I was out of the country last week, which is why there was no post, and now it’s the All Star break. Rather than discussing the Nats, which could be very depressing, I’m going to go for a different topic. I mean, talking about the team, am I going to discuss how much they suck, how there are maybe 3 position players on the non-DL part of the roster that are part of the future, or how nobody knows because nobody’s watching? No thanks. So, for the 100th post of this blog first let me say thank you to everyone who reads this. I appreciate it, truly, as I would no longer be writing if people weren’t interested. Anyway, instead of the usual, someone put a comment on the Stats page asking me to explain what WARP and WARP-3 are, and I thought this might be a good diversion for Nats fans…
Who is this guy?
WARP stands for Wins Above Replacement Player, and like the stat VORP, compares a player to the imaginary “replacement-level” player. What is replacement level? Well, there is the long answer and the short answer. Here is the long answer, along with a definition of VORP. I prefer the short answer – he’s an average fielder, and as a hitter, he’s has the level of production a team would expect if they called up a minor leaguer. This is an important distinction. A replacement level player is not an average hitter, he is worse. If a team was full of average hitters, they would be a pretty good team, especially if they could pitch. If they were filled with replacement level players, they would stink. We’re not talking 60 wins here, we’re talking 30 or 40.
So, on to WARP
WARP, again, stands for Wins Above Replacement Level. It is complicated to calculate, I’ve never done it, and despite taking lots of math and stats classes in my life, I have no desire to. For those of you who are curious, it is based upon batted, fielding and pitching runs above the replacement level, as well as the number of runs needed for a win. I am content with knowing it is calculated, and in the answer lies the key. It is more useful in casual baseball conversation than VORP because it gets to the heart of how a player helps. If a player has a WARP of 5, he will provide your team, over the course of the season, with 5 more wins than a replacement level player. Despite the difficulty in calculating or defining it, this concept is beautiful in its simplicity.
Let’s look at an example: According to Baseball Prospectus, in 2007, Ryan Zimmerman had a WARP of 8.0, while the Orioles’ Melvin Mora had a WARP of 3.3. This means that the Nationals had about 4.7 more wins thanks to Zimmerman playing third over Mora. Meanwhile, Chipper Jones had a WARP of 8.5 and AL MVP third baseman Alex Rodriguez had a WARP of 10.7.
That’s confusing enough, can we stop there?
NO! There is also something called WARP-2 and WARP-3. None of these have anything to do with faster-than-light propulsion, which is, from what I understand, extremely more complicated than studying baseball (but not as hard as actually hitting a major league pitch). Baseball Prospectus uses WARP-1 to be the stat that we’ve already discussed, what we called WARP.
WARP-2 is like WARP-1, but the difficulty of the league is factored in as well.
WARP-3 factors in season length as well. This is useful not so much for someone like Ruth (who had 154 game seasons instead of 162) where there is little difference between what the total WARP would be for someone in Ruth’s time and now. It is great, though, for looking at someone like Moises Alou of the Expos in the strike shorted 1994 season, who had a WARP-1 of 6.2 but a WARP-3 of 7.9.
As you may have figured out, since 0 is replacement level, and ARod had a 10.7 last year, WARP numbers are in a concentrated spectrum. Usually the 3-5 range is what an average player would have, while the 7 or 7.5-10 range are All Star caliber players. Above 10 is pretty much the top few players in the league, and if you get above 12 or 13, it would be a pretty historically unreal season. And remember that the rest of the league factors into these numbers. In 1926, Babe Ruth hit .372/.516/.737, very similar to 2003 when Bonds hit .341/.529/.749. But while Ruth’s WARP that year was 15.3, Bonds’ in his season was only 12.7. That is because in 2003 the average player hit much better than than in 1926, causing the difference in this player-average based number