In baseball statistics, you’ll often hear the phrase: “small sample size” thrown around early in the season. Most teams in baseball have played about 15 games so far — give or take a game. 15 played games results in 9.2% of your total baseball season. At this point, if you’re a full-time position player you’ve probably accrued about 50-ish PA’s on the season. The point is that in an early season small sample size like 50 PA’s an 0-4 day can (or think of it this way, nearly 10% of your total seasonal PA’s to this point) can put a significant dent in your numbers. On the flip-side, a 4-4 day can boost things considerably.
For example, look no further than Edgar Renteria. In his non-super high BABIP years, he’s been about a +2 win player — ie: entirely serviceable and worthy of a major league roster spot. But, to start this season Renteria has largely struggled. Heading into April 21st, he was hitting a Bockcockian line of : .200/.289/.250. That’s pretty ugly no matter how you slice it. But he compiled those numbers in a scant 45 plate appearances. That’s a drop in the bucket when compared to an entire season’s worth of PA’s. Then, on the night of April 21st, Renteria went 3-4 with a grand slam and 5 RBI’s. His line then went to: .250/.327/.364 in just one night. That’s an OPS increase of nearly 130 points between two games.
That’s why it’s not a good idea to put too much stock (ie: Zito’s finally figured things out! or Fred Lewis’ mighty .500+ OBP is here to stay!) in early season performances. They are very volatile and subject to some pretty big swings. It’s not to say that a player’s first 50 PA’s (or 30 IP if you’re a starter) don’t mean anything. They do mean something — but they probably don’t mean as much as most would like.
With that out of the way, I thought it would be interesting to post some early season Small Sample Size All-Stars or SSSAS by FanGraph’s Win Values.
This All-Star team has some pretty good talent on it. But, there are a couple of SSS guys on here. I think most would be shocked to learn that Marco Scutaro is the best SS in baseball right now. His line of: .262/.410/.525 is no doubt impressive but it far exceeds anything else he’s done in his career in over 2,400 PA’s. Marco has put together that line in just 78 PA’s or about 11.1% of a full season of play for a position player. Or, to think of it another way, he’s got nearly 90% left of his season to hit, field, and run the bases.
Raul Ibanez is a very good hitter, but his terrible fielding — he’s been anywhere from -5 runs to -20 runs in LF over the last three years with no signs of improvement — has brought down his total win values to +2-2.5 over the last three seasons. In just 57 PA’s to start the season, Ibanez has already put up +1 win of value. In large part because his terrible fielding hasn’t caught up to him yet — he’s playing an average LF right now. If you extrapolated his early season +1 win total to a full season of play, he’d end up as a +12 win player. That’s not going to happen.
On the other side, here’s the list of position players by lowest WAR and position.
I’m cheating a little here to make my point. Jimmy Rollins isn’t the worst SS in all of baseball, but he’s in the top three. Some of the names on this list are legitimately bad players. Cody Ransom probably is a AAAA-type at this point. The jury is out on Felix Pie — but his defense should save him to an extent. But, Conor Jackson is coming off a +3 win season and Brian Giles has aged quite well as a player. Sure, he could be on his way out this year, but there’s no way of knowing this early in the season.
I’ll spin this back to a Giants context. For the first month or so of the season, try to relax when you’re considering player performances. There’s definitely good signs to look for (and historic context matters, too) but I’ve seen more than one Internet posting about how Tim Lincecum is broken after just two disappointing starts or how Edgar Renteria is the worst signing ever. Just be calm, relax, and things will work themselves out.
Though, it’s OK to think Eugenio Velez is pretty bad.