One of the great pleasures of writing about fantasy baseball is the opportunity to interact with readers, particularly when these readers have theories about the game and insight to provide.  I encountered one such reader last week, who sent David Gonos and me an email with his theories on pitcher strategy in DraftDay games.  To excerpt:

“I've done some numbers testing, nothing with remotely big enough sample size to be scientific, but it seems to me DD is one of the most punitive sites for pitching. What I mean by this is DD's pitchers scoring is unique in that when a pitcher has a bad game, it really hurts you in comparison to other sites.

Looking at the numbers posted in your article Derek, DD has a -10% for H, BB, and HBP. Other sites are -8% or -9%. Also, there is a much larger penalty for a loss. I think the cumulative effect, is when a pitcher has a bad game, the negative effect is magnified much more than on other sites…

Taking a look at David’s combined pitcher scores for the week [he tested cheap pitchers]: 7, 66, 69, 32, 82, 8, 60. This is absolutely amazing to look at. Two scores less than 10, four scores 60 or better, and ONE in between, ONE. Again I think this is due to the harsh nature of DD’s scoring. It is just not as common to have a middle to good score, think 15-20 pts on DD. Either your pitcher gets the win, which is +10, which probably means more innings, fewer hits/Rs allowed, and hits the 30 pt mark, or, gets the loss for -5, gives up more hits, runs etc, and ends up in the -5 to 10 range.”

When I was developing the format for my “Who To Target” series, I toyed with a metric that didn’t make the final cut for fear of stat overload but which examines the exact phenomenon this reader points out.  It’s essentially just an intrasite standard deviation with all sites scaled to the same scoring system, which explains how spread out or how closely clustered players' scores are on a given site.  We can call it "Boom or Bust" score for the sake of finding something catchier than "intrasite standard deviation."



You can see the results for all sites in the table at left, and you’ll see that this reader’s suspicions match up perfectly with the actual numbers. DraftDay indeed has the highest Boom or Bust score by a huge margin -- the gap between DraftDay and the next-highest site is more than double the gap between any other two sites -- which means that pitcher scores in DraftDay are widely dispersed with a relatively large chunk at the extremes and few scores in the average range.  There might be a bunch of pitchers that score either a 60 or a 0 on a given day, but there are far fewer in the 30 range than there are on other sites.

I’ve already discussed the kinds of pitchers that I think are most valuable in DraftDay when the site took its turn in the “Who To Target” series, coming to these conclusions:

“If you’re playing on DraftDay, it would be wise to target pitchers with high-powered offensive support to minimize the chance of taking a loss.  It would also be wise to shell out for the high-priced, elite pitching options who have a higher chance of throwing shutouts and complete games.  While it’s a myth that high strikeout rates drive up pitch counts, high walk rates definitely do, so you’re better off going for Cliff Lee (1.4 BB/9) or Doug Fister (1.4 BB/9) than Gio Gonzalez (3.6 BB/9) or A.J. Burnett (3.5 BB/9), since they’ll be more likely to pitch deep into games and ultimately finish them out.”

Our reader friend proposed a similar solution himself:

“I think for DD specifically, this means pitchers such as Ubaldo Jimenez -- high ERA, high Ks -- are worth much less on DD than other sites as the effect of the Ks don't offset the higher penalty for hits, runs, and losses.”

You’ll notice this is pretty much what I said, just phrased differently.  The penalties for hits and runs aren't that much higher than other sites, but the penalty for losses is significant.  Pitchers who strike out a lot of batters but who also walk a lot of batters, driving up their ERAs and subsequently increasing their chances of taking a loss, are definitely going to be less valuable on DraftDay (in relative terms) than guys with modest strikeout rates but excellent walk rates.

It's also worth noting that, given this "Boom or Bust" quality of starting pitching in DraftDay games, optimal strategy will change wildly from head-to-head contests to tournaments.  In tournaments, where you want to embrace risk and volatility, boom or bust is the exact strategy that should be employed.  Now, while all pitchers have this quality to some extent in DraftDay, it's my theory that low-priced pitchers are best utilized in tournaments since they are "boom or bustier," as it were.  If you find one with a good matchup and he turns in a start like the dirt-cheap Wily Peralta did yesterday (2 hits in 8 IP and 41 DraftDay points, which would have been 51 with any run support at all -- if only someone had recommended you select him :) -- you're going to be in excellent shape since the money you saved likely went to elite hitting.

And in DraftDay, where wins are so valuable and (moreover) losses so detrimental, all you need your cheap pitcher to do is avoid a loss and he'll probably have a pretty good stat line.  In other words, it seems as though it's a lot easier for a cheap pitcher to put a good score in DraftDay than he can on other sites.  They'll put up plenty of bad ones, but when they put up good ones, they'll be really good.

All in all, my take on optimal DraftDay strategy hasn’t changed.  Our findings today just add another piece of evidence to support this strategy.