# Thread: Why are some baseball players inconsistent by the month? Reasons?

1. ## Re: Why are some baseball players inconsistent by the month? Reasons?

This is something the Reds would be wise to research, as the team as currently comprised is heavily streaky (past dialogues with nate aside). I would think an above average offense is one with both streaky and consistent hitters.

3. ## Re: Why are some baseball players inconsistent by the month? Reasons?

Originally Posted by Spring~Fields
I am not sure how to write or say this. But, is there a way to adjust players for that variance to maybe a more consistent core number? Kind of like saying for player X who OPS one month .1000, another month OPS .875, another month .925, then another month .773 and so forth. Is there a way to find out what they really are as a day to day player adjusted OPS? Or is that getting out there on the bizarre walk?

The law of large numbers says that the bigger your sample gets, the closer it gets to reality. Big samples = better guesses. And because a players true talent does change over time, we give more credit to the most recent observations. This is how you go about making a single best guess. But that's still pretty imprecise.

So we have confidence intervals. They basically say, we're X% sure that the real answer is within a certain range of our best guess. As the sample size increases (given a certain of confidence), the range gets smaller. It never goes away completely, but it gets smaller as we include more plate appearances. Even after a player's entire career we can't be 100% certain about his talent level, but we have a really really good idea. It's that simple.

Let's pretend we have a player and don't yet have a good guess of his true talent level (as measured by OPS). As we look at his OPS by day, we see a range from 0.000 to 4.000. The average day is .775, but the range on any day is huge. His weeks range from .250 to 2.000. Again, the average is .775, but the range is still very big -- for all we know, his true sill is anwhere in that range. His months range from .600 to 1.000. Yet again, a .775 average month, but this time the range is getting smaller. Yet, we still don't know whether he's a star or a scrub. But when we look at their years, they vary from .675 to .900. Now we're starting to get the picture. In fact, his last 3 years are .900, .700, .800. Now, because talent does change over time, we weight the most recent observations more heavily. Studies put it about 60-30-10.

Do the math and our imaginary player comes out to .780. That doesn't mean he's exactly a .780 guy, but that's our best guess. And because we're using a good chunk of data, that range is now pretty small. So we could say that we think he's a .780 OPS guy and we're 95% sure the actual level (which, again we can never know with absolute certainty) is between .740 and .820.

One day samples tell you nothing, statistically speaking. One month samples tell you very little. Half seasons tell you a bit. Full seasons tell you a good deal and multiple years give you a really good idea.

The trick is to realize that there's nothing special about the calendar, just the amount of observations you have. The 25 games from May 1 to June 1 are no more or less important in identifying his true talent level than the games from May 15 to June 15. The month cutoffs are completely arbitrary as it relates to performance.

An easy, if quite simplified, way to think about it is that if you're trying to predict the next day, week, month, or year -- those OPS's are all going to be the same based on what you think the guy's true talent is -- the average of that range or of all of those possible ranges. But how confident you can be with that guess is based on how big your sample is. And your best guess of his true talent is based on a big sample of what he's done in the recent past.

The big takeaway however should be to simply stop looking at monthly splits. They just make it confusing and don't really tell us anything about what's likely to happen in future months, be it the next month or the same month next year.

4. ## Re: Why are some baseball players inconsistent by the month? Reasons?

Baseball is a very random game, and the difference between great, good and below average players is very little (as little as 10% separates a .300 hitter from a .270 hitter to a .240 hitter).

Luck and randomness can easily account for wide variances between a player's months. It's actaully to be expected. It would be impossible for a player to hit his average every month over a long period of time. What you might expect to see is more consistency in a hitters walk rate and K rate, but the rest is too random to be consistent.

5. ## Re: Why are some baseball players inconsistent by the month? Reasons?

