Monday, December 29, 2008

Marcel the Projector Can See Inside Your Soul

If you thought the only benefit of the baseball stats revolution was nerdclusters emerging twice a year from their basement lairs to prop up the stripper and ear medicine industries, you'd be wrong. The real fun: projections!

Imagine the Farmer's Almanac if it were, much like the farmer and his children, on crystal meth. Take a wide range of rate stats, age curves, monkey blood stains and voila! Bill James accurately predicts Vernon Wells's slugging percentage to within 20 points! Add all the individual projections up, you can simulate the entire season! Why even bother playing??? These numbers make for good fun and the occasional fistfight, but how much stock should we place in them? More importantly, who should we believe?

The good men and women at Bluebird Banter tackled the Bill James and ZiPS projections earlier in the year. It would seem Bill James is quite the Blue Jays booster. Fangraphs also makes the Marcel projections available, and they got my attention for a number of reasons. Mostly because they differ from the other two so profoundly on two of the most important Jays in 2009: Travis Snider and Adam Lind.

Travis SniderAt BatsHome RunsOPSwOBA
Bill James44819.804.351
Adam LindAt BatsHome RunsOPSwOBA
Bill James59723.843.365

Encouraging signs? They both agree that Travis Snider is a raking machine. Discouraging? Marcel only gives him 213 at bats, and only 374 to Lind! Does this algorithm have secret access to JP's inner sanctum? Does the mythical Marcel and tangotiger's mathematical machinations foresee a high profile DH riding into town, depriving these young studs of plate appearances? Also, why the disparity? Marcel sees Adam Lind producing 2.1 runs below average, while Bill James believes in 10 runs above! What about the 30 point difference between their projected averages on balls into play? Does Marcel believe Adam Lind will suddenly stop hitting line drives all together?

In all seriousness, this is a big year for Adam Lind. Much like Vernon Wells and Alex Rios before him, this spring presents itself as the long hyped phenom's opportunity to prove he belongs. In some respects, they all are similar hitters. Not particularly patient (though Vernon is much better at getting the bat on the ball), similar line drive and home run/fly ball rates. An outfield of the three of them wouldn't lead the league in home runs, but hardly a weak spot among them.

There is another player that comes to mind when I think of Adam Lind, a player that we can all agree would be a nice ceiling for a player like Lind: Garret Anderson. As the linked chart shows, Garret Anderson didn't really take off as a player until his 26 year old season. They're both left handed left fielders without much patience but enough pop to hold down an outfield spot. If Adam Lind can develop Garret Anderson's consistent production, I for one would be content.

So which projection do we put the most stock in? Which is the profit and which is the lunatic screaming on the streetcorner? Perhaps I will just take the optimistic route for each, project them both as 5 win players and camp out for playoff tickets. For argument's sake; Marcel projected Alex Rios as a .354 wOBA player with 18 home runs and 35 doubles. Rios actually put up 15 home runs, a not-talked-about-enough 47 doubles (tying him with Aaron Hill for the fourth highest Blue Jay total ever) and a .350 wOBA. Pretty accurate, more so than Bill James who overestimated Rios' OPS by 40 points. So we're fucked.


  1. I only half-understand about 10% of the stats and figures these projections use. I usually go with what someone famous said at one point about a place(I think Hollywood) "no one really knows anything."(or something like that. Unless Marcel knows that Lind and Snider are going to have unfortunate run-ins with garden appliances or tire irons when they reach those AB levels. I also forgot what I was talking about a long time ago.

  2. LOL Brendan, that was hilarious...thanks for the laugh!

  3. Marcel is notorious for leg breaking in order to keep his projections in line. He is also notorious for being a monkey.

  4. I just re-read my comment from yesterday...I apologize for the weird wording and hanging elipses. Not sure what was going on. I hadn't even had my meds yet.

  5. I'm pretty sure Marcel doesn't use minor league numbers, so they are completely useless for Snider...This might have something to do with why he's just over 200 innings, too:

    FAQ: "But, what about a player who's never played MLB? Where's his forecast?" That's simple. His forecast is the league mean over 200 PA, 60 IP (starter) or 25 IP (reliever).

  6. His 34% line drive rate can't help normalize his numbers either

  7. The thing about Rios wOBA is he was mechanically fucked for about a month. Thats the thing these projection system fail to take into account is injuries/abnormalities. If you subtract all of Rios PA's from the month of May, his wOBA skyrockets to around .385, which places him amonsgt the top 25-30 hitters in the league!


Send forth the witticisms from on high