Photo

O's Scoring Distribution


  • Please log in to reply
31 replies to this topic

#21 BSLSteveBirrer

BSLSteveBirrer

    Soccer Analyst

  • Members
  • PipPipPipPipPip
  • 4,217 posts
  • LocationMS and ID

Posted 17 May 2018 - 12:28 PM

CA

 

Clearly my last post wasn't clear enough yet. I was not trying to dismiss offensive design. My point was the first step is to uncover if there is a potential issue. I believe I have done that. The second step would be to further evaluate that issue to see if there is a way to minimize it. So the next step would be to evaluate OBP and see if it does or does not correlate to run scoring variance.

 

If it doesn't then perhaps there is some other factor to look at. Or perhaps it isn't related to offensive construct. But if it does then there would be some evidence that the all or nothing offense has some issues that reduce Ws.

 

So I apologize for not being clear. I was just trying to determine if there was any foundation for an issue that my eye test told me that high variance (which the O's have) was a problem. Now that I have satisfied myself that they do indeed have high variance it begs further evaluation.

 

Its really Systems Engineering 101. No need to dig to what you suspect is the ultimate factor right out the gate. Start at upfront premise and see if there is anything there or not. If you find that your original premise wasn't really an issue then you can stop digging and spending resources on a non issue.

 

At this point really the next step if I was doing a systems evaluation would be to look at other teams and see if there are other teams with high variances and how that has impacted their Ws and Ls. Then I would know I was really onto something (and its more my interest and curiosity. I suspect MLB data guys have looked at this in depth). I am not suggesting at all that I have thought of some new thing.


  • BSLChrisStoner likes this

#22 Pedro Cerrano

Pedro Cerrano

    SportsGuy's Muse

  • Members
  • PipPipPipPipPip
  • 26,904 posts
  • LocationEllicott City, MD

Posted 17 May 2018 - 12:32 PM

Oh man. Belkast and his WOW offers.

There is baseball, and occasionally there are other things of note

"Now OPS sucks.  Got it."

"Making his own olive brine is peak Mackus."

"I'm too hungover to watch a loss." - McNulty

@bopper33


#23 CA-ORIOLE

CA-ORIOLE

    HOF

  • Members
  • PipPipPipPipPip
  • 5,589 posts
  • LocationSOCAL

Posted 17 May 2018 - 01:04 PM

OK Steve, that's fair enough. 


  • BSLSteveBirrer likes this

#24 Mackus

Mackus

    HOF

  • Members
  • PipPipPipPipPip
  • 35,466 posts

Posted 17 May 2018 - 02:02 PM

Steve, I think you'll find that any attempts at "reducing your variance" will end up either causing you to score more or less runs overall.  If they cause you to score more runs, they are a good thing, regardless of if your variance goes up or down.  If they cause you to score fewer runs, they are a bad thing, regardless of if your variance goes up or down.  Its runs scored, not variance, that is the determining factor. 

 

Variance would be a good tool for estimating statistically how a team that scores a certain amount of runs will perform compared to their expected W/L record.  A high variance offensive team will tend to underperform their pythagorean record.  A high variance pitching team will tend to overperform.

 

I don't think reducing variance is a good objective for going about how you try to build your team.  Nobody tries to build a team with the least amount of run scoring variance.  You try to build a team that scores the most amount of runs.



#25 CA-ORIOLE

CA-ORIOLE

    HOF

  • Members
  • PipPipPipPipPip
  • 5,589 posts
  • LocationSOCAL

Posted 17 May 2018 - 02:32 PM

I don't think reducing variance is a good objective for going about how you try to build your team.  Nobody tries to build a team with the least amount of run scoring variance.  You try to build a team that scores the most amount of runs.

I doubt it amounts to anything. Power is much cheaper than OBP. (disregarding the recent stupid contracts we gave out for Dumbo and Trumbo), So from an economic perspective I think it (variance) could provide you some insight, even if very small. 



#26 BSLSteveBirrer

BSLSteveBirrer

    Soccer Analyst

  • Members
  • PipPipPipPipPip
  • 4,217 posts
  • LocationMS and ID

Posted 17 May 2018 - 04:22 PM

Steve, I think you'll find that any attempts at "reducing your variance" will end up either causing you to score more or less runs overall.  If they cause you to score more runs, they are a good thing, regardless of if your variance goes up or down.  If they cause you to score fewer runs, they are a bad thing, regardless of if your variance goes up or down.  Its runs scored, not variance, that is the determining factor. 