Originally Posted by Scrap Irony
This is something the Reds would be wise to research, as the team as currently comprised is heavily streaky (past dialogues with nate aside). I would think an above average offense is one with both streaky and consistent hitters.
I know this is a conversation that we've had many times around here, but nobody has yet that Reds players are any more streaky than any other players. It certainly may seem that away as we watch them, but that's because we're exposed to nearly every event, whereas we have more exposure to other players in their aggregate totals.

I won't dismiss the possibility that our players are streaky, but the human eyes and brain a horrible tool to assess streakiness -- we're not wired for it. So what is your definition of streaky, how do Reds players come out in that calculation and how does it compare to other players? What is the "average" level of streakiness across MLB?

6. ## Re: Why are some baseball players inconsistent by the month? Reasons?

I will say that...although there may no be way to actually prove it, Arroyo strikes me as one of the streakiest pitchers I have ever seen.

7. ## Re: Why are some baseball players inconsistent by the month? Reasons?

Originally Posted by fearofpopvol1
I will say that...although there may no be way to actually prove it, Arroyo strikes me as one of the streakiest pitchers I have ever seen.
Depends on how many pitchers you've seen in your life.

And I'm not being snarky. It really does matter.

8. ## Re: Why are some baseball players inconsistent by the month? Reasons?

Originally Posted by RedsManRick

The law of large numbers says that the bigger your sample gets, the closer it gets to reality. Big samples = better guesses. And because a players true talent does change over time, we give more credit to the most recent observations. This is how you go about making a single best guess. But that's still pretty imprecise.

So we have confidence intervals. They basically say, we're X% sure that the real answer is within a certain range of our best guess. As the sample size increases (given a certain of confidence), the range gets smaller. It never goes away completely, but it gets smaller as we include more plate appearances. Even after a player's entire career we can't be 100% certain about his talent level, but we have a really really good idea. It's that simple.

Let's pretend we have a player and don't yet have a good guess of his true talent level (as measured by OPS). As we look at his OPS by day, we see a range from 0.000 to 4.000. The average day is .775, but the range on any day is huge. His weeks range from .250 to 2.000. Again, the average is .775, but the range is still very big -- for all we know, his true sill is anwhere in that range. His months range from .600 to 1.000. Yet again, a .775 average month, but this time the range is getting smaller. Yet, we still don't know whether he's a star or a scrub. But when we look at their years, they vary from .675 to .900. Now we're starting to get the picture. In fact, his last 3 years are .900, .700, .800. Now, because talent does change over time, we weight the most recent observations more heavily. Studies put it about 60-30-10.

Do the math and our imaginary player comes out to .780. That doesn't mean he's exactly a .780 guy, but that's our best guess. And because we're using a good chunk of data, that range is now pretty small. So we could say that we think he's a .780 OPS guy and we're 95% sure the actual level (which, again we can never know with absolute certainty) is between .740 and .820.

One day samples tell you nothing, statistically speaking. One month samples tell you very little. Half seasons tell you a bit. Full seasons tell you a good deal and multiple years give you a really good idea.

The trick is to realize that there's nothing special about the calendar, just the amount of observations you have. The 25 games from May 1 to June 1 are no more or less important in identifying his true talent level than the games from May 15 to June 15. The month cutoffs are completely arbitrary as it relates to performance.

An easy, if quite simplified, way to think about it is that if you're trying to predict the next day, week, month, or year -- those OPS's are all going to be the same based on what you think the guy's true talent is -- the average of that range or of all of those possible ranges. But how confident you can be with that guess is based on how big your sample is. And your best guess of his true talent is based on a big sample of what he's done in the recent past.

The big takeaway however should be to simply stop looking at monthly splits. They just make it confusing and don't really tell us anything about what's likely to happen in future months, be it the next month or the same month next year.
RedsManRick this is really insightful and helpful. Thank you for taking the time to show us this in some very good detail. With this team then we cannot fairly asses most of the players then? Would that be correct? Yet we would have a better read on perhaps some of the older players, though there is few of them on this team presently. I am wrong at this stage in their careers if I personally draw concluscions about them, is what I am seeing here and also sense intuitively. Again like I was doing last year, I am looking at too small of samples for the current group. Right ?

