Spring Training Stats - Absolutely Meaningless - Ask Luke Hughes and Matt Maloney

The games will start soon.

My challenge to the Twin Daily bloggers is to pledge not to use any data in spring training. The samples are way too small. The quality if competition way too varied. The data is absolutely meaningless.

Uncertain?

Look at last year's numbers. Luke Hughes tore it up. People on this site were arguing the Mike Hollimon had earned a job. Brian Dinkelman and JR Towles joined them with an OPS's over .900. On the pitching side, Matt Maloney, Nick Blackburn and Francisco Liriano dominated.

Still uncertain? Look at 2011 data.

That is not to say that a player can't earn a job. If he does, it must be through the eyes of the staff watching every at bat in the cage, looking at the quality of approach at the plate, seeing the range in the field during drills...

We won't see it. We can't measure it. We can pledge not to use the data that is so obviously useless. Don't agree? Let's argue it out now before we get to the games.

Blackburn, Maloney and Hughes are the poster children for Spring Training stats being meaningless. I hope we don't see Hughes redux this year with Dozier, regardless of his stats in ST they need to stick with him throughout a good portion of the season to get the best idea if he's the guy.

When you say "bloggers," do you mean the "commenters?"

I was aiming more at those who take the time and make the effort in writing articles. That energy and effort is wasted if supported by spring training data.

I think that the bloggers understand that spring training stats don't have anything to do with what happens in the regular season. That said, the Opening Day roster can be somewhat affected by them. So, I think it's more than fair to talk and write about them. It isn't to say that anyone thinks that they're meaningful, but we do want to write about who will/should make the roster, and it's one factor in how that is determined.

The statistics may be meaningless but individual matchups could be important. It is not so important how the prospect does in a game against the fellow prospect, but what they do against the top of the line players could be important. How else would you determine some semblance of readiness for the next level? Or total overmatch?

Quote:

---

Originally Posted by old nurse

The statistics may be meaningless but individual matchups could be important. It is not so important how the prospect does in a game against the fellow prospect, but what they do against the top of the line players could be important. How else would you determine some semblance of readiness for the next level? Or total overmatch?

---

If you mean what they do as measured by numbers, any numbers will be meaningless. If you mean what they do in practice, in approach, in games, against quality opposition as seen through the eyes of the staff, it will have meaning.

Quote:

---

Originally Posted by Seth Stohs

I think that the bloggers understand that spring training stats don't have anything to do with what happens in the regular season. That said, the Opening Day roster can be somewhat affected by them. So, I think it's more than fair to talk and write about them. It isn't to say that anyone thinks that they're meaningful, but we do want to write about who will/should make the roster, and it's one factor in how that is determined.

---

Spring data (the numbers) should play no role in any decision made by the Twins management.

Decisions must be made based on performance as seen through the eyes of the staff combined with longitudinal data.

I think it's completely open to interpretation. What if we get to spring training and Jeff Clement hits .423 with 8 homers in extended action... it's certainly possible he could then make it as a bench option and carry over his success to the regular season. Yes you pointed out a few examples of guys who had great springs in 2012 but bombed in the regular season, but on the flip side I'm sure there are plenty of cases where just the opposite happens. It's interesting reading and fun to think about the possibilities after four months of nothing.

Quote:

---

Originally Posted by glanzer

I think it's completely open to interpretation. What if we get to spring training and Jeff Clement hits .423 with 8 homers in extended action... it's certainly possible he could then make it as a bench option and carry over his success to the regular season. Yes you pointed out a few examples of guys who had great springs in 2012 but bombed in the regular season, but on the flip side I'm sure there are plenty of cases where just the opposite happens. It's interesting reading and fun to think about the possibilities after four months of nothing.

---

Correct, and if that scenario plays out, no one should expect that Clement will put up the same numbers in the regular season, but he may be an acceptable bench bat.

Listen, it has played out time and again that spring training statistics are meaningless. They're meaningless for all the guys who have a certain roster spot. I mean, Joe Mauer could go 0-50 this spring, and he's still going to make the team. The numbers (along with the things we can't see like the quality of swings, the practice efforts, the attitude and all that) matter for guys trying to make a final roster spot or trying to make an impression on the coaching and front office staff. If someone gets hurt early in the regular season and they need someone to call up, making a strong spring impression might be enough to get you a call up. Do you think Chris Colabello or James Beresford and Anthony Slama don't want to make an impression. One way to do that is to put up numbers when given the opportunity.

As with Maloney and Jeff Gray last year, being on the 40 man roster has a lot more to do with making the Opening Day roster than any of the stats people put up.

The case can be made that the stars won't carry over to the regular season but that doesn't mean that they aren't important for filling out a roster to some extent. If butera goes 0-30 and throws out 10% of attempted base stealers, and Herrmann (sorry for spelling error) goes 10-30 and throws out 40%, would you not be a little confused if butera makes the 25 man as a 3rd catcher?

