Spring Training win-loss records, and you

Some people think that they can project how well a team will perform based on said team's win-loss record during Spring Training. Those people, by and large, are clinically insane.

Spring Training is a lot of things. It is fun. It is useful for players to get to know their teams, and teammates. It's great for working out the kinks of a full off-season of rest. It's even good for scouting purposes -- players legitimately change over the offseason sometimes. And while players tinker with those mechanical and body adjustments, sometimes things can shift that may, or may not, affect a guy's true talent levels going forward.

Mostly, Spring Training is fun. It's fun for fans, analysts, writers, and even --occasionally the players. It's certainly fun for the towns in Florida and Arizona where March games are played.

Do you know what isn't fun? Trying to project a team's overall capability based on their Spring Training win-loss record. There are several reasons for us to think that Spring Training win-loss records don't matter, and most of them fall under one big category.

Winning isn't always priority No. 1 during Spring Training games.

Devastating, I know. I'm not standing here trying to say that the teams and the players aren't playing to win -- I'd bet that it many cases, the players are giving something close to 100% on the field. Many of them are trying to win a regular-season job, make the case to stick in an organization, or playing for a callup down the line. Spring Training games are littered with prospects, non-prospects, and guys who management is trying to identify as contributors or scout.

To me, one of the most fun things about a game in Spring Training is the mish-mash of players who litter a lineup. Often, you'll see a mix of prospects, guys on the cusp of washing out, bench guys, and a few of the day-in, day-out regulars you've come to know and love. It's a rarity to see a team's all-out, Opening Day lineup take the field. And when you do see something akin to a full squad of A-listers, how many times does that entire lineup actually stay on the field for most of the game?

No, teams are trying to get reps, and see what players can offer, not win a game. ST is time for the big-league regulars to knock the rust off, get some reps, and occasionally have a rest day and glad-hand the fans. Most teams would rather see the fringy quys, the question marks, get into a game, face live pitching, show what they can do in the field, and then take the rest of the day to work. It's not about winning

Don't get me started on pitchers. Really, don't. These are guys that need to be ramped up -- starters may only pitch a couple of innings at a time early in camp, and may not break five innings at any point before April. Teams aren't likely to burn an arm the way they might in the regular season, guys are on mixed-up rest, and some pitchers may be focusing on a particular pitch or mechanical trick.

Anyways, we all know this particular bit of wisdom, right? But do we know this?

We've gotta test it.

...

Okay, so Lewie Pollis tested it three years ago here at Beyond the Box Score. You can check it out here.

...

Let's update it for today!

If you haven't taken the time to read Lewie's article above, we're going to follow the same methodology he used back in 2012, but we'll add some good stuff at the end.

The math is hardly #gorymath, but it did require a little bit of data gathering. I pulled up all of the regular-season winning percentages for each of the last five seasons (2010-2014), then plotted them against all of the Spring Training winning percentages for every team. I came up with this visualization:

As far as scatter plots go, not my favorite. (Come on, we all have a favorite scatter plot.) It's basically a mess, and you can probably guess that this means those two things -- regular-season winning percentage and ST winning percentage -- aren't correlated very closely. And you'd be right. I ran a quick correlation (to see if the two variables track together), and I got a correlation coefficient of .219.

For the record, a coefficient of .000 means no correlation, and a coefficient of 1.000 means perfect correlation. A coefficient of .219? It means there's very little likelihood that one has much to do with the other during the same season.

So yeah, that's a very weak correlation -- and it's a little different from what Lewie found three seasons past. Lewie found a correlation coefficient of .325 when comparing winning percentages between 2007 and 2011. In my sample, the correlation is weaker, but not markedly different.

I wanted to take it just a step further. We all know how much luck plays a part in the regular season -- so why not see if Spring Training wins correlate any better with a team's Pythagorean winning percentage*, instead of the team's actual winning percentage.

* - The short version: Pythagorean winning percentage was posited by Bill James, and uses a team's runs scored and runs allowed to determine how many games that team should have won. Pythagenpat is better, but I had access to B-R's Pythagorean data faster.

Did that look any better? Take a look for yourself:

Mmm ... not really. Believe it or not, the correlation coefficient for Spring Training winning percentage and same-season Pythagorean is .244. That's slightly better than what we had for ST W% against W% -- but it's still not much of a difference. It's still not much of a correlation. Just because it's better, well, that doesn't mean that it is good. Of course, Pythagorean record is -- probably -- a closer approximation of true talent than regular winning percentage, so I guess it makes some intuitive sense.

Enjoy your Spring Training. Soak up the sun, play a game of catch, talk to the players. Have a blast. Just do everyone, including yourself, a favor, and don't get super-excited when your team has a great Grapefruit League run, and don't get melancholic just because your guys are at the bottom of the Cactus League.

While there's plenty of worthwhile baseball stories during Spring Training, wins and losses aren't a tale worth telling.

. . .

Special thanks to Steven Silverman for checking my math. All statistics courtesy of FanGraphs, Baseball Prospectus, and Baseball Reference.

Bryan Grosnick is the Managing Editor of Beyond the Box Score. You can follow him on Twitter at @bgrosnick.