They will all have to have more AB/PA over time to tell then and to be more fair.

9. ## Re: Why are some baseball players inconsistent by the month? Reasons?

Originally Posted by PuffyPig
Baseball is a very random game, and the difference between great, good and below average players is very little (as little as 10% separates a .300 hitter from a .270 hitter to a .240 hitter).

Luck and randomness can easily account for wide variances between a player's months. It's actaully to be expected. It would be impossible for a player to hit his average every month over a long period of time. What you might expect to see is more consistency in a hitters walk rate and K rate, but the rest is too random to be consistent.
This is very helpful too. What you are speaking about is something that I can't really see with an emotional set of fans eyes when looking at the team play, it is something that has to be researched and reviewed per player over a larger amount of AB/PA and time.

The guy could be hitting the ball very well and be making outs, which doesn't necessarily say that he is a poor hitter, am I understanding luck and randomness correctly? I was trying to look and see who is and who is not good with this team, but the whole story or truth, is not in the month to month is it?

10. ## Re: Why are some baseball players inconsistent by the month? Reasons?

Originally Posted by RedsManRick
I know this is a conversation that we've had many times around here, but nobody has yet that Reds players are any more streaky than any other players. It certainly may seem that away as we watch them, but that's because we're exposed to nearly every event, whereas we have more exposure to other players in their aggregate totals.

I won't dismiss the possibility that our players are streaky, but the human eyes and brain a horrible tool to assess streakiness -- we're not wired for it. So what is your definition of streaky, how do Reds players come out in that calculation and how does it compare to other players? What is the "average" level of streakiness across MLB?
So if we are saying "streaky" or "inconsistent" we are trying to imply that the players are not performing up to our personal expectation then right? So I really cannot call them streaky or inconsistent, suggesting that these players are under performing scrubs. I don't really know that then do I, unless we can get a larger sample size on these players which would call for them to get more AB/PA.

How do we end up telling what we have on this team then? Can we find some valid hope in some of them being better or more than "we think we might be seeing now"? I mean I don't think it is uncommon to be wrong about a player, speaking for myself only, they do improve or regress. These guys right now mostly have small samples.

11. ## Re: Why are some baseball players inconsistent by the month? Reasons?

Originally Posted by Spring~Fields
RedsManRick this is really insightful and helpful. Thank you for taking the time to show us this in some very good detail. With this team then we cannot fairly asses most of the players then? Would that be correct? Yet we would have a better read on perhaps some of the older players, though there is few of them on this team presently. I am wrong at this stage in their careers if I personally draw concluscions about them, is what I am seeing here and also sense intuitively. Again like I was doing last year, I am looking at too small of samples for the current group. Right ?

They will all have to have more AB/PA over time to tell then and to be more fair.
A few things:

1) We should differentiate between the concepts of predicting and projecting. Not that you've used these incorrectly, but they sometimes get used interchangeably in these conversations, which adds to the confusion. "Projecting" is taking a set of assumptions and looking at how things would play out given those. A prediction is choosing a specific projection that you believe is most likely to occur. Just an FYI.

2) At the major league level, we always have a body of work from which we can make some "best guess" about a player's true ability. Even with rookies, there are ways to translate their minor league numbers and use them to project major league performance. (e.g. a .300/.350/.450 line in AAA might translate to a .270/.325/.400 line in the majors) Admittedly, minor league translations aren't as reliable as major league numbers, but they're better than nothing. We had a pretty good idea that Votto was likely to be a better hitter than Adam Rosales.