 

Variance would be a good tool for estimating statistically how a team that scores a certain amount of runs will perform compared to their expected W/L record.  A high variance offensive team will tend to underperform their pythagorean record.  A high variance pitching team will tend to overperform.

 

I don't think reducing variance is a good objective for going about how you try to build your team.  Nobody tries to build a team with the least amount of run scoring variance.  You try to build a team that scores the most amount of runs.

Mackus I agree in general, that you want a team that scores the most amount of runs. But only to a point. From what I see I disagree that reducing variance isn't a good objective. Just look at the O's numbers. Take a fair number of runs away but decrease variance and they would have a good shot at more wins. Here's the fallacy with just scoring the most runs. If is irrelevant if you win 3-1 or 20-1. Now perhaps with further in depth analysis you might discover that reducing variance isn't something you can really do or that it wouldn't make a difference in the W-L totals. However, I do see enough indication here that I don't think it can be summarily dismissed.

 

At this point I simply don't have the analysis to determine that for instance increasing OBP wouldn't have an impact on reducing variance. But it dang sure might.



#27 BSLRobShields

BSLRobShields

    Sr. Orioles Analyst

  • Moderators
  • 71,695 posts
  • LocationBaltimore

Posted 17 May 2018 - 04:28 PM

None of this really matters.

What matters is that even when the Orioles were winning, they had a mediocre offense that was built poorly.

They need to improve on that.
@BSLRobShields

#28 BSLSteveBirrer

BSLSteveBirrer

    Soccer Analyst

  • Members
  • PipPipPipPipPip
  • 4,217 posts
  • LocationMS and ID

Posted 17 May 2018 - 04:40 PM

None of this really matters.

What matters is that even when the Orioles were winning, they had a mediocre offense that was built poorly.

They need to improve on that.

How do you know it doesn't matter?

How do you know they are built poorly?

What do you do differently? How can you answer this question without looking at the why they are built poorly. Can't fix what you don't understand. Is it as simple as get more OBP? Maybe.

 

Here's whats interesting. There are guys getting paid a fair bit of money to analyze stuff like this for professional sports teams. A lot of money in fact. Here we are sitting around discussing it for free.  :-P



#29 BSLRobShields

BSLRobShields

    Sr. Orioles Analyst

  • Moderators
  • 71,695 posts
  • LocationBaltimore

Posted 17 May 2018 - 05:38 PM

How do you know it doesn't matter?
How do you know they are built poorly?
What do you do differently? How can you answer this question without looking at the why they are built poorly. Can't fix what you don't understand. Is it as simple as get more OBP? Maybe.

Here's whats interesting. There are guys getting paid a fair bit of money to analyze stuff like this for professional sports teams. A lot of money in fact. Here we are sitting around discussing it for free. :-P


All that matters is scoring runs. The Orioles don’t score enough runs. You can’t score if you don’t get on base...the os suck at getting on base. There’s definitely a correlation there.
@BSLRobShields

#30 BSLRobShields

BSLRobShields

    Sr. Orioles Analyst

  • Moderators
  • 71,695 posts
  • LocationBaltimore

Posted 17 May 2018 - 05:42 PM

And btw, if you believe they are scoring enough runs, then run prevention becomes more important.
@BSLRobShields

#31 BSLSteveBirrer

BSLSteveBirrer

    Soccer Analyst

  • Members
  • PipPipPipPipPip
  • 4,217 posts
  • LocationMS and ID

Posted 17 May 2018 - 07:13 PM

Rob we will just agree to disagree. I think you are way too quickly discounting the potential that variance is a real issue. Of course, if they scored 1200 runs it wouldn't matter. But they don't and it might.



#32 BSLRobShields

BSLRobShields

    Sr. Orioles Analyst

  • Moderators
  • 71,695 posts
  • LocationBaltimore

Posted 17 May 2018 - 09:08 PM

Rob we will just agree to disagree. I think you are way too quickly discounting the potential that variance is a real issue. Of course, if they scored 1200 runs it wouldn't matter. But they don't and it might.


If you have a +80 or greater run differential, you will a lot of games. That’s the real thing you need.
@BSLRobShields




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

Partners