This is a recap of my spring training. I personally have a beautiful swing (I do however miss the ball a lot but I look good doing it). I can also run down pretty much any ball hit in my general area (but I tend to make a throwing error or 2 regularly). I am also a great club house guy. You wouldn't believe the jokes I come up with off the top of my head. Gardy absolutely loves the way I get after it. At the same time hicks has better stats than I have but he's kinda awkward around the rest of the guys. And humans Gardy don't have much in common.

Should I beat hicks out for the center field job? Are you ok with me batting lead off?

Want to talk about a Spring Training Hero? Edwin Encarnacion. For the longest time it looked like his spring numbers were not telling the truth, but then the hitting coach said, pretend every day is spring training.

I just want to put a plug in for thrylos98. He's advertised that he'll be tracking all the spring stats.

The numbers have no meaning. They can't have meaning at this small sample size. They won't tell you who his a better hitter or pitcher. They won't be predictors of the upcoming season. They are utterly useless.

Players will win jobs this spring. I look forward to Parker's video analysis of the hitters and pitchers. Has Dozier changed his approach? Parker might see it in the video. The Twins staff will be watching for it. The numbers will not tell us. Whether he hits .200 or .400 this spring, it doesn't matter. The numbers will not describe whether he has improved or not.

If we are going to discuss who should win the spots on the roster, we need to be able to do it without the support of spring training numbers.

I think most organizations understand sample size. I hope the Twins do. It does worry me that they seem to base the rule 5 draft on numbers in the very short season AFL. I just hope that they are making decisions based on the eyes of their scouts and personnel and do not give any consideration to data from a small sample.

Quote:

---

Originally Posted by jorgenswest

Decisions must be made based on performance as seen through the eyes of the staff combined with longitudinal data.

---

How about latitudinal data instead?

Small Sample Size alright, but it is better than non-sample size and making (subjective) decisions based on perceived "stuff" and the way "the ball comes out of ones' hand". There are some players who are better than what the scouts thing. If you don't believe me, ask Mr Slama whose performance in actual games that mattered mirrored that of ST and really has been against Scouts' opinions.

And some players, mind you, do better in ST because they play every day. If they are used as once every third day PHs or every 4th day starters (Luke Hughes is a prime example, so is former Twin Garrett Jones btw) often fail. Some players need regular time to produce. And most of them get that during Spring.

Some pitching data mater a whole lot. Believe me. If you cannot find the plate in ST and you are throwing ball after ball after ball, you will likely not find it during regular season. And if you consistently miss the strike zone it does not matter who is at bat whether it is a little leaguer or Miguel Cabrera. If you can steal every time you go for it in ST, you will likely have a great SB% during the season. If you cannot get to fly balls in ST, you likely will not get to fly balls during the season.

So, it's a combination. And you cannot take data away when there are close fights because otherwise your opinions are subjective. It has to be a balance. Otherwise, let them make the roster now.

spring training stats in my opinion have value when used correctly, sort of like siracha sauce lol! a little bit is good, but relying to much on it makes you a crappy cook!! one could argue even more crappy then the twins bench the last two years!

Another thing regarding sample size:
the sample size of someone's performance in ST as reflected by numbers is actually greater than the sample size of someones' performance as reflected by "scouting opinions" because scouts do not see every single pitch and every single swing a player has.

Chew on that...

Major League players go up against each other for the first three innings, if that, for at least the first half of spring training; in split squad games, you don't even get that. Pitchers have shorter stints than the position players (usually).

The one area where they're a bit "meaningful" is if your postion player goes out and stinks it up .... you're going to worry if that happens.

Quote:

---

Originally Posted by thrylos98

Another thing regarding sample size:
the sample size of someone's performance in ST as reflected by numbers is actually greater than the sample size of someones' performance as reflected by "scouting opinions" because scouts do not see every single pitch and every single swing a player has.

Chew on that...

---

There is nothing to chew on. Sample is relevant to data. An at bat resulting in a single as recorded in data can result from thousands of possible events and settings. A member of the Twins staff watching that at bat gathers significantly more information than simply a check in the single column.

Our understanding of sample size and reliability of data appears to vary significantly. I appear to be the on the side of the minority.

Clement could hit 400 this spring with 8 home runs and still likely won't the team. If he did a lot of that damage the last 2 weeks of ST he might have a chance, but it really doesn't change what he is. The Twins likely have better bench options than Clement. He will go down to AAA and if the need arises and he is hitting well, he could conceivably be called up.

What the Twins will be looking for is for the young guys like Hicks, Dozier, Arcia to show what they can do. That is take good AB's, show that you hit a major league fastball, not get fooled by off speed stuff, take good routes to the ball, make the routine plays. If all that shows up in the stats, well and good. But the stats don't really matter much. Over 30 or so AB's a few scratch hits can be the difference between a 300 average and and a 250 average. A couple of home runs hit early in the spring, with the wind blowing out, against a guy still trying to locate his fastball aren't going mean much.

The Twins aren't going to worry if a guy like Harden gets kicked around his first few outings. They aren't going to be extremely impressed if some young guy strikes out the side. I expect it maybe quite difficult this spring to evaluate all the pitchers. There are some guys coming off injuries, some guys they won't want to expose to waivers, so I doubt the ST numbers will match up with who they decide to start the year with.