That said, our estimation of a guy's talent level based on his track record (using something like the 60-30-10 method) is just a starting point. For a young guy, we might adjust our projections up from what his history would suggest because we know historically that young players tend to improve. Or we might adjust down for a guy who is aging and whom history tells us is likely to decline. But for any player, the most likely thing for him to do in the near future is perform at his talent level. So we look back at the last few years, weight them by recency, and that gives us our best guess for his current talent level.

I think it's wise to always to take our understanding of a player's true ability with a grain of salt. The reality is that there is a simply a lot of "noise", variation due to player talent changes over time (a guy legitimately improves or declines), varying opportunity (park effects or bad pitchers), and simply experiences of extended streaks of success and/or failure. The biggest mistake we make is is underestimating what really counts as a "small sample" -- particularly when what we're seeing confirms our predictions...

12. ## Re: Why are some baseball players inconsistent by the month? Reasons?

Originally Posted by Spring~Fields
So if we are saying "streaky" or "inconsistent" we are trying to imply that the players are not performing up to our personal expectation then right? So I really cannot call them streaky or inconsistent, suggesting that these players are under performing scrubs. I don't really know that then do I, unless we can get a larger sample size on these players which would call for them to get more AB/PA.
Well, we should clear up what streaky means. Is Aaron Harang "streaky" as a hitter because he makes lots of outs in a row? I don't think that's what people mean. They mean that whatever a guy's average stats are, he gets them in runs instead of evenly distributed. Let's use a coin as an example.

We know that there is a 50% chance of heads and a 50% chance of tails on any given toss. When we say streaky in general, we mean that we get long runs of heads and tails. But what's funny is that the way randomness works, streaks are to be expected. If you were to toss a coin 10 times, you would not expect HTHTHTHTHT. That is actually an unusual occurrence.

We should actually expect to see something with some streaks, like HTTTHHTHHHT. "More streaky" would mean something like HHHHHTTTTT.

So people see one of our guys go on a cold streak or hot streak and start saying that we have "streaky" players. Well, yeah, I suppose so. But every single player in MLB is streaky by that definition. Every single player, regardless of how talented they are, experiences long streaks of relative success and failure. If we want to say that our players are more streaky than the average MLB player, we need to find a way to define just how streaky the average player is. How many streaks of what scale and length are "normal"? Because no player, whether he's an .950 OPS star or a .650 OPS scrub puts up a .950 OPS every day, week, month, or year. What's the measure of streakiness? Until we define that, it's just a silly claim based on observations -- and the human brain is wired to see patterns everywhere, even where they're not.

How do we end up telling what we have on this team then? Can we find some valid hope in some of them being better or more than "we think we might be seeing now"? I mean I don't think it is uncommon to be wrong about a player, speaking for myself only, they do improve or regress. These guys right now mostly have small samples.
Two things: firstly don't look just at the results a guy is getting, but how he's getting them. As you said above, sometimes you hit the ball well and it just finds gloves. Over time, those things even out, but it takes a lot of time. ERA is notorious for this as well. Sometimes you get lucky and strand lots of baserunners. You can have a shiny ERA for awhile, but eventually it will catch up with you (Justin Lehr anybody?) Find the stats that actually predict future performance. That's one reason why people like FIP. It's based on the things pitchers can actually control and filters out the things they can't. And it ends up being a better predictor a guy's future ERA than his current ERA is....

Secondly, stop looking at split stats other than platoon splits which have been established over at multiple seasons, the samples are just too small and the amount of "noise" too high for them to tell us anything.

Bottom line, look at what a guy has done the past few years, in aggregate and stick with it. Player's talent levels just don't change overnight. Sure, the improve or decline, but generally speaking, that's the best predictor we've got. This is why people were upset with bringing Hairston back. Yeah, he had a great year last year. But he had a very long track record of being a .700 OPS type guy and he didn't fundamentally improve any of his skills. He simply got better results for one season. This season? He's back at his .714 OPS.

Again though, minor league translations are important and where systems like PECOTA are so handy. They make this translations and do the math for us. PECOTA (and others) even account for the fact that young guys are likely to get better and old guys decline. Sure they're not perfect, but they're better than our gut guesses. Unless we have reason to believe a good projection is missing something which has fundamentally changed a guy's talent level, they're much better than we can get by trying to pick apart every little thing we see. The human eye is great at picking apart visual details, but the human brain is too easily influenced by what it has seen recently or personally experienced to make good calculations about things that have lots of noise.

It is possible that some of our current players are underperforming? Sure. Rosales has had bad luck on balls in play. But PECOTA had him for a .236/.300/.398 50th percentile (most likely) projection. So adjust his low BABIP and you're pretty much there and still in sub replacement territory. Nix, this is who he is -- decent power, low OBP. Taveras? No surprise. Janish, this was to be expected.

Make no mistake, anybody who projected more than a .500 record from this team wasn't using reasonable projections. Just look at all of the OBP black holes and it becomes quite clear that this team needed both EE and Bruce to have solid seasons and neither did. The rest of the guys have performed pretty much as expected. The two minor bright spots, Hanigan and Dickerson struggled to get regular playing time. Nix can't get on base, and never could. The bench was poor to begin with and now it's starting everyday.

These guys might have small samples in the majors, or this year, but we know roughly what we've got. And right now, it's not pretty.

Thanks RMR.

14. ## Re: Why are some baseball players inconsistent by the month? Reasons?

Originally Posted by Dom Heffner
I just don't think the turning of the calendar effects most players.
I think it has more of an effect than most. Granted it is just a day, it is used because it is an easy way to create a split.

But I think it has more to do psychological than anything else. If you have having an awful stretch of things, the turning of a calendar may be a way to put your bad month behind you. It may be a way to start fresh, turn the page on an awful month, and go into the next month with a renewed focus.

I look at is somewhat similar with golf. When you play 18 holes why does it matter when you make the turn? Why are some golfers able to turn a great back nine after a horrible front nine? Hole 9 and 10 count the same, but when you make the turn to the back nine you can star over fresh.

15. ## Re: Why are some baseball players inconsistent by the month? Reasons?

Originally Posted by bucksfan2
I think it has more of an effect than most. Granted it is just a day, it is used because it is an easy way to create a split.
Being easy to measure doesn't make it significant.

I look at is somewhat similar with golf. When you play 18 holes why does it matter when you make the turn? Why are some golfers able to turn a great back nine after a horrible front nine? Hole 9 and 10 count the same, but when you make the turn to the back nine you can star over fresh.
Because they play better?

16. ## Re: Why are some baseball players inconsistent by the month? Reasons?

Originally Posted by nate
Because they play better?
What changes so much between hole 9 and 10? Its similar to the same thing that changes going from July 31st to August 1st. When you are able to put a bad month, 9, week, etc. behind you it enables you to start fresh. Monthly stats are easy to measure because there is a start date and and end date. Too often many cast aside psychological things because you can't measure them.

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•

 Board Moderators may, at their discretion and judgment, delete and/or edit any messages that violate any of the following guidelines: 1. Explicit references to alleged illegal or unlawful acts. 2. Graphic sexual descriptions. 3. Racial or ethnic slurs. 4. Use of edgy language (including masked profanity). 5. Direct personal attacks, flames, fights, trolling, baiting, name-calling, general nuisance, excessive player criticism or anything along those lines. 6. Posting spam. 7. Each person may have only one user account. It is fine to be critical here - that's what this board is for. But let's not beat a subject or a player to death, please. Thank you, and most importantly, enjoy yourselves!

RedsZone.com is a privately owned website and is not affiliated with the Cincinnati Reds or Major League Baseball

Contact us: Boss | GIK | cumberlandreds | dabvu2498 | Gallen5862 | Plus Plus | RedlegJake | RedsfaninMT | redsfan1995 | The Operator | Tommyjohn25