.

[andymac]

Originally Posted by TonyJ

Yes, bingo exactly on the money. He has something like Greg Maddux control and Randy Johnson k rate.

Which I guess would lead us to believe that OOTP probably places too much weight on the movement rating.

[Joshv02]

This is a great find. I think this could be a concern. We'd first need to find the real relationship b/w SO/BB rates and HR rate. Then, compare that to what the game produces.

[Crapshoot]

I'm trying to recall- I think Tango (or Tippett when he discovered the actual limits of DIPS) looked at this relationship.

[lynchjm24]

Originally Posted by Aadik

I present you to the case of Michael Cullum. Mike's good- real good. Looking at K/BB, he's at the gold standard of the best seasons all time - as in Pedro's 1999-2000 seasons.

Overall G GS W L SV ERA IP HA R ER BB K CG CG% SHO QS QS%
18 18 10 4 0 3.83 127.0 122 63 54 7 105 5 27.8 0 10 55.6

Mike is prone to the HR - he's given up 17 HR - or averaging 1.3 per 9 IP. That being said, his BAPIP is .312 - the league average is set to .295. He has behind him a defense that is above average(6 or higher) at every position except 3rd, and an OF which often goes 9-10-9. Without Cato installed sadly, the only fielding stats I can offer is that Milwaukee leads the league in fielding percentage, and has committed the second fewest errors in the league. I'm not inclined enough to calculate range factors for the league yet...

That being said, a player with a K/BB ratio of 15! has an ERA of 3.83 (not even in the top 15 in the league) - though admittedly sample size issues exist.

League ERA in 2005: 4.69

Admittedly, this is not exactly a huge drawn out arguement - but it seems amazing to me that a player with a K/BB ratio of 15 has an ERA that while good, is hardly reflective of his dominance.

That's a pretty small sample of innings. What do you have Milwaukee's HR Park Factor set up at and where has he made his road starts? If they are 9 home starts and at Cinci, 2 at Colorado it might just be the fact it's only 18 starts/120+IP.

[rem]

Originally Posted by Joshv02

Yeah, but it could also mean that he throws the ball over the plate (i.e., strikes) and he throws hard -- which, I could imagine, leads to high HR rates. I haven't looked it up, I will now, but I am thinking of a BJ Ryan.

eek - extreme rim shot

[Crapshoot]

Originally Posted by lynchjm24

That's a pretty small sample of innings. What do you have Milwaukee's HR Park Factor set up at and where has he made his road starts? If they are 9 home starts and at Cinci, 2 at Colorado it might just be the fact it's only 18 starts/120+IP.

Milwaukee has a park factor of 102 for HR, 103 for average. Cullum's road stats are an ERA of 3.12, 69.1 IP,5/59 BB/K ratio (3/60 at home, where he the higher 4.15 ERA in 73.2 IP).


Update- he's pitched 3 straight complete games and now has an ERA of 3.65.

[gmo]

Originally Posted by andymac

Which I guess would lead us to believe that OOTP probably places too much weight on the movement rating.

I remember this being noted with V6 (at least earlier, maybe recent patches have addressed it as I have not examined it closely) when it comes to some pitchers being too good at not allowing HRs. People have noted examples and I have seen them in my solo league where pitchers would go entire seasons without allowing a HR. Perhaps with a great reliever you might see that once in a while in RL, but in the game it could be seen with SPs and multiple relievers and over consecutive seasons.

[TonyJ]

Originally Posted by gmo

I remember this being noted with V6 (at least earlier, maybe recent patches have addressed it as I have not examined it closely) when it comes to some pitchers being too good at not allowing HRs. People have noted examples and I have seen them in my solo league where pitchers would go entire seasons without allowing a HR. Perhaps with a great reliever you might see that once in a while in RL, but in the game it could be seen with SPs and multiple relievers and over consecutive seasons.

In other words, the home runs allowed ability may be all messed up. It would be interesting to see the results of a serious study of this issue.

[gmo]

Originally Posted by TonyJ

In other words, the home runs allowed ability may be all messed up. It would be interesting to see the results of a serious study of this issue.

Your dissatisfaction may vary, but yeah, to me it seems both tails are a bit too heavy in the sense that too many pitchers give up too few home runs and too many give up too many, either absolutely or maybe like in the example above considering the other stats. Just my impression, but I don't really think it is too bad, though of course the issues at the extremes are very noticeable.

I remember at least Skipaway advocating linking GB% to HRA rating because that is about as good a correlation as you can come by. But that would require GB% becoming a real rating and not just something randomly assigned from between 30-70%.

[TonyJ]

Originally Posted by gmo

I remember at least Skipaway advocating linking GB% to HRA rating because that is about as good a correlation as you can come by. But that would require GB% becoming a real rating and not just something randomly assigned from between 30-70%.

My first post in this thread may have come off as too harsh, considering. Markus made a great improvement by switching to a DIPS based system in v6, something that neither Baseball Mogul or Pure Sim has done, to my knowledge. That shows both guts and intelligence and there is no doubt that pitchers are modeled much better in 6 than they were in 5, or in any other career sim for that matter. DIPS was, of course, the biggest statistical breakthrough in evaluating pitchers in a great many years and, I would argue, one of the most revolutionary changes in evaluating players of any type, period.

That said, Skipaway is on the mark. We think we know (as with all science) a few things about pitchers now. To any significant degree, pitchers have influence over walks and home runs allowed and strikouts recorded and the groundball/flyball ratio of balls in play. Adding this to the game and, perhaps, tweaking the amount of home runs allowed amongst the outliers, would move the game that much closer to perfectly modeling the real game. Perfection in evaluation is, of course, impossible but that doesn't mean that the game can't reflect the best tools we have available.

Granted, as I conceded in my previous post, the issue of home runs allowed simply begs further study. I am sure that there is some study out there of the relationship between home runs allowed and walks allowed or strikeouts recorded, and the game can surely reflect that. First, however, we have to know what we are dealing with in the game right now. I'm not a math whiz, nor do I have the time or, frankly, the will to do such an in-depth study myself. Someone smarter and more willing than I will have to handle that end of the bargain. Once we know the answer to those two questions, however, and then include a working GB/FB model, OOTP would be completely up to speed with our working knowledge of pitching.

You may say I'm a dreamer, but I doubt I'm the